Combustion Handbook - Fundamentals For Power, Marine & Industrial Applications - The Coen & Hamworthy

THE COEN & HAMWORTHY
COMBUSTION HANDBOOK
Fundamentals for Power, Marine & Industrial Applications
STEPHEN B. LONDERVILLE
CHARLES E. BAUKAL, JR.
Editors
COMBUSTION HANDBOOK
IndustrIal combustIon serIes
Series Editors:
Charles E. Baukal, Jr.
The Coen & Hamworthy Combustion Handbook:

Stephen Londerville and Charles E. Baukal, Jr.
The John Zink Hamworthy Combustion Handbook, Second Edition
Volume 1— Fundamentals
Volume II— Design and Operations
Volume II1— Applications
Industrial Burners Handbook
The John Zink Combustion Handbook
Computational Fluid Dynamics in Industrial Combustion
Charles E. Baukal, Jr., Vladimir Gershtein, and Xianming Jimmy Li
Heat Transfer in Industrial Combustion
Oxygen-Enhanced Combustion
COMBUSTION HANDBOOK
Edited by
STEPHEN B. LONDERVILLE
CHARLES E. BAUKAL, JR.
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2013 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Version Date: 20121220
International Standard Book Number-13: 978-1-4398-7334-2 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and
information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact
the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides
licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment
has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation
without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
This book would not have been possible without the feedback from the thousands of Coen and Hamworthy
customers, who have been requesting this handbook for many years. In many cases, our innovative
products were initiated by our customers’ special needs and requirements. Therefore, we humbly dedicate
this book to all our customers worldwide who we have been proud to serve for over 100 years.
Contents
List of Figures.............................................................................................................................................................................ix
List of Tables.......................................................................................................................................................................... xxiii
Foreword............................................................................................................................................................................... xxvii
Preface......................................................................................................................................................................................xxix
Acknowledgments.................................................................................................................................................................xxxi
Editors................................................................................................................................................................................... xxxiii
Contributors.......................................................................................................................................................................... xxxv
1. Introduction......................................................................................................................................................................... 1
Stephen B. Londerville, Timothy Webster, and Charles E. Baukal, Jr.
2. Engineering Fundamentals............................................................................................................................................ 15
Stephen B. Londerville
3. Combustion Fundamentals............................................................................................................................................ 25
Stephen B. Londerville, Joseph Colannino, and Charles E. Baukal, Jr.
4. Fuels..................................................................................................................................................................................... 71
John Ackland, Jeff White, Richard T. Waibel, and Stephen B. Londerville
5. Oil Atomization................................................................................................................................................................ 95
I.-Ping Chung and Stephen B. Londerville
6. Solid Fuel Combustion in Suspension........................................................................................................................115

Stephen B. Londerville and Timothy Webster
7. Heat Transfer................................................................................................................................................................... 127

Stephen B. Londerville, Charles E. Baukal, Jr., and Jay Karan
8. Fundamentals of Fluid Dynamics............................................................................................................................... 153

Stephen B. Londerville and Wes Bussman
9. CFD-Based Combustion Modeling............................................................................................................................ 183

Michael A. Lorra and Shirley X. Chen
10. Pollutant Emissions.........................................................................................................................................................211

Charles E. Baukal, Jr., I.-Ping Chung, Stephen B. Londerville, James G. Seebold, and Richard T. Waibel
11. Noise.................................................................................................................................................................................. 267

Wes Bussman, Jay Karan, Carl-Christian Hantschk, and Edwin Schorer
12. Combustion Controls, Burner Management, and Safety Systems....................................................................... 301

Vladimir Lifshits
13. Blowers for Combustion Systems............................................................................................................................... 335

John Bellovich and Jim Warren
14. Burners and Combustion Systems for Industrial and Utility Boilers................................................................. 359
Vladimir Lifshits
vii
viii Contents
15. Duct Burners.................................................................................................................................................................... 397

Peter F. Barry, Stephen L. Somers, Stephen B. Londerville, Kenneth Ahn, and Kevin Anderson
16. Air Heaters....................................................................................................................................................................... 421

Nigel Webley
17. Marine and Offshore Applications............................................................................................................................. 431

Richard Price
Appendix A: Units and Conversions................................................................................................................................. 443
Appendix B: Physical Properties of Materials................................................................................................................. 449
Appendix C: Properties of Gases and Liquids�� 465
Appendix D: Properties of Solids...................................................................................................................................... 509

List of Figures
Figure 1.1 Flow through a boiler, highlighting radiant and convection sections...................................................... 3
Front elevation of “D” type boiler—the furnace and drum locations form the letter
Figure 1.2
“D,” hence the name......................................................................................................................................... 4
Front elevation of “A” type boiler—the furnace and drum locations form the letter
Figure 1.3
“A,” hence the name......................................................................................................................................... 4
Front elevation of “O” type boiler—the furnace and drum locations form the letter
Figure 1.4
“O,” hence the name......................................................................................................................................... 4
Figure 1.5 “D” type package boiler................................................................................................................................... 5
Figure 1.6 “O” type package boiler................................................................................................................................... 5
Figure 1.7 Typical field-erected boiler.............................................................................................................................. 6
Figure 1.8 Opposed wall fired boiler................................................................................................................................ 8
Figure 1.9 TXU Electric Collin Station Power Plant, a tangential fired boiler............................................................ 9
Figure 1.10 Typical duct burner location.......................................................................................................................... 10
Figure 1.11 Ecojet burner in an HRSG.............................................................................................................................. 11
Figure 2.1 Celsius and Fahrenheit relation.................................................................................................................... 20
Figure 2.2 Velocity distribution next to a boundary.................................................................................................... 21
Figure 3.1 “O” type package boiler................................................................................................................................. 27
Figure 3.2 Carbon atom with six protons, neutrons, and electrons........................................................................... 28
Figure 3.3 Periodic table.�� 29
Composition of air by volume....................................................................................................................... 32
Figure 3.4
Species concentration versus excess air for the following fuels: (a) CH4, (b) natural gas
Figure 3.5
(c) simulated refinery gas (25% H2, 50% CH4, 25% C3H8), (d) C3H8, (e) No. 2 oil, and (f) No. 6 oil..............36
Figure 3.6 A
diabatic flame temperature versus equivalence ratio for air/H2, air/CH4, and air/C3H8
flames, where the air and fuel are at ambient temperature and pressure.............................................. 50
Figure 3.7 A
diabatic flame temperature versus air preheat temperature for stoichiometric air/H2, air/
CH4, and air/C3H8 flames, where the fuel is at ambient temperature and pressure............................. 50
Figure 3.8 A
diabatic flame temperature versus fuel preheat temperature for stoichiometric air/H2, air/
CH4, and air/C3H8 flames, where the air is at ambient temperature and pressure............................... 51
Figure 3.9 A
diabatic flame temperature versus fuel blend (CH4/H2 and CH4/N2) composition for
stoichiometric air/fuel flames, where the air and fuel are at ambient temperature and pressure..... 51
Figure 3.10 A
diabatic flame temperature versus fuel blend (CH4/H2) composition and air preheat
temperature for stoichiometric air/fuel flames, where the fuel is at ambient temperature
and pressure..................................................................................................................................................... 52
Figure 3.11 Sample Sankey diagram showing distribution of energy in a combustion system.............................. 52
Figure 3.12 A
vailable heat versus gas temperature for stoichiometric air/H2, air/CH4, and air/C3H8
flames, where the air and fuel are at ambient temperature and pressure.............................................. 53
ix
x List of Figures
Figure 3.13 A
vailable heat versus air preheat temperature for stoichiometric air/H2, air/CH4, and air/
C3H8 flames at an exhaust gas temperature of 2000°F (1100°C), where the fuel is at ambient
temperature and pressure............................................................................................................................ 54
Figure 3.14 A
vailable heat versus fuel preheat temperature for stoichiometric air/H2, air/CH4, and air/
C3H8 flames at an exhaust gas temperature of 2000°F (1100°C), where the air is at ambient
temperature and pressure............................................................................................................................ 54
Figure 3.15 Graphical representation of ignition and heat release............................................................................. 55
Figure 3.16 S
pecies concentration versus stoichiometric ratio for the following fuels: (a) CH4, (b) natural gas
(c) simulated refinery gas (25% H2, 50% CH4, 25% C3H8), (d) C3H8, (e) No. 2 oil, and (f) No. 6 oil..........59
Figure 3.17 Adiabatic equilibrium reaction process..................................................................................................... 62
Figure 3.18 A
diabatic equilibrium calculations for the predicted gas composition as a function of the
O2:CH4 stoichiometry for air/CH4 flames, where the air and CH4 are at ambient temperature
and pressure.................................................................................................................................................. 62
Figure 3.19 A
diabatic equilibrium stoichiometric calculations for the predicted gas composition of the
major species as a function of the air preheat temperature for air/CH4 flames, where the CH4
is at ambient temperature and pressure.................................................................................................... 63
Figure 3.20 A
minor species as a function of the air preheat temperature for air/CH4 flames, where the
CH4 is at ambient temperature and pressure............................................................................................ 63
Figure 3.21 A
major species as a function of the fuel preheat temperature for air/CH4 flames, where the air
Figure 3.22 A
minor species as a function of the fuel preheat temperature for air/CH4 flames, where the air
Figure 3.23 A
major species as a function of the fuel blend (H2 + CH4) composition for air/fuel flames,
where the air and fuel are at ambient temperature and pressure......................................................... 65
Figure 3.24 A
minor species as a function of the fuel blend (H2 + CH4) composition for air/fuel flames,
Figure 3.25 A
major species as a function of the fuel blend (N2 + CH4) composition for air/fuel flames,
Figure 3.26 A
minor species as a function of the fuel blend (N2 + CH4) composition for air/fuel flames,
Figure 3.27 E
quilibrium calculations for the predicted gas composition of the major species as a
function of the combustion product temperature for air/CH4 flames, where the air and fuel
are at ambient temperature and pressure................................................................................................. 67
Figure 3.28 E
quilibrium calculations for the predicted gas composition of the minor species as a
function of the combustion product temperature for air/CH4 flames, where the air and fuel
are at ambient temperature and pressure................................................................................................. 67
Figure 4.1 Simplified process flow diagram for hydrogen reforming/PSA...................................................................75
Figure 4.2 Simplified process flow diagram for flexicoking..................................................................................... 77
List of Figures xi
Figure 4.3 Burning velocity for various gaseous fuels................................................................................................. 84

Figure 4.4 Refinery flow diagram................................................................................................................................... 86
Figure 4.5 Flow diagram of UOP fluid catalytic cracking complex........................................................................... 88
Figure 4.6 Burner firing heavy oil no. 1.......................................................................................................................... 89
Figure 4.7 Burner firing heavy oil no. 2.......................................................................................................................... 90
Figure 4.8 Naphtha distillation curve............................................................................................................................ 90
Figure 4.9 Crude oil distillation curve........................................................................................................................... 91
Figure 4.10 Viscosity of fuel oils........................................................................................................................................ 92
Figure 5.1 Viscosity of mid-continent oils..................................................................................................................... 96
Figure 5.2 L
iquid disintegration of a cylindrical jet caused by wave formations on liquid surface either
by sinuous wave (a) or dilational wave (b).................................................................................................. 98
Figure 5.3 A hollow-cone swirl spray with high viscosity liquid (ν = 6 mm2/s).................................................... 100
Figure 5.4 John Zink Spray Laboratory equipped with a PDPA.............................................................................. 101
Figure 5.5 Spray angle relative to a stable oil flame................................................................................................... 101
Figure 5.6 A gun with a 90° machine angle. Its spray angle actually is about 30°................................................ 102
Figure 5.7 Patternator to collect water sprayed out of an oil gun............................................................................. 102
Figure 5.8 Patternation measurements for a gun shown in Figure 5.6.................................................................... 102
Figure 5.9 Simplex swirl atomizer................................................................................................................................ 103
Figure 5.10 Simplex swirl atomizer with return flow................................................................................................... 103
Figure 5.11 John Zink EA oil gun.................................................................................................................................... 104
Figure 5.12 John Zink MEA gun..................................................................................................................................... 104
Figure 5.13 John Zink HERO gun................................................................................................................................... 105
Figure 5.14 Y-jet atomization principle.......................................................................................................................... 105
Figure 5.15 W
DH waste aqueous gun design with one liquid exit port surrounded with eight atomizing

ports...............................................................................................................................................................................105
Figure 5.16 Patternation comparisons for HERO and WDH guns............................................................................. 106
Figure 5.17 Coen elliptical cap slots for low-NOx......................................................................................................... 106
Figure 5.18 Droplet size measurements of the MEA oil gun at different air pressures.......................................... 106
Figure 5.19 Droplet size measurements of the MEA oil gun at similar air–water differential pressures........... 107
Figure 5.20 D
roplet size measurements of MEA oil gun at the same air–water differential
pressure but different mass ratios.............................................................................................................. 107
Figure 5.21 Droplet size comparison measured by PDPA for different oil gun designs........................................ 107
Figure 5.22 Steam consumption for different oil gun designs................................................................................... 108
Figure 5.23 Steam consumption curve for a constant steam–oil differential pressure oil gun............................. 108
Figure 5.24 A typical oil gun capacity curve. It shows oil gun turndown ratio...................................................... 108
Figure 5.25 Comparison of NOx emissions for the HERO and MEA oil guns......................................................... 109
Figure 5.26 Diagram of Y-jet.............................................................................................................................................110
xii List of Figures
Figure 5.27 CFI versus CCR, RCR and asphaltenes...................................................................................................... 112

Figure 6.1 Subbituminous char burnout Coen code A = 60 and E = 17,150.............................................................119
Figure 6.2 Pet coke char burnout Coen code A = 15 and E = 19,000......................................................................... 120
Figure 6.3 Coal dust flame velocity versus equivalence ratio................................................................................... 120
Figure 6.4 Fuel introduction for conveying options................................................................................................... 122
Figure 6.5 Front of Coen biomass burner..................................................................................................................... 123
Figure 7.1 Specific heat of air and flue gas in Btu/lbmole°F versus temperature..................................................... 129
Figure 7.2 Specific heat of air and flue gas in Btu/lbm°F versus temperature......................................................... 129
Figure 7.3 Nitrogen gas mix chamber.......................................................................................................................... 130
Figure 7.4 Boiler example............................................................................................................................................... 131
Figure 7.5 Plane wall conduction.................................................................................................................................. 138
Figure 7.6 Temperature drop due to thermal contact resistance.............................................................................. 138
Figure 7.7 Equivalent thermal circuit for a series composite wall........................................................................... 139
Figure 7.8 Calculated casing heat flux losses using 0.8 emissivity for various exterior velocities...................... 144
Figure 7.9 Total emissivity of water vapor at reference state.................................................................................... 145
Figure 7.10 Total emissivity of carbon dioxide at reference state............................................................................... 146
Figure 7.11 Correction for water vapor and carbon dioxide....................................................................................... 147
Figure 7.12 Maximum emissivity versus C/H mass ratio........................................................................................... 149
Figure 8.1 Velocity profile of fluid along a solid surface. .......................................................................................... 155
Figure 8.2 U-tube manometer........................................................................................................................................ 157
Figure 8.3 Inclined manometer..................................................................................................................................... 158
Figure 8.4 Bourdon-tube gage....................................................................................................................................... 158
Figure 8.5 O
sborn Reynold’s experimental apparatus used to study the transition from laminar
to turbulent flow............................................................................................................................................ 159
Figure 8.6 Force components on fluid particle in direction of streamline.............................................................. 160
Figure 8.7 Water flowing in pipe from point 1 to 2.....................................................................................................161
Static, velocity, and total pressure measurements inside a pipe.............................................................162
Figure 8.8
Figure 8.9 Moody diagram............................................................................................................................................. 166
Figure 8.10 Coefficient loss for various fittings............................................................................................................. 168
Figure 8.11 Series flow in a conduit................................................................................................................................ 171
Figure 8.12 Common arrangement for parallel flow problems.................................................................................. 173
Figure 8.13 Complex parallel nozzle............................................................................................................................... 173
Figure 8.14 Parallel flow with spinner............................................................................................................................ 173
Figure 8.15 Combined series parallel flow in a burner.................................................................................................174
Figure 8.16 Series and parallel flow circuit diagram.................................................................................................... 175
Figure 8.17 Discharge coefficients for different internal nozzle designs...................................................................176
Figure 8.18 Factors affecting discharge coefficient........................................................................................................176
List of Figures xiii
Figure 8.19 Photograph of a gas exiting a nozzle......................................................................................................... 181

Figure 8.20 Mixing downstream of a free jet................................................................................................................ 181
Figure 8.21 General structure of a turbulent free jet.................................................................................................... 181
Figure 9.1 Investigation of an isothermal flow field................................................................................................... 184
Figure 9.2 Original topographic data........................................................................................................................... 184
Figure 9.3 R
epresentation of topographical data in a CFD model (blue showing lower elevation,
red showing higher elevation).................................................................................................................... 185
Figure 9.4 Close-up view of a burner in a test furnace.............................................................................................. 185
Figure 9.5 R
epresentation of a burner, colored by temperature (blue showing low temperatures, red high
temperatures)................................................................................................................................................. 185
Figure 9.6 Point measurement of a scalar in a turbulent flow.................................................................................. 187
Figure 9.7 Plot of the β-function for several values of Z and Z″................................................................................ 197
Figure 9.8 Representation of a luminous flame utilizing a soot model................................................................... 201
Figure 9.9 Discretized geometry of a typical burner................................................................................................. 202
Figure 9.10 Discretized geometry of a typical boiler burner...................................................................................... 203
Figure 9.11 Close-up view of primary and secondary tips......................................................................................... 203
Figure 9.12 R
endered view inside an ethylene cracker showing flow patterns near the premixed radiant
wall burners................................................................................................................................................... 205
Figure 9.13 Illustration of a flame envelope defined as an iso-contour of 2500 ppm CO....................................... 205
Figure 9.14 Illustration of combustion products indicating poor mixing between fuel and oxidizer.................. 206
Figure 9.15 Smaller combustion product envelopes indicate improved mixing between oxidizer and fuel...... 206
Figure 9.16 Velocity contours of multiburner boiler..................................................................................................... 207
Figure 9.17 Predicted temperature distribution for stabilizer wing on a duct burner element....................................207
Figure 9.18 Temperature distribution in flame for a tangential fired boiler............................................................. 207
Figure 9.19 (a) Path lines colored with velocity before improved turning vanes added. (b) Path lines
colored with velocity showing improved flow distribution with turning vanes................................ 208
Figure 9.20 (a) Windbox velocity distribution to burners before improved baffle system. (b) Improved
velocity to burners with improved baffle location................................................................................... 208
Figure 10.1 N
umber of people (in millions) living in counties with air quality concentrations above the
level of the primary (health-based) National Ambient Air Quality Standards (NAAQS) in 2008...... 213
Figure 10.2 C
omparison of growth measures (gross domestic product, vehicle miles traveled, population,
and energy consumption) and emissions (CO2 and aggregate emissions) from 1970 to 2010 in
the United States........................................................................................................................................... 213
Figure 10.3 Distribution of air pollution emissions by pollutant type and source category..................................214
Figure 10.4 Schematic of NO exiting a stack and combining with O2 to form NO2................................................ 218
Figure 10.5 Schematic of acid rain.................................................................................................................................. 218
Figure 10.6 Acid rain deterioration examples............................................................................................................... 219
Figure 10.7 Schematic of smog formation...................................................................................................................... 219
Figure 10.8 NO
x emissions in the United States between 1970 and 1999 based on the process............................ 220
xiv List of Figures
Figure 10.9 Schematic of fuel NOx formation pathways............................................................................................ 221

Figure 10.10 Adiabatic equilibrium NO as a function of equivalence ratio for air/fuel flames............................ 222
Figure 10.11 A
diabatic equilibrium NO as a function of gas temperature for stoichiometric air/fuel
flames............................................................................................................................................................ 223
Figure 10.12 A
diabatic equilibrium NO as a function of air preheat temperature for stoichiometric
air/fuel flames............................................................................................................................................. 223
Figure 10.13 A
diabatic equilibrium NO as a function of fuel preheat temperature for a stoichiometric
air/CH4 flame............................................................................................................................................... 224
Figure 10.14 A
diabatic equilibrium NO as a function of fuel composition (CH4/H2) for a stoichiometric
air/fuel flame............................................................................................................................................... 224
Figure 10.15 A
diabatic equilibrium NO as a function of fuel composition (CH4/N2) for a stoichiometric
air/fuel flame............................................................................................................................................... 225
Figure 10.16 C
onversion ratio of fuel-bound nitrogen to NO2 of various nitrogen-containing fuels as a
function of fuel–nitrogen content............................................................................................................. 225
Figure 10.17 Conversion rate of fuel-bound nitrogen to NOx for two different oil-fired burners......................... 226
Figure 10.18 Relative NOx versus air/fuel ratio for premix and diffusion flames................................................... 226
Figure 10.19 Schematic of HiTAC furnace..................................................................................................................... 227
Figure 10.20 Example of a staged fuel burner............................................................................................................... 229
Figure 10.21 Example of a staged air burner................................................................................................................. 229
Figure 10.22 Schematic of furnace gas recirculation.................................................................................................... 230
Figure 10.23 Example of a burner incorporating furnace gas recirculation............................................................. 230
Figure 10.24 Remote stage fuel tip.................................................................................................................................. 230
Figure 10.25 Illustration showing how the remote stage method provides lower NOx emissions....................... 231
Figure 10.26 Radiant wall burners firing (a) with and (b) without remote staging................................................. 231
Figure 10.27 H
istory of low NO burner development for (a) round flame burners and (b) radiant wall
burners, firing on gaseous fuels............................................................................................................... 234
Figure 10.28 COOLstar burner........................................................................................................................................ 235
Figure 10.29 Computational fluid dynamic modeling of the COOLstar burner..................................................... 235
Figure 10.30 Adiabatic equilibrium CO as a function of equivalence ratio for air/fuel flames............................ 236
Figure 10.31 A
diabatic equilibrium CO as a function of gas temperature for stoichiometric air/fuel
flames............................................................................................................................................................ 236
Figure 10.32 A
diabatic equilibrium CO as a function of air preheat temperature for stoichiometric
air/fuel flames............................................................................................................................................. 237
Figure 10.33 A
diabatic equilibrium CO as a function of fuel preheat temperature for a stoichiometric air/
CH4 flame..................................................................................................................................................... 237
Figure 10.34 A
diabatic equilibrium CO as a function of fuel composition (CH4/H2) for a stoichiometric
air/fuel flame............................................................................................................................................... 238
Figure 10.35 A
diabatic equilibrium CO as a function of fuel composition (CH4/N2) for a stoichiometric
air/fuel flame............................................................................................................................................... 238
Figure 10.36 Bacharach smoke tester included a hand pump, filter papers, and spot scale sheet........................ 240
List of Figures xv
Figure 10.37 Particulate sampling train......................................................................................................................... 241

Figure 10.38 Sampling at different isokinetic variations............................................................................................. 242
Figure 10.39 Minimum number of traverse points for particulate traverses........................................................... 243
Figure 10.40 Type S pitot tube and manometer assembly........................................................................................... 244
Figure 10.41 BERL experimental facility........................................................................................................................ 247
Figure 10.42 Convection section simulator (CSS)......................................................................................................... 248
Figure 10.43 Conventional diffusion flame burner (CDFB)......................................................................................... 249
Figure 10.44 Low-NOx diffusion flame burner (LDFB)................................................................................................ 249
Figure 10.45 C
onventional diffusion flame burner total hydrocarbon emissions versus heating value of
HC fuel mixture.......................................................................................................................................... 249
Figure 10.46 C
onventional diffusion flame burner total hydrocarbon emissions versus combustion zone
stoichiometry............................................................................................................................................... 250
Figure 10.47 C
onventional diffusion flame burner total hydrocarbon emissions versus propylene and
ethylene spikes............................................................................................................................................ 250
Figure 10.48 C
onventional diffusion flame burner total hydrocarbon emissions versus hydrogen content
of HC fuel mixture...................................................................................................................................... 250
Figure 10.49 Conventional diffusion flame burner total PAH at stack outlet.................................................251
Figure 10.50 C
onventional diffusion flame burner total PAH and benzo(a)pyrene at furnace outlet
compared to stack outlet............................................................................................................................ 251
Figure 10.51 Lagrangian jet model predictions............................................................................................................ 252
Figure 10.52 Conventional diffusion flame burner photoionization current (pA) versus theoretical air (%)...... 253
Figure 10.53 R
ange and average of emissions at the stack outlet for the conventional diffusion flame
burner........................................................................................................................................................... 254
Figure 10.54 Range of measurements of HAPs at the stack outlet for the conventional diffusion flame burner....... 255
Figure 10.55 E
missions for refinery fuel gas (16% H2, propane, natural gas) for the conventional diffusion
flame burner................................................................................................................................................ 256
Figure 10.56 R
ange of emissions for natural gas and refinery fuel gas for the conventional diffusion flame
burner and the ultralow-NOx diffusion burner..................................................................................... 256
Figure 10.57 Emission factor comparison for low-NOx burner and conventional burner...................................... 257
Figure 10.58 Total PAH emissions 4 rings and greater versus stoichiometric ratio................................................ 258
Figure 10.59 B
enzene (blue) and PAH (red) emissions versus stoichiometric ratio for the conventional
diffusion flame burner............................................................................................................................... 258
Figure 10.60 C
O (green) and PAH (red) emissions versus stoichiometric ratio for the conventional
Figure 10.61 H
C (yellow) and PAH (red) emissions versus stoichiometric ratio for the conventional
Figure 10.62 H
C, aldehyde, VOC, and PAH emissions versus stoichiometric ratio for the conventional
Figure 10.63 T
otal heavy VOC emissions versus stoichiometric ratio for the conventional diffusion flame
burner........................................................................................................................................................... 260
xvi List of Figures
Figure 10.64 Typical process heater, petroleum refinery emissions factors............................................................. 261
Figure 11.1 Community located close to an industrial plant.................................................................................... 268
Figure 11.2 Tree falling in the forest............................................................................................................................. 269
Figure 11.3 Pressure peaks and troughs. .................................................................................................................... 270
Figure 11.4 Cross section of the human ear................................................................................................................. 270
Figure 11.5 Relationship of decibels to watts.............................................................................................................. 271
Figure 11.6 Calculating SPL at a distance r.................................................................................................................. 272
Figure 11.7 Threshold of hearing in humans.............................................................................................................. 273
Figure 11.8 Threshold of hearing and threshold of pain in humans....................................................................... 273
Figure 11.9 A-weighted scale for human hearing threshold..................................................................................... 273
Figure 11.10 A-weighted burner noise curve................................................................................................................ 273
Figure 11.11 Weighting curves A, B, C, and D...............................................................................................................274
Figure 11.12 Block diagram of a sound level meter.......................................................................................................274
Figure 11.13 Same sound spectrum on three different intervals............................................................................... 275
Figure 11.14 Typical burner noise curve........................................................................................................................ 276
Figure 11.15 Nomogram for noise level addition.......................................................................................................... 277
Figure 11.16 Atmospheric attenuation for octave bands with center (geometric-mean) frequencies................... 278
Figure 11.17 Typical earplugs and muffs....................................................................................................................... 280
Figure 11.18 Test flare at John Zink test site in Tulsa, OK........................................................................................... 282
Figure 11.19 Typical noise signature emitted from a flare.......................................................................................... 282
Figure 11.20 Photograph of a high-pressure and low-pressure flare burning the same fuel................................ 283
Figure 11.21 Shadow photograph of a burning butane lighter................................................................................... 283
Figure 11.22 Engineer measuring flare noise................................................................................................................ 284
Figure 11.23 P
WL LW calculated from measured noise data, plotted versus heat release rate, Q̇combust, for
different types of industrial flares under various operating conditions............................................ 284
Figure 11.24 Predicted sound pressure field contour plots for a multipoint LRGO flare system......................... 285
Figure 11.25 S
PL emitted from a steam-assisted flare operating at normal conditions and at over-steamed
conditions..................................................................................................................................................... 286
Figure 11.26 Burner SPL normal and with instability.................................................................................................. 287
Figure 11.27 Development of orderly wave patterns within a high-speed gas jet.................................................. 288
Figure 11.28 Illustration showing the region of maximum jet mixing noise........................................................... 288
Figure 11.29 Photograph showing shock waves downstream of an air jet............................................................... 288
Figure 11.30 Screech tone emissions.............................................................................................................................. 289
Figure 11.31 Noise radiating from a valve..................................................................................................................... 289
Figure 11.32 Photograph of two enclosed flares........................................................................................................... 290
Figure 11.33 A steam-assisted flare with a muffler...................................................................................................... 291
Figure 11.34 Steam jet noise emitted with and without muffler................................................................................ 291
List of Figures xvii
Figure 11.35 E
xample for noise abatement in steam-assisted flares by reducing the amount of steam
required to ensure smokeless combustion.............................................................................................. 292
Figure 11.36 Water injected into a high-pressure flare................................................................................................ 292
Figure 11.37 Noise spectrum from a high-pressure flare with and without water injection................................ 293
Figure 11.38 Sound pressure versus frequency for a burner operating with and without a muffler................... 293
Figure 11.39 Illustration showing two different muffler designs............................................................................... 294
Figure 11.40 Illustration showing a common plenum chamber for floor burners in a furnace............................ 294
Figure 11.41 Noise emissions from a steam control valve........................................................................................... 294
Figure 11.42 Illustration used for burner noise example............................................................................................. 295
Figure 11.43 SPL spectrum for high-pressure flaring.................................................................................................. 296
Figure 11.44 Noise contributions separately based on the mathematical model..................................................... 296
Figure 11.45 Effect of distance on flare noise................................................................................................................ 297
Figure 12.1 Symbols of SAMA logic diagrams........................................................................................................... 304
Figure 12.2 Steam pressure control logic..................................................................................................................... 305
Figure 12.3 Schematic of SPP......................................................................................................................................... 307
Figure 12.4 COEN AC (adjustable characteristic) fuel valve designed for SPP controls....................................... 307
Figure 12.5 Parallel positioning of fuel and air........................................................................................................... 308
Figure 12.6 Fully metered controls with cross limiting between fuel and air....................................................... 309
Figure 12.7 Fully metered controls with parallel prepositioning and cross limiting between fuel and air...... 312
Figure 12.8 Operating ranges and NOx performance of a single zone lean premixed burner............................ 313
Figure 12.9 Operating ranges and NOx performance of a lean premixed burner with auxiliary fuel zones...... 313
Figure 12.10 Typical characteristic of a butterfly type valve....................................................................................... 317
Figure 12.11 COEN i-scan™ flame scanner.................................................................................................................... 319
Figure 12.12 Example of the setup panel of COEN i-scan™ flame scanner.............................................................. 320
Figure 12.13 A
n example of furnace pressure after the moment of flame detachment with and without
subsequent detonation for high fire operation of a typical 150,000 lb/h package boiler............ 321
Figure 12.14 Single-element feedwater control............................................................................................................. 325
Figure 12.15 Three-element feedwater control............................................................................................................. 326
Figure 12.16 Pressure transmitter (left) and pressure gauge (right).......................................................................... 327
Figure 12.17 Thermocouple.............................................................................................................................................. 327
Figure 12.18 Thermowell.................................................................................................................................................. 327
Figure 12.19 High-velocity thermocouple...................................................................................................................... 328
Figure 12.20 F
low coefficient C for square edge orifices with flange taps for larger pipes per ASME MFC
#M1989.......................................................................................................................................................... 329
Figure 12.21 View of a portable TESTO analyzer......................................................................................................... 331
Figure 12.22 E
rror in CARI approximation based on wet O2 measurements optimized for CH4 and CH4
blend with N2.............................................................................................................................................. 332
xviii List of Figures
Figure 12.23 C
orrelation of WI and approximated CARI for hydrocarbon fuels, CO, H2, and blends of CH4
with N2, CO2, and CO in 50/50 (vol.) ratio............................................................................................... 332
Figure 13.1 Centrifugal fan............................................................................................................................................ 336
Figure 13.2 Fan wheel designs....................................................................................................................................... 337
Figure 13.3 Vane axial fan.............................................................................................................................................. 337
Figure 13.4 Purge air blower on the side of a combustion chamber........................................................................ 338
Figure 13.5 Multistage high speed centrifugal blower for a landfill application................................................... 338
Figure 13.6 Fan drive arrangements for centrifugal fans AMCA standard 99-2404-03........................................ 339
Figure 13.7 Arrangement 4 fan...................................................................................................................................... 340
Figure 13.8 Basic centrifugal fan curve........................................................................................................................ 341
Figure 13.9 Basic vane axial fan curve.......................................................................................................................... 341
Figure 13.10 Basic centrifugal fan curve with horsepower......................................................................................... 342
Figure 13.11 Forward tip blade operating curve for 1780 RPM, 70°F, and 0.075 lb/ft3 density.............................. 342
Figure 13.12 Backward curved blade operating curve for 1780 RPM 70oF and 0.075 lb/ft3 density..................... 343
Figure 13.13 One primary and one backup fan in the field with ducting................................................................ 343
Figure 13.14 Six-blade vane axial fan in the field.......................................................................................................... 344
Figure 13.15 Outlet damper effects on fan performance............................................................................................. 346
Figure 13.16 Inlet damper effects on fan performance................................................................................................ 346
Figure 13.17 Centrifugal fan with inlet and outlet dampers...................................................................................... 346
Figure 13.18 Speed change effects on fan performance............................................................................................... 347
Figure 13.19 Variable and controlled pitch change effects on fan performance...................................................... 347
Figure 13.20 Close-up of variable pitch blades on a vane axial fan........................................................................... 347
Figure 13.21 Close-up of a flexible coupling.................................................................................................................. 349
Figure 13.22 Belt-driven centrifugal blower.................................................................................................................. 349
Figure 13.23 Oil lubricated bearings with reservoir. .................................................................................................. 349
Figure 13.24 Maintenance of arrangement 8 bearings................................................................................................. 350
Figure 13.25 Fan foundation............................................................................................................................................ 351
Figure 13.26 Inlet and outlet expansion joints for vibration isolation of ducting.................................................... 351
Figure 13.27 Outlet damper fan curve with horsepower............................................................................................ 354
Figure 13.28 Inlet damper fan curve with horsepower............................................................................................... 354
Figure 13.29 Speed control fan curve with horsepower.............................................................................................. 355
Figure 14.1 Approximate rate of thermal (Zeldovich) NO formation...................................................................... 362
Figure 14.2 Approximate relation of NOx reduction with FGR for fuels without FBN......................................... 364
Figure 14.3 NO
x performance of a typical Coen premix burner firing natural gas in a package boiler............ 367
Figure 14.4 Coen QLA burner performance at high fire........................................................................................... 367

Figure 14.5 Coen DAZ™ burner.................................................................................................................................... 370
Figure 14.6 Coen Variflame™ burner........................................................................................................................... 371
List of Figures xix
Figure 14.7 Hamworthy DFL® burner.......................................................................................................................... 371

Figure 14.8 Coen DAF™ burner.................................................................................................................................... 372
Figure 14.9 Coen Delta NOx™ burner.......................................................................................................................... 372
Figure 14.10 Hamworthy ECOjet® gas-only burners.................................................................................................... 373
Figure 14.11 Hamworthy ECOjet® flame........................................................................................................................ 373
Figure 14.12 Schematic of a Coen QLN™ burner..........................................................................................................374
Figure 14.13 Coen QLN burner flame with 20 ppm NOx -firing natural gas without FGR......................................374
Figure 14.14 Coen RMB™ burner equipped with an air isolation sliding barrel damper..................................... 375
Figure 14.15 Fuel risers of a Coen RMB burner............................................................................................................ 375
Figure 14.16 Enhanced images of gas-fired RMB flame.............................................................................................. 376
Figure 14.17 Coen D-RMB® burner mounted inside the windbox............................................................................. 377
Figure 14.18 Coen D-RMB™ burner performance in a large package boiler........................................................... 377
Figure 14.19 Coen QLA burner schematic..................................................................................................................... 378
Figure 14.20 Assembled Coen QLA burner (side)........................................................................................................ 378
Figure 14.21 Assembled Coen QLA burner (front)....................................................................................................... 379
Figure 14.22 Coen QLA burner flame with 7 ppm NOx (natural gas firing)............................................................ 379
Figure 14.23 Peabody LVC™ burner for firing BFG..................................................................................................... 379
Figure 14.24 M
odified Coen LCF burner for simultaneous low NOx firing of multiple fuels of variable
composition.................................................................................................................................................. 380
Figure 14.25 NO
x reduction with FGR mixed with combustion air or fuel.............................................................. 380
Figure 14.26 Large Coen QLN burners mounted inside windboxes......................................................................... 381
Figure 14.27 Coen QLN-II burner.................................................................................................................................... 382
Figure 14.28 R
ow of boilers equipped with Coen QLN-II™ burners rated to 63 × 106 and 90 × 106 Btu/h
(18.5–26 MWt)............................................................................................................................................... 382
Figure 14.29 Typical NOx performance of QLN-II™ burner with FGR..................................................................... 383
Figure 14.30 Large 350 × 106 Btu/h (100 MW) Coen DAF™ burner for firing syngas and natural gas............... 383
Figure 14.31 Flames of Coen DAF™ burner firing natural gas (a) and syngas (b).................................................. 383
Figure 14.32 Schematic of Coen Delta Power™ burner............................................................................................... 384
Figure 14.33 E
xamples of NOx reduction with air staging and FGR in utility boilers when firing natural
gas................................................................................................................................................................. 387
Figure 14.34 Effect of fuel biasing on the NOx.............................................................................................................. 387
Figure 14.35 Effect of FGR on thermal portion of NOx in different utility boilers.................................................. 388
NOx emissions firing #6 oil with 0.54% FBN.......................................................................................... 388
Figure 14.36
Figure 14.37 Low NOx natural gas flame....................................................................................................................... 390
Figure 14.38 Main components of a fixed geometry burner for a gas and oil T-fired boiler.................................. 390
Figure 14.39 Spinners for tilting burners....................................................................................................................... 390
Figure 14.40 Flame stabilizers and adjacent buckets of a tilting (a) gas- and (b) oil-fired burner......................... 391
xx List of Figures
Figure 14.41 Corner of a T-fired boiler with tilting burners with some heat damage............................................. 391
Figure 14.42 Coen warm-up gas burners....................................................................................................................... 392
Figure 14.43 Conceptual design of low CO flue gas reheat system for refinery gas firing.................................... 393
Figure 14.44 Coen ProLine™ burner flames at low and high fire operation............................................................. 393
Figure 15.1 Typical plant schematic.............................................................................................................................. 399
Figure 15.2 Cogeneration at Teesside, England........................................................................................................... 400
Figure 15.3 Combination (oil and gas) fired duct burners at Dahbol, India........................................................ 401
Figure 15.4 Typical location of duct burners in an HRSG......................................................................................... 401
Figure 15.5 Schematic of HRSG at Teesside, England................................................................................................ 402
Figure 15.6 Fluidized bed startup duct burner........................................................................................................... 402
Figure 15.7 Inline burner................................................................................................................................................ 403
Figure 15.8 Linear burner elements.............................................................................................................................. 403
Figure 15.9 Gas flame from a grid burner.................................................................................................................... 403
Figure 15.10 Oil flame from a side-fired oil gun........................................................................................................... 403
Figure 15.11 Approximate requirement for augmenting air....................................................................................... 405
Figure 15.12 Drawing of a duct burner arrangement.................................................................................................. 405
Figure 15.13 Comparison of flow variation with and without straightening device.............................................. 407
Figure 15.14 Physical model of duct burner system..................................................................................................... 407
Figure 15.15 Sample result of CFD modeling performed on an HRSG inlet duct................................................... 408
Figure 15.16 Drilled pipe duct burner............................................................................................................................ 408
Figure 15.17 Low-emission duct burner......................................................................................................................... 409
Figure 15.18 Flow patterns around flame stabilizer..................................................................................................... 410
Figure 15.19 Effect of conditions on CO formation.......................................................................................................411
Figure 15.20 Typical main gas fuel train: single element or multiple elements firing simultaneously.................414
Figure 15.21 Typical main gas fuel train: multiple elements with individual firing capability..............................414
Figure 15.22 Typical pilot gas train: single element or multiple elements firing simultaneously.......................... 415
Figure 15.23 Typical pilot gas train: multiple elements with individual firing capability...................................... 415
Figure 15.24 Typical main oil fuel train: single element...............................................................................................416
Figure 15.25 Typical main oil fuel train: multiple elements........................................................................................ 417
Figure 15.26 Typical pilot oil train: single element....................................................................................................... 417
Figure 15.27 Typical pilot oil train: multiple elements................................................................................................. 418
Figure 16.1 Peabody twin shell air heater.................................................................................................................... 422
Figure 16.2 Twin shell, twin-inlet reverse flow combustion air................................................................................ 424
Figure 16.3 Twin shell single-inlet high-temperature air heater............................................................................... 425
Figure 16.4 Twin shell twin-inlet all-metal air heater................................................................................................ 425
Figure 16.5 Peabody scroll burner................................................................................................................................. 426
Figure 16.6 Pressurized air heater................................................................................................................................. 426
List of Figures xxi
Figure 16.7 Pressurized air heater with an internally mounted burner.................................................................. 427
Figure 16.8 Low NOx QL burner for air heaters.......................................................................................................... 429
Figure 17.1 Hamworthy Combustion ElectroTec® rotary-cup burner...................................................................... 433
Figure 17.2 Hamworthy Combustion DF register burner......................................................................................... 434
Figure 17.3 Heavy-fuel oil sprayer and twin-fluid Y-jet atomizer............................................................................ 434
Figure 17.4 Hamworthy Combustion HXG dual-fuel register burner..................................................................... 435
Figure 17.5 Chentronics® high-energy igniter............................................................................................................. 435
Figure 17.6 Hamworthy Combustion AMOxsafe® GCU........................................................................................... 436
Figure 17.7 AMOxsafe® GCU flow schematic............................................................................................................. 437
Figure 17.8 AMOxsafe® GCU temperature profile..................................................................................................... 438
Figure 17.9 Hamworthy Combustion dual-fuel DF register burner......................................................................... 438
Figure 17.10 Fuel gas valve enclosure for FPSO engine room boiler......................................................................... 439
Figure 17.11 Hamworthy Combustion’s triple 120 m-ton/h (132 U.S. ton/h) steam boiler module for FPSO......... 440
Figure 17.12 Hamworthy Combustion’s high-pressure steam boiler module for power generation.................... 440
Figure 17.13 Hamworthy Combustion fuel gas knock-out pot on an FSO................................................................ 441
Figure 17.14 Hamworthy Combustion DFL low NOx register burner....................................................................... 441
List of Tables
Table 2.1 Common English Units.................................................................................................................................... 16

Table 2.2 Common SI Units.............................................................................................................................................. 16
Table 2.3 Prefixes and Abbreviations............................................................................................................................. 17
Table 2.4 Dynamic Viscosity of Common Fluids Referenced at Normal Conditions............................................. 21
Table 2.5 Samples of the Two Decimal Notations........................................................................................................ 22
Table 2.6 Conversion Factors........................................................................................................................................... 22
Table 3.1 Density Comparison of Average Air and 79% N2 21% O2 Assumption.................................................... 33
Table 3.2 Common Gaseous Fuels.................................................................................................................................. 34
Table 3.3 Combustion Data for Hydrocarbons.............................................................................................................. 39
Table 3.4 Specific Gravity and Properties of Common Liquid Fuels......................................................................... 46
Table 3.5 Liquid Fuel Properties by API Gravity as well as Common Coals........................................................... 47
Table 3.6 Adiabatic Flame Temperatures....................................................................................................................... 49
Table 4.1 Example Pipeline-Quality Natural Gas......................................................................................................... 72
Table 4.2 Commercial Natural Gas Components and Typical Ranges of Composition......................................... 73
Table 4.3 Composition of a Typical Refinery Gas......................................................................................................... 74
Table 4.4 Typical Composition of Steam Reforming/PSA Tail Gas........................................................................... 75
Table 4.5 Typical Composition of Flexicoking Waste Gas........................................................................................... 76
Table 4.6 Volumetric Analysis of Typical Gaseous Fuel Mixtures............................................................................. 78
Table 4.7 Physical Constants of Typical Gaseous Fuel Mixtures................................................................................ 79
Table 4.8 Physical Constants of Typical Gaseous Fuel Mixture Components.......................................................... 80
Table 4.9 Quantitative Listing of Products Made by the U.S. Petroleum Industry................................................. 85
Table 4.10 General Fraction Boiling Points...................................................................................................................... 88
Table 4.11 Requirements for Fuel Oils (per ASTM D 396)............................................................................................. 89
Table 4.12 Typical Analysis of Different Fuel Oils......................................................................................................... 89
Table 4.13 Naphtha Elemental Analysis........................................................................................................................... 90
Table 5.1 Liquid Fuel Properties...................................................................................................................................... 97
Table 5.2 Effect of Dimensionless Parameters on Liquid Breakup Length............................................................. 100
Table 5.3 Combustion Performance of the HERO Gun.............................................................................................. 109
Table 5.4 Combustion Test Results for the MEA Oil Gun......................................................................................... 109
Table 5.5 Effect of Operational Parameters on Pollutant Emissions........................................................................ 109
Table 6.1 Coal Analyses on As-Received Basis............................................................................................................116
Table 6.2 Class of Coals...................................................................................................................................................116
Table 6.3 Fuel Properties of Biomass and Solid Fuels.................................................................................................117
xxiii
xxiv List of Tables
Table 7.1 Specific Heat Curve Fit Equation.................................................................................................................. 128

Table 7.2 Enthalpy of N2 Gas as Function of Temperature........................................................................................ 130
Table 7.3 Enthalpy Values for Saturated Steam and Water Based on Temperature............................................... 131
Table 7.4 Enthalpy Values for Saturated Steam and Water Based on Pressure...................................................... 131
Table 7.5 Thermal Conductivity of Common Materials............................................................................................ 137
Table 7.6 Properties of Various Substances at above 32°F (0°C)................................................................................ 137
Table 7.7 Typical Convective Heat Transfer Coefficients........................................................................................... 139
Table 7.8 Normal Emissivities, ε, for Various Surfaces.............................................................................................. 147
Table 7.9 Mean Beam Lengths Le for Various Gas Geometries................................................................................. 147
Table 8.1 Fluid Properties of Various Gases................................................................................................................ 154
Table 8.2 Specific Weight and Viscosity of Various Fluids at Atmospheric Pressure........................................... 156
Table 8.3 Equivalent Roughness for New Pipes......................................................................................................... 166
Table 9.1 Universal “Empirical” Constants Used in the Standard k−ε Turbulence Model................................... 188
Table 9.2 Cartesian Differential Equation Set............................................................................................................. 189
Table 9.3 Cylindrical Differential Equation Set.......................................................................................................... 190
Table 9.4 Discrete Ordinates for the SN-Approximation (N = 2, 4, and 6)............................................................... 200
Table 10.1 U.S. National Ambient Air Quality Standards (NAAQS) as of October 2011........................................ 212
Table 10.2 Combustion Emission Factors (lb/106 Btu) by Fuel Type...........................................................................214
Table 10.3 NOx Emission Factors by Fuel Type............................................................................................................. 221
Table 10.4 Uncontrolled NOx Emission Factors for Typical Process Heaters............................................................ 221
Table 10.5 NOx Reductions for Different Low-NOx Burner Types............................................................................. 233
Table 10.6 Location of Traverse Points in Circular Stacks........................................................................................... 243
Table 11.1 Speed of Sound in Different Media and at Different Temperatures....................................................... 270
Table 11.2 The 10 Octave Bands....................................................................................................................................... 271
Table 11.3 Octave and One-Third Octave Bands.......................................................................................................... 275
Table 11.4 A-Weighting of the Burner Sound Curve from Figure 11.14..................................................................... 276
Table 11.5 Addition Rules................................................................................................................................................. 277
Table 11.6 Sound Levels of Various Sources.................................................................................................................. 280
Table 11.7 OSHA Permissible Noise Exposure............................................................................................................. 280
Table 11.8 Calculation of the Typical Combustion Noise Spectrum of a Stable Burning Flare from the
Overall Sound Pressure Level (OASPL)....................................................................................................... 283
Table 11.9 OASPL Determined Experimentally and Using the Mathematical Model............................................ 297
Table 12.1 Difference in C with Flange Taps between a 20 in. (50 cm) Diameter Pipe and 2 in. (5 cm)
Diameter Pipe (Independent of Re).............................................................................................................. 329
Table 13.1 Relative Characteristics of Centrifugal Blowers......................................................................................... 338
Table 13.2 Effects of Temperature and Pressure on Volume and Horsepower........................................................ 340
Table 13.3 Effects of Density on Horsepower................................................................................................................ 341
List of Tables xxv
Table 13.4 Fan Bearing Vibration Limits........................................................................................................................ 350

Table 13.5 Fan Vibration Diagnostic Clues.................................................................................................................... 352
Table 13.6 Control Options Relative to Design Rate.................................................................................................... 353
Table 13.7 Potential Controls Cost Savings.................................................................................................................... 353
Table 13.8 Blower Trouble Shooting Guide.................................................................................................................... 356
Table 15.1 Typical NOx and CO Emissions from Duct Burners.................................................................................. 413
Foreword
Combustion has routinely been defined as the rapid involved in applications of full-size combustion sys-
oxidation of a fuel producing an exothermic reaction.1,2 tems. The applications are provided with state-of-the-art
A first order of complexity can be added to this definition solutions. Several practical and solvable examples are
as two solutions of the Rankine–Hugoniot equations, also provided.
resulting in detonation and deflagration combustions.3 Several chapters contain significant discussions on
Detonation combustion results in a shock wave of super- emissions. The formation, reduction, and prediction
sonic velocities and can loosely be described as an explo- of emissions from combustion systems are examined
sion. Deflagration combustion is a very fast process and in detail. The impact of external variables is also dis-
results in subsonic flame v elocities. The e ditors of this cussed. The reader can thus make intelligent choices on
book restrict the discussion to subsonic deflagration. fuels, burners, and combustion chambers and clearly
A primary division of combustion categories relevant understand the impacts of the many variables.
to our target reader is premixed and non-premixed
flames, both of which are considered in this book. Stephen B. Londerville
A further subcategory is turbulent and laminar flames.
Industrial flames are generally designed for turbulent
flows, but can be laminar at turndown or near the surface
of small solid or liquid droplets4,5; these are dealt with in
this text. Another variant to the treatment of combustion References
covered in this book is the generation of useful heat via
combustion. Efficiencies and computing the amount of 1. C.E. Baukal (ed.), The John Zink Combustion Handbook,
useful heat absorbed are discussed as they are critical to CRC Press, Boca Raton, FL, 2001.
2. A.H. Lefebvre, Gas Turbine Combustion, 2nd edn.,
the usefulness of combustion in industrial applications.
Taylor & Francis, New York, 1999, p. 33.
The rigorous treatment of combustion can at times be 3. F.A. Willams, Combustion Theory, 2nd edn., Perseus
so complex that the kinetic variables, fluid turbulence Books, Reading, MA, 1985.
factors, luminosity, and other factors cannot be defined 4. A. Linan and F.A. Williams, Fundamental Aspects of
well enough to find a realistic solution. It is our inten- Combustion, Oxford University Press, New York, 1993.
tion in this text to simplify the processes and not to cre- 5. L.D. Smoot and P.J. Smith, Coal Combustion and
ate more complexities. This book is written for those Gasification, Plenum Press, New York, 1985.
xxvii
Preface
The last Coen & Hamworthy Combustion Handbook was • Control all the flows and operational parameters
written by Arthur H. Light in 1920.1 In the late 1980s, • Design a burner/combustion chamber to achieve
Coen authorized the publication of a new combustion performance levels such as emissions and heat
handbook at the request of its customers, but day-to-day transfer
work never allowed time for the book.
• Avoid excessive noise and vibration and pro-
The origins of this book started during a business trip.
vide long, durable equipment life under adverse
In early 2000, one of the editors (Steve) was on a trip to
conditions
visit a major Coen customer with a young Coen appli-
cation engineer. During dinner, the engineer told Steve
As a result, the topics in this book include units, chem-
a story of a number of engineers debating the units of
istry, fluid flow, heat transfer, atomization, solid fuels
molecular weight for hours. It was at this moment that
including handling, liquid and gaseous fuels, pollution
Steve decided a unique kind of book was needed that
emissions, CFD, noise and practical discussions on con-
did not exist for combustion engineers. The book needed
trols, auxiliary support, and burner selection criteria.
to be all inclusive and cover the very basics in building
This book is designed to be a review of the critical,
block form up to more complex combustion topics for
relevant elements of combustion science required to

everyone involved in combustion systems to use.
apply simple calculations and more advanced computa-
This book is designed for all engineers and profession-
tions. It is especially targeted at engineers and profes-
als involved in the field of industrial and utility com-
sionals in the field of combustion who need a review
bustion systems. It is roughly divided into two parts,
of fundamentals so they can make calculations and
consisting of a total of 17 chapters, sequentially cover-
decisions on proper design features, computations,
ing relevant and important combustion engineering
emissions, fuel choices, controls, burner selection, and
basics and then specific computations and applications.
burner/furnace combinations. In addition to the build-
Each chapter is roughly organized from simple to more
ing block organization, users can go directly to individ-
complex, thus allowing the reader to absorb as much as
ual chapters concerned with specific applications to get
they may need before moving on to the next chapter.
information on different applications without reading
Practical examples are also included.
the preceding chapters.
The intent is to have a ready reference combined with
a practical review needed for engineers in the field of
combustion. The practical aspects of all combustion sys-
tems include by necessity a variety of subsystems that
include, as a minimum, methods to
Reference
• Transport and introduce fuel and air to a system 1. A.H. Light, Efficient Oil Burner, Coen Company,
• Safely monitor the combustion system San Francisco, CA, 1920.
xxix
Acknowledgments
The authors collectively thank the Coen, John Zink, Londerville, for a stimulating work environment and
and Hamworthy companies for their help and support opportunity to expand combustion technology during
during the preparation of this book. Many colleagues his 23-year career with Coen and for the invitation to
helped with ideas, content, and the preparation of fig- write chapters of this book on burners and controls.
ures and tables. We especially thank Rick Ketchum, Steve Londerville acknowledges Chet Binasik for hir-
Andrew Walter, Vincent Wong, and Jeffrey Ma for their ing him at Coen in 1977 and providing a long, fulfill-
help in preparing the materials for this book. ing career in R&D; Temple Voorhees for his continuous
Chuck Baukal thanks his wife, Beth, and his daugh- guidance in the drive for new products with a passion
ters, Christine, Caitlyn, and Courtney, for their con- that was unparalleled in this industry; Sherman Eaton,
tinued support. He also thanks the good Lord above, his mentor for many years, who paved the way for his
without whom this would not have been possible. career and accomplishments; Jim Church for driving
Wes Bussman thanks his family, Brenda, Sean, and into his mind that engineers need to understand finan-
Zach, for their support. He also thanks all of his col- cials and for their many discussions; his wife Pure who
leagues at the John Zink Company for their encour- never complained about, “Sorry I have work this week-
agement and for the knowledge they have shared with end”; and finally John Baxter, his lifelong friend since
him throughout his career. Vladimir Lifshits acknowl- childhood, who always said, “Just get it done, and then
edges Coen Company, Inc. and his long-term boss, Steve we can have some fun.”
xxxi
Editors
Charles E. Baukal, Jr., PhD, is the director of the Dr. Baukal earned his PhD in mechanical engineer-
John Zink Institute for the John Zink Company, LLC ing from the University of Pennsylvania (Philadelphia,
(Tulsa, Oklahoma), where he has been since 1998. He Pennsylvania) and is a licensed professional engineer
has also been the director of R&D and the director of in the state of Pennsylvania, a board-certified environ-
the R&D Test Center at Zink. He previously worked mental engineer (BCEE), and a qualified environmental
for 13 years at Air Products and Chemicals, Inc. professional (QEP). He has served as an expert witness
(Allentown, Pennsylvania) in the areas of oxygen- in the field of combustion, has 11 U.S. patents, and is
enhanced combustion and rapid gas quenching in the a member of numerous honorary societies and Who’s
ferrous and nonferrous metals, minerals, and waste Who compilations. He is also a member of the American
incineration industries. He worked for Marsden, Inc. Society of Mechanical Engineers, the American Society
(a burner supplier in Pennsauken, New Jersey), for for Engineering Education (ASEE), and the Combustion
five years in the paper, printing, and textile indus- Institute. He serves on several advisory boards, holds
tries and for Selas Corp. (a burner supplier in Dresher, offices in the Combustion Institute and ASEE, and is a
Pennsylvania) in the metals industry, both in the area reviewer for combustion, heat transfer, environmental,
of industrial combustion equipment. He has over 30 and energy journals.
years of experience in the fields of industrial com-
bustion, pollution control, and heat transfer and has Stephen B. Londerville is currently chief engineer at
authored more than 100 publications in those areas. Coen Company, San Mateo City, California. He previ-
Dr. Baukal is an adjunct instructor at Oral Roberts ously served as chief technical officer since 1978 at Coen,
University, the University of Oklahoma, the University vice president R&D, director R&D, and chief engineer.
of Tulsa, and the University of Utah. He is the author/ During the past 35 years, he has been involved with all
editor of eight books in the field of industrial com- aspects of product development at Coen Company. He
bustion, including Oxygen-Enhanced Combustion earned his mechanical engineering degree from San
(1998), Heat Transfer in Industrial Combustion (2000), Jose State University in 1977. He holds 7 patents and
Computational Fluid Dynamics in Industrial Combustion has authored 16 publications. He is a member of ASME,
(2001), The John Zink Combustion Handbook (2001), AIChE, the Combustion Institute, Tau Beta PI, and the
Industrial Combustion Pollution and Control (2004), Institute for Liquid Atomization and Spray Systems
Handbook of Industrial Burners (2004), Heat Transfer from (ILASS) and was past officer in the board of directors
Flame Impingement Normal to a Plane Surface (2009), and at ILASS. He was recognized as Engineer of the Year by
Industrial Combustion Testing (2011). ASME, Santa Clara Valley Section.
xxxiii
Contributors
John Ackland worked at the John Zink Company, Shirley X. Chen is a senior CFD engineer at the John Zink
LLC (Tulsa, Oklahoma), and John Zink International Company’s Simulation Technology Solutions Group. She
Luxembourg Sarl for more than 13 years in various has over ten years of experience in the power generation
technical and business roles within the process burner and eight years of experience in the petrochemical
and process flare departments. He earned his BSc in industries. Her areas of expertise include radiative heat
chemical engineering from the University of Tulsa. transfer, flow and heat transfer in porous media, and
combustion simulations. She has published over 15 papers
Kenneth Ahn has worked in various technical roles at in peer-reviewed journals and conference proceedings.
the Coen Company for 23 years. He worked in research Shirley earned her PhD in aerospace and mechanical
& development for 18 years and holds two patents in engineering from the University of Oklahoma.
duct burner design. He currently serves as a senior
design engineer at the Coen Company. He received his I.-Ping Chung, PhD, is a senior development engineer
BS in mechanical engineering from the University of in the Technology and Commercial Development Group
California at Berkeley in 1988. at the John Zink Company, LLC (Tulsa, Oklahoma). She
has worked in the field of industrial combustion and
Kevin Anderson, PE, is the design engineering
equipment, fluid dynamics, atomization and sprays,
manager for Coen Company, Inc. In this capacity, Kevin
spray combustion, and laser diagnosis in combustion
is responsible for the proper design and operation of
and earned her PhD in mechanical and aerospace
Coen-supplied combustion equipment. He leads Coen’s
engineering from the University of California, Irvine.
CFD modeling efforts and has substantial experience
She has 24 publications and 9 patents and is a registered
using CFD modeling in the development of new
professional engineer of mechanical engineering in
combustion products. Kevin has worked in the field
California and Iowa states.
of industrial combustion for 16 years and received his
MS in mechanical engineering from Sacramento State
University. He is a licensed professional engineer in the Joseph Colannino is chief technology officer of Clear
state of California. Sign Combustion Corporation—a company that applies
low power electric fields to flames to improve their
Peter F. Barry is the former director of duct burners for the shape and emissions. Prior to joining ClearSign, he was
John Zink Company, LLC (Tulsa, Oklahoma). He earned director of research and development at the John Zink
his BS in mechanical engineering from Lehigh University. Company, LLC (Tulsa, Oklahoma), where he led global
R&D efforts. His responsibilities included management
John Bellovich is the manager of the Combustion of intellectual property, oversight of John Zink’s testing
Rental Group at the John Zink Company, LLC (Tulsa, facility, and leading knowledge management efforts.
Oklahoma). He earned his BSME from the University Colannino has more than 25 years of experience in the
of Tulsa and has more than 20 years of experience in combustion industry and has authored or contributed
the industrial combustion industry. He has written or to several books, including Industrial Combustion Testing,
cowritten three published articles. The Air Pollution Control Guide, The John Zink Combustion
Handbook, and Modeling of Combustion Systems—A
Wes Bussman, PhD, is a senior research and development Practical Approach. He is a registered professional
engineer for the John Zink Company, LLC (Tulsa, engineer and has written and reviewed problems
Oklahoma). He earned his PhD in mechanical engineering appearing on the NCEES professional engineering
from the University of Tulsa. He has 20 years of experience exam given in all 50 states for professional engineering
in basic scientific research work, industrial technology licensure. Colannino’s areas of expertise include R&D
research and development, and combustion design management, combustion, pollutant formation and
engineering. He holds ten patents and has authored several control, and statistical experimental design. Past and
published articles and conference papers. He has also been present memberships include the American Institute of
a contributing author to several combustion-related books Chemical Engineers, the American Chemical Society, the
and has taught engineering courses at several universities. Air and Waste Management Association, the American
He is a member of the Kappa Mu Epsilon Mathematical Statistical Association, and the National Association
Society and Sigma Xi Research Society. of Professional Engineers. He earned his BSc from the
xxxv
xxxvi Contributors
California Polytechnic University in Pomona and his he also completed his PhD. He gained experience in
master’s degree in knowledge management from the NOx reduction techniques, especially in reburning
University of Oklahoma. technology, and developed his own software code for
the computation of turbulent reacting flow problems
Carl-Christian Hantschk, PhD, has been working using laminar flamelet libraries. During his time at
as a consulting engineer in industrial acoustics for John Zink, he specialized in CFD models for thermal
Müller-BBM GmbH (Munich, Germany) since 2001. oxidizers and flare systems. He currently works as a
He was promoted to managing director in 2009. He senior process engineer/CFD specialist at John Zink
works on industrial acoustics in general, including KEU GmbH in Krefeld, Germany.
theoretical and applied acoustics, environmental
acoustics, aero-acoustics, and numerical acoustics, Richard Price is the sales director of Hamworthy
with special focus on the interdisciplinary field Combustion (United Kingdom) for the marine and
between combustion and acoustics. He holds a offshore industry and has over 18 years of experience
diploma in mechanical engineering and received in marine-related combustion applications. He earned
his doctorate in thermodynamics from the Technical his master’s degree in mechanical engineering from
University Munich, Germany. His research has Southampton University in the United Kingdom. He
focused on combustion-driven acoustic oscillations in is a chartered engineer and member of the Institute of
burners and combustion–acoustic interactions. He has Marine Engineering, Science & Technology.
also given lectures on chemical thermodynamics,
thermal radiation, and heat transfer and acoustics Edwin Schorer, PhD, has been working as a consulting
at his university, at international conferences, and engineer in industrial acoustics for Müller-BBM
for industrial clients. His work has resulted in 30 GmbH (Munich, Germany) since 1989. He earned his
publications and 4 invention disclosures. As one of degree in electrical engineering and his doctorate in
his main research projects, he codeveloped an active psychoacoustics from the Technical University Munich,
acoustic feedback control for industrial combustion Germany. He was promoted to managing director in
systems. 2006. His research interests are in industrial acoustics
in general, including theoretical and applied acoustics,
Jay Karan (Jaiwant Jayakaran) is director of thermal with special focus on noise predictions for flare
oxidizer systems at the John Zink Company, LLC noise and fan noise, fluid mechanics, ship acoustics,
(Tulsa, Oklahoma). He has over 25 years of experience and acoustic optimization of postal automation
in the fields of combustion, petrochemicals, and systems. His work has resulted in 15 publications on
power, with responsibilities in R&D, plant operations, psychoacoustics as well as industrial and technical
and engineering. Jay earned his MS in mechanical acoustics. Dr. Schorer is a member of the German
engineering. He has authored many technical articles Institute for Standardization, the Noise Control and
and papers over the years and has several patents. Vibration Engineering Standards Committee, and the
German Acoustical Society. His research focuses on a
Vladimir Lifshits is a principal development engineer functional schematic of just noticeable frequency and
of the Design Engineering Group at Coen Company, amplitude variations. He has worked as a temporary
Inc. He joined the company in 1989. For eight years up academic counsel at his university, lecturing
to 2008, he was a director of research and development electroacoustics and technical acoustics. He has also
at Coen Company. Lifshits pioneered the development acted as a supervising tutor for the student’s diploma
of low NOx burners with lean premix combustion. theses and practical trainings.
His many other developments include advanced
combustion products and systems for boilers and James G. Seebold, Chevron (retired), has more than
steam generators for enhanced oil recovery and air 40 years of experience in the design, operation, and
heaters. He has authored many technical papers and maintenance of burners, fired heaters, furnaces,
patents. He earned his MS in mechanical engineering incinerators, boilers, and flares, including low-NOx burn
and physics from St. Petersburg Polytechnic Institute ers, selective catalytic and noncatalytic NOx reduction
of Russia. systems, and hazardous air pollutant emissions. He
conceived and led the 4-year, $7 million, 20-participant,
Michael A. Lorra, PhD, has been a CFD engineer for industry–government–university collaboration that pro
the John Zink Company, LLC (Tulsa, Oklahoma), vided the basis for this volume’s chapter on pollutant
since 1999. Previous to that, he worked at Gaswaerme emissions. He earned his PhD in mechanical engineering
Institut, Essen, Germany, e.V, for eight years, where from Stanford.
Contributors xxxvii
Stephen L. Somers was a senior process engineer at vaneaxial blowers. He presently serves on the Mechanical
the John Zink Company, LLC (Tulsa, Oklahoma). He Engineering Advisory Board at the University of Tulsa.
has over 30 years of experience in combustion and
process design, with 15 of those years in the application Nigel Webley has been group technical director at
and design of duct burners for HRSG supplementary Hamworthy Combustion since 2006. He has over 30
firing. He earned his MS in chemical engineering from years of experience in combustion, including design and
the University of Oklahoma and his BS in chemical project management of thermal oxidizer systems and
engineering from the University of Tulsa. burners for large industrial and utility boilers. He earned
his BSc in environmental chemical engineering from
Richard T. Waibel, PhD, was a senior principal engineer the University of Salford in the United Kingdom and a
in the Burner Process Engineering Group at the John postgraduate diploma in management studies. He is a
Zink Company, LLC (Tulsa, Oklahoma). He worked in chartered engineer and a member of the Energy Institute.
the field of burner design and development and received
his doctorate in fuel science from The Pennsylvania State Timothy Webster is the president of Coen Company, Inc.
University. He has published over 70 technical papers, He has worked in the field of industrial combustion for
publications, and presentations. Dr. Waibel was the 18 years and earned his master’s degree in engineering
chairman of the American Flame Research Committee from the University of Wisconsin. He has over 40
for many years, starting in 1995. publications and is a licensed professional mechanical
engineer in California.
Jim Warren is the manager, Mechanical Engineering
Group, for the Thermal Oxidizer/Flare Division at the Jeff White is a senior flare design engineer at the John
John Zink Company, LLC (Tulsa, Oklahoma). Warren Zink Company, LLC (Tulsa, Oklahoma). He has worked in
earned his BSME from the University of Tulsa and has the field of flare system design at the John Zink Company
over 23 years of experience at Zink. His area of expertise for 30 years. He earned his master’s degree in mechanical
is refractory and rotating equipment. He holds API-936 engineering from the University of Texas at Austin. He
certification in Refractory Installation Quality Control and has published two articles, one on flare radiation methods
is responsible for equipment selection for centrifugal and and the other on flow measurement by ASME nozzles.
1
Introduction
Stephen B. Londerville, Timothy Webster, and Charles E. Baukal, Jr.
CONTENTS
1.1 Introduction........................................................................................................................................................................ 1
1.2 History of Coen Company................................................................................................................................................ 2
1.3 History of Hamworthy Combustion............................................................................................................................... 3
1.4 Boiler Basics........................................................................................................................................................................ 3
1.4.1 Industrial Boilers................................................................................................................................................... 3
1.4.2 Package Boilers....................................................................................................................................................... 4
1.4.3 Field-Erected Boilers.............................................................................................................................................. 6
1.4.4 Power Generation Industry.................................................................................................................................. 6
1.5 Utility Boilers...................................................................................................................................................................... 7
1.6 Utility Boiler/Burner Design........................................................................................................................................... 7
1.7 Utility Boiler Types............................................................................................................................................................ 7
1.7.1 Wall Fired Burner Installation............................................................................................................................. 7
1.7.2 Opposed Fired Installation.................................................................................................................................. 8
1.7.3 Tangentially Fired Installations........................................................................................................................... 8
1.8 Air Heaters.......................................................................................................................................................................... 9
1.9 Duct Burners....................................................................................................................................................................... 9
1.10 Burners.............................................................................................................................................................................. 10
1.10.1 Competing Priorities........................................................................................................................................... 10
1.10.2 Design Factors...................................................................................................................................................... 11
1.10.2.1 Fuel.......................................................................................................................................................... 12
1.10.2.2 Oxidizer.................................................................................................................................................. 12
1.10.2.3 Custom-Engineered Solutions............................................................................................................ 12
References................................................................................................................................................................................... 13
type of heating process. While this book is concerned

specifically with industrial/utility combustion, all of
1.1 Introduction
these combustion processes share many features in
While fire has been existent since the beginning of time, common.
much remains to be learned about it. Because the sci- The last Coen combustion handbook was written in
ence of combustion combines heat transfer, thermody- 1920 by Arthur H. Light.1 This current handbook was
namics, chemical kinetics, and multiphase turbulent inspired from an internal Coen week-long engineering
fluid flow to name a few areas of physics, the study of workshop that was developed as an in-depth review
industrial combustion is interdisciplinary by necessity. for engineers specializing in combustion applications.
The field of industrial combustion is very broad and The course was required for all company engineers and
touches directly, or indirectly, nearly all aspects of life. included a comprehensive final exam. This workshop
Electronic devices are generally powered by fossil- formed the basis for the topics and order of introduc-
fuel-fired power plants. Automobiles use internal com- tion in this handbook. This handbook is intended to be
bustion engines. Planes use jet-fuel-powered turbine a review of basic engineering topics, followed by more
engines. Most materials have been made through some detailed topics and practical examples.
1
2 The Coen & Hamworthy Combustion Handbook
a complete line of coal-fired burners and its fuel feed

system for rotary kiln firing and industrial boilers. The
1.2 History of Coen Company
system provided higher output, lower fuel consump-
The origins of Coen Company can be traced back to tion, and better quality product than any other coal fir-
1912, when Garnet Coen first fashioned an innovative ing system available at the time. This was also a time
device which would heavily impact the fuel burning of alternative fuels and efficiency innovations. Coen
industry. His invention—the adjustable tip mechanical developed biomass burners and slurry burners together
oil burner, a unique device that could maintain atomi- with complicated microprocessor control systems into
zation quality at low supply pressure via an adjustable packages. At the time, it was projected that natural gas
tip—was what ultimately spurred the company to its would cease to be used as a boiler fuel.
present success.2 In recent years, the trend toward the use of natural
In 1914, Coen employed Joseph Voorheis, a mechani- gas has increased dramatically, due to lower relative
cal engineer who at the time was working for Shell Oil cost, higher availability, and lower pollution emissions.
Company. Voorheis sought after Coen’s burner design Coen responded to this need with the development of
for the retrofit of a mechanical burner onto a tugboat and Low NOx gas burners and Ultra Low NOx gas burn-
soon after began working for the company. With his help, ers and the associated controls required. Unknown in
the first “Coen system” of mechanical oil burning was the 1970s, Coen was converting kilns, industrial boil-
created consisting of an oil heater, a duplex strainer, and ers, and utility boilers to natural gas firing. Although
simple integrated pumping systems ahead of the burner. seemingly simple, conversion to fire natural gas in
From the beginning, it was obvious the company many cases required a detailed heat transfer analysis
intended to offer more than just a product or system. due to changes in heat losses and luminosity of the
It was offering engineering expertise to accommodate flames. These conversions were uniquely engineered
the combustion requirements and capabilities of any solutions that by then had become a common Coen
furnace, at sea or on land. This early emphasis on engi- supplied offering.
neering would serve the company well in the decades Coen remained under the control of Garnet Coen until
ahead as new technologies created expanding markets 1934 when Joe Voorheis acquired principal ownership
and opportunities. of the company. Coen continued as a privately owned
It was not long before the efficiency engineers of large company (Voorheis) when in April 2007 it was acquired
oil and industrial companies recognized the adaptabil- by another privately owned company: Koch Chemical
ity of the Coen systems for operation of the boiler and Technology Group, a division of Koch Industries. This
refinery heater furnaces of their stationary installations. acquisition merged Coen products and Todd combus-
By 1921, hundreds of Coen burner systems were fueling tion products. In 2011, Coen acquired Hamworthy
boilers and furnaces in the oil companies’ pumping sta- Peabody Combustion, further expanding its combus-
tions and refinery heaters, in breweries, power plants, tion products and offerings. On August 2, 2012, Coen
foundries, smelters, and institutions throughout the celebrated 100 years since its incorporation on August 2,
United States. 1912. For all those years, Coen has remained innovative
By the early 1950s, a new trend was being established and privately owned.
in the boiler industry—packaged boilers. Packaged Although Coen began as a company providing burn-
water tube boilers were shop-assembled and could be ers for industrial boiler and process heaters, it has
transported and installed within days after unloading. expanded since the inception 100 years ago to include
Prior to this, multi-burner boilers were erected on site
and took months to construct. Packaged boilers resolved
these issues and could be shipped on a flat car to the site, • Utility/industrial/process burners and associ-
thus saving time, labor, and money. However, these new ated support products
long furnaces did not permit the application of multiple • Advanced combustion controls and burner
burners, common with field-erected boilers. To overcome managements systems
this problem, Coen developed new forced draft large • Applications and systems for a wide variety of
single burners custom engineered for these new fur- solid, liquid, and gaseous fuels
naces. Further, Coen provided package burners for these
• Pump sets, fuel trains, and custom systems
boilers, complete with piping and controls, so the entire
such as solid conveying
boiler/burner/controls became a package unit. Coen has
provided thousands of these package burner units. • Cement, lime and ore kilns, air heaters, incin-
In the 1970s, Coen realized that oil and gas prices erators, and duct burners systems
would drive large fuel users, such as cement plants and • Unique custom solutions to one of a kind com-
larger boilers, to the use of coal. The company developed bustion applications
Introduction 3
Flame Radiant section
1.3 History of Hamworthy Combustion

While Coen, based in San Francisco, California, was
retrofitting boilers with a novel mechanical atomizer
around 1912, two brothers, Percy and Sidney Hall,
incorporated the Hamworthy Engineering Company on
April 16, 1914, located 5350 miles away in Poole England.
Convection
Hamworthy designed and built oil engines initially for section
marine use plus pumps and compressors.3
After surviving two wars and the great depression,
Hamworthy realized the trend in switching from coal
fuels to oil and established the British Combustion Refractory or steel baffle
Equipment (BCE) group in 1946. By 1956 the BCE had
become the driving force in the Hamworthy group. FIGURE 1.1
Flow through a boiler, highlighting radiant and convection sections.
From 1960 to 1980, Hamworthy had close links with the
British Central Generating Board (CEGB) in the supply
of utility-grade oil-fired burners to new oil-fired util- the steam drum. At the steam drum, both water and steam
ity furnaces. In the early 1990s, Hamworthy acquired exist. Separators are used (not shown) that separate steam
Peabody Combustion, another burner-based company from water. The result is saturated steam at the exit.
founded in the United States in 1920. The boiled water (steam) is usually controlled at a
From the mid-1980s to the present time, Hamworthy much higher pressure than ambient, for example, maybe
either acquired or developed a complete line of burners for 150–500 PSIG (10–35 barg). This means the temperature
gas and oil firing, ignition systems, and package burners of the steam is higher than atmospheric boiling (212°F =
systems. Together, Coen and Hamworthy will be celebrat- 100°C). The steam generated by boilers can be saturated
ing nearly 200 years of combined combustion experience or superheated. Saturated steam is in equilibrium with
at the time of this publication. the hot water at any given pressure, meaning both liquid
and steam exist in the same volume/space. Saturated
steam temperature is a direct function of pressure, that
is, 212°F (100°C) at atmospheric pressures. At 350 PSIG
(24 barg), the saturated temperature for boiling is at
435°F (223°C). Heating the resultant saturated steam to
1.4 Boiler Basics
higher temperatures will result in superheated steam by
Included later is a significant discussion on boilers in returning saturated steam back to the boiler from the
this section because this is the largest application of steam drum to super heater tubes for further heating.
large burners and the burner design and placement is Although the firing rate of the burner defines the
essential for the boiler performance. Simply stated, boil- boiler steam output, the burner design and placement
ers convert water to steam; however, the actual process is must be engineered to achieve the boiler essential oper-
very complex. Water tube boilers can be simply described ation such as
as an upper steam drum and a lower water drum (mud
drum). These drums are connected by a multitude of • Prevention of flame impingement
tubes forming both a large open volume called a radi- • Emission compliance, NOx, CO, HCs. VOC,
ant section and a multitude of closely spaced tubes form- particulate, opacity, etc.
ing a convection section (see Figure 1.1). Some of these
• Efficiency, limiting excess air, turndown
tubes can be “risers” and some “down comers” causing
natural circulation in the boiler water/steam system. • Noise from piping and combustion roar
Larger boilers may have forced circulation using pumps. • Control of superheat temperature
Flames are introduced in the radiant section, producing
mostly radiant heat transfer. The post-flame gases enter
1.4.1 Industrial Boilers
the convection section where the dominant mode of heat
transfer is convection (see Chapter 7). Burners designed for use in industrial boilers are most
Boilers receive treated feed water (liquid). This is added commonly of the circular register design, range in size
to the lower “mud” drum at 212°F–300°F (100°C–150°C). from heat inputs of 20 to 400 × 106 British Thermal Units
The heat from flame or post-flame gases heats the water (Btus)/h (6–115 MW), operate with forced draft (FD) fans
which is converted into steam in the “rising tubes” up to supplying the combustion air, and burn gas or liquid
hydrocarbon fuels. This section covers a variety of different

sizes and types of industrial boilers. The most commonly
encountered boilers would be single-burner industrial-
packaged boilers and wall fired field-erected boilers with
up to six burners per boiler. Industrial-packaged boilers
consist of a furnace and boiler, which is a self-contained
system that can basically be shipped as a unit, which
became prevalent in the 1950s as a method to save time
and labor. While some larger field-erected boilers with
more burners, and designs such as the tangentially fired,
turbo-fired, or cyclonic-fired boilers, can occasionally be
found in these applications, they make up a very small
percentage of the total number of installed units.
1.4.2 Package Boilers

The simplest form of burner arrangement is to have
only one burner that provides all of the necessary heat
input to the boiler. The most simple package boiler
consists of an FD fan, burner and wind box, breech-
FIGURE 1.3
ing, and stack. This is possible in industrial-package Front elevation of “A” type boiler—the furnace and drum locations
boilers generating up to 300,000 lb/h (136,000 kg/h) of form the letter “A,” hence the name.
steam, where single burners can reach heat inputs up
to 400 × 106 Btu/h (115 MW). The burner is located on
the end wall of the boiler, which can be fully refrac-
tory lined or a combination of refractory around the
burner throat and tubes comprising the rest of the
wall. Industrial-package boiler configurations are typ-
ically denoted by the drum configuration, with “D”
style (see Figure 1.2), “O” style (see Figure 1.3), and “A”
style (Figure 1.4) boilers being the most common.
FIGURE 1.4
Front elevation of “O” type boiler—the furnace and drum locations
form the letter “O,” hence the name.
A “D” style boiler has two drums located directly

above each other on either the right or the left side of
the boiler, with the tubes extending out to form the fur-
nace in a “D” configuration. This boiler type has only a
single convection bank located on the same side as the
FIGURE 1.2
drums (see Figure 1.5). An “O” style boiler is similar to
Front elevation of “D” type boiler—the furnace and drum locations a “D” style in that it utilizes two drums centered over
form the letter “D,” hence the name. each other; however, in the case of an “O” style boiler
Introduction 5
FIGURE 1.5
“D” type package boiler. (Courtesy of Victory Energy Operations, LLC.)
the drums are located over the center of the furnace. As the capacity of the boilers gets larger, the most
In this configuration, a convective bank is located on constrained dimension is typically the width of the
each side of the furnace so the flow out of the furnace boiler, which is restricted to allow shipment by road or
is split evenly between these two convective banks rail. Space constraints at the site may also constrain the
(see Figure 1.6). An “A” style boiler has three drums, a allowable length of the boiler. In these cases, the firebox
single steam drum in the top center of the furnace, and dimensions may not be sufficient to accommodate the
two mud drums located in each corner of the furnace. flame geometry required from a single burner. In some of
Similar to the “O” style boiler, the “A” style also has two these cases, given the right firebox geometry, two burners
convection banks located on each side of the furnace. can be supplied in a common wind box and operated as
FIGURE 1.6
“O” type package boiler. (Courtesy of Victory Energy Operations, LLC.)
a single unit, called unison firing. This also can be used to For multiple burner applications, each burner can be
reach heat inputs higher than those available from one brought in and out of service independently, allowing
burner only. In the case of unison firing, both burners greater flexibility in operating turndown. Typically
operate as a single unit with the loss of flame on either all burners in service are controlled by a single fuel-
burner causing a shutdown of the entire system. control valve and therefore operate at the same heat
input. For added flexibility on units with several rows
or columns of burners, a flow-control valve can be sup-
1.4.3 Field-Erected Boilers
plied for each row or column, allowing more flexibility
As boiler capacities get larger than the physical size that in controlling heat input and distribution within the
can be shipped as an assembled unit, the boilers are firebox. Burners should be brought into service sym-
shipped in pieces and erected at the site. These field- metrically about the boiler drum centerline to provide
erected boilers example (shown in Figure 1.7) are usually balanced heating to the boiler and to minimize drum
a large furnace and boiler system with multiple ancillary level fluctuations.
systems. The system is shipped to the field in pieces and
erected on site. Sometimes, portions are fabricated on site.
1.4.4 Power Generation Industry
To minimize the footprint of these “field-erected” boil-
ers, the fireboxes get taller as the boiler capacity grows, Boilers are used for a variety of purposes in an assort-
while the width and depth of the boiler do not typically ment of applications. Common uses include producing
grow proportionally with capacity. Since the depth of hot water or steam for heating, producing steam for use
the firebox that is available to accommodate the flame within a plant such as atomizing oil for oil-fired burn-
length is constrained, these boilers will utilize multiple ers, and producing steam to generate power in large
burners, from 4 to as many as 16, arranged in rows on a power plants. Applications range from small single-
single wall or on opposed firing walls. The number and burner uses in hospitals, schools, and small businesses
arrangement of the burners is based on the required heat up to large multi-burner boilers in power plants.
input and the available width and depth to accommodate Both duct burners and boiler burners are used in the
the flame geometry. power generation industries. Duct burners (see Chapter 15)
Burner spacing is important to ensure that no flame- are burners that are inserted into large ducts to boost
to-flame interaction occurs which can increase emissions the temperature of the gases flowing through the ducts.
and flame lengths, leading to impingement. This can These burners are frequently used in cogeneration
vary based on the different designs of burners employed projects, electrical utility peaking stations, repower-
and the design pressure drop across the burner. ing programs, and in industrial mechanical driver sys-
tems employing gas turbines with site requirements for
steam. They are also used in fluidized bed combustors
Type PFT integral-furnace boiler
and chemical process plants. The efficiency of a duct
burner to supply additional heat approaches 100% (on a
lower heating value basis), which is much higher than,
for example, a backup boiler system in generating more
steam. Duct burners are often easily retrofitted into
existing ductwork. Several important factors in-duct
burner applications include low pollutant emissions,
safe operation, and uniform heat distribution from the
duct burners to the gases flowing through the duct, get-
ting uniform gas distribution through the duct burn-
ers, and having adequate turndown to meet fluctuating
demands. Duct burners typically use gaseous fuels, but
occasionally fire on oil.
Boiler burners (see Chapter 14) are used to combust
fuels, commonly natural gas or fuel oil, in the production
of steam, which is often used to produce electrical energy
for power generation. These burners produce radiation
and convection used to heat water flowing through the
boiler. The water is vaporized into steam. Sometimes the
steam is used in the plant in the case of smaller indus-
FIGURE 1.7 trial boilers. Larger utility boilers produce steam to drive
Typical field-erected boiler. turbines for electrical energy production. While boiler
Introduction 7
burners have been around for many years, there have • Minimum parasitic power requirements through
been many design changes in recent years due to the cur- low pressure drop of combustion air and flue gas
rent emphasis on minimizing pollutant emissions. systems, especially the burner register draft loss
• Simplicity of burner maintenance and adjustment
• Flame dimensions to match the dimensions of
the furnace, with no flame impingement on any
furnace wall
1.5 Utility Boilers
The burners are typically located in common wind
Burners designed for use in utility boilers are very boxes which supply air to all of the burners located on
similar to those found in industrial boilers, with a few each firing wall (or corner). Proper air flow distribu-
minor differences. Utility boilers typically employ tion to each burner must be ensured during the system
multiple burners, from 4 to as many as 48 or more. design and is typically accomplished through physical
These burners can be brought in and out of service or computational fluid dynamic (CFD) modeling (see
as needed, allowing the firing rate of the boiler to Chapter 9) of the air delivery system and wind box. The
be varied greatly without needing to achieve a high goal is to provide equal combustion air flow between
turndown ratio on any single burner. Therefore, each burners, uniform peripheral velocity distributions at the
burner on a utility boiler does not need to cover as burner inlets, and the elimination of tangential veloci-
wide a size range as the burner on industrial boiler, ties within each burner. If the unit has been designed
and typically range in size with heat inputs of 50–400 × with FGR, the O2 content must be equal between the
106 Btu/h (15–115 MW). They operate with forced draft burners, and this is accomplished by balancing the FGR
fans supplying the combustion air, which is typically distribution to each burner.
preheated to between 400°F (200°C) and 650°F (340°C) As the burners are taken in and out of service, to
as compared to ambient temperature air in most indus- maximize boiler efficiency, it is desirable to limit the
trial applications. They burn gaseous, liquid, or solid air flow through the out-of-service burners. This is
fuels, and may need to be fuel-flexible to take advan- typically done by including an air damper or register
tage of changing fuel costs. on each burner which can be closed when the burner
is out of service to allow only enough air flow through
the burner as needed for cooling and purging. These
dampers can be automated, along with the burner fuel
shutoff valves, so that operators can take burners in and
out of service from the control room or even automate
1.6 Utility Boiler/Burner Design this function as part of the boiler’s combustion controls.
Utility boiler/burner designs, both conventional and In multiple burner installations, NOx reduction can
low NOx, employ similar design techniques as those for be achieved by biasing the fuel to some burners. This
industrial boilers/burners. The differences that sepa- causes some burners to operate fuel rich and others to
rate utility applications are generally based on their operate fuel lean. This may include shutting off the fuel
intended operational parameters, such as the fuels to be completely to one or more burners, which is called burn-
fired and level of air pre-heat they are designed to han- ers-out-of-service (BOOS). The optimum amount and
dle. Due to the importance of maintaining high electri- pattern of biasing or BOOS is often very boiler-depen-
cal system reliability and minimizing generation costs, dent, with the best NOx reduction (see Chapter 10) results
utility boiler burners must offer found through a series of tests during commissioning.
• High reliability during long-term operation

• Simplicity and reliability of fuel ignition
• High flame stability across the operational
turndown range 1.7 Utility Boiler Types
• Fuel flexibility, including the ability to co-fire 1.7.1 Wall Fired Burner Installation
multiple fuels
The simplest form of burner arrangement is to have all
• High thermal efficiency by minimizing excess of the burners located on only one wall of the furnace
air levels in a common wind box. The burners will be located in
• Minimizing emissions through operation with rows and columns, based on the number of burners
flue gas recirculation (FGR) and over fire air (OFA) required. The simplest configuration would be a four
burner arrangement with two rows and two columns of of the furnace, and in some cases boilers have been
burners. Very large boilers may have up to 24 burners, designed with burners located on all four walls. The
or more, which could be arranged in a pattern of 4 rows burners are located on opposite walls and are there-
each containing 6 burners. fore called “opposed fired” boilers (see Figure 1.8). In
Spacing of the burners is important to ensure that no opposed fired applications, not only is the spacing
flame-to-flame interaction occurs which can increase between burners important, but also, the interaction
emissions and flame lengths, leading to impingement. between the flames of the opposed burners meet-
This can vary based on the different designs of burners ing in the center of the furnace. Depending on the
employed and the design pressure drop across the burner. boiler design, the burners may be directly opposite
For multiple-burner applications, each burner can be each other or may be staggered to help avoid inter-
brought in and out of service independently, allowing action with the burners from the opposite wall. In
greater flexibility in operating turndown. Typically, all some instances, the boilers may include a “division
burners in service are controlled by a single fuel-control wall” along the centerline of the furnace that extends
valve and therefore operate at the same heat input. For from the bottom to some height in the furnace. This
added flexibility on units with several rows or columns wall eliminates some concerns over opposed burner
of burners, a flow-control valve can be supplied for each flames interacting, but does present concerns for
row or column, allowing more flexibility in controlling flame impingement.
heat input and distribution within the firebox. Burners
should be brought into service symmetrically about the
1.7.3 Tangentially Fired Installations
boiler drum centerline to provide balanced heating to
the boiler and to minimize drum level fluctuations. Combustion Engineering developed a boiler design that
places all of the burners at the corners of the boiler, fir-
ing tangentially toward a pitch circle in the center of the
1.7.2 Opposed Fired Installation
firebox. This tangentially fired or T-fired boiler design
Larger utility boilers are designed to have burners utilizes a vertical column of burners in each corner of the
on two of the four walls and firing toward the center firebox (see Figure 1.9). The burner flames all converge
Boiler
Economizer
Advanced Electrostatic Electrostatic

overfire air precipitator precipitator
(AOFA)
ports Air preheater
Combustion Air
Stack
air
Air
Windbox AOFA ports
Windbox
Pulverized
pulverized
coal
coal
Concentric
firing
system Concentric Dry waste to disposal
firing
system
Ash
FIGURE 1.8
Opposed wall fired boiler.
Introduction 9
firebox. By controlling the location of the fireball rela-

tive to the super heater tubes located at the top of the
furnace, superheated steam temperature can be con-
trolled. This can also be used to control the residence
time of the combustibles in the furnace to assure car-
bon monoxide (CO) burnout on harder-to-combust
fuels.
1.8 Air Heaters

Air heaters (see Chapter 16) are used in a wide range of
industries—for preheating of process heaters in the pet-
rochemical and refining industries, for detergent, coal or
fertilizer drying, and for other heating and drying appli-
cations, for example, in the chemical, soap, paper, food,
and cement industries. They are used mainly for high-
temperature drying, where the combustion products
mix with air and are used to heat the product directly.
Typical applications include drying of detergents, min-
erals, fertilizers, animal feeds, and coal. Air heaters are
also used as an indirect source of heat for drying of food-
stuffs such as milk powder and dried potatoes, roasting
of coffee beans, and production of food-grade chemical
additives.
The most common type of air heater is the Peabody
twin shell design. This type of air heater is used for
conventional drying applications where the inlet air
stream is at a low temperature, less than 100°C (212°F)
FIGURE 1.9 and the required outlet temperature is no more than
TXU Electric Collin Station Power Plant, a tangential fired boiler. 800°C (1500°F). The twin shell design gets its name-
sake from an annular passage through which diluted
air passes between the boiler outer shell and the com-
into a swirling “fireball” in the center of the furnace. bustion chamber, before mixing with hot combus-
The number of burners in each column is the same and tion products to achieve the final outlet temperature.
is dependent on the capacity of the boiler and number Burners for air heaters are generally round, require
of fuels being fired. high turndown, and sometimes must operate with very
The burners originally supplied by the boiler OEM high excess air. The high excess air used in air heaters
for these boilers consisted of square burner “buckets” can make low NOx operation a challenge. On the other
that were either fixed or tilting. The vertical column hand, lean premix burners can be easily employed to
would contain some buckets dedicated to each particu- reduce NOx.
lar fuel and some buckets that supplied only air. Some
of the fixed (non-tilting) bucket applications have been
retrofitted in the field to accommodate round burners,
although in most cases burner retrofits and upgrades
involve modifying or replacing the fuel components
1.9 Duct Burners
and buckets with components that fit into the existing
burner geometry. Duct burner systems (see Chapter 15) can be loosely
In some applications, the burners are designed to be described as large cross-sectional ducts with high
tilted up and down by plus or minus 30° from level. flows that require uniform heat addition for rela-
The burners are all tilted at the same angle, which tively small temperature increases. The flows can be
allows the fireball to be moved higher or lower in the air, fumes, or oxygen-depleted streams. The burners
used are also called ribbon burners, linear burners, or

duct burners designed so that the heat input can be
1.10 Burners
distributed over a relatively large cross section. The
typical location of this type of burner can be seen in The burner is the device that is used to combust the fuel
Figure 1.10. with an oxidizer to convert the chemical energy in the
Linear and in-duct burners were used for many fuel into thermal energy. A given combustion system
years to heat air in drying operations before their may have a single burner or many burners, depending
general use in cogeneration systems. Some of the on the size and type of the application. A typical round
earliest systems premixed fuel and air in an often low NOx burner is shown in Figure 1.11. Boilers and other
complicated configuration, which fired into a recir- combustion chambers come in various sizes and configu-
culating process air stream. The first use for high- rations. It is the design and orientation of the burner that
temperature, depleted oxygen streams downstream will define proper operation of the equipment. There are
of gas turbines in the early 1960s was to provide many factors that go into the design of a burner. This sec-
additional steam for process use in industrial appli- tion will briefly consider some of the important factors
cations and electrical peaking plants operating that are considered in designing burners. A detailed dis-
steam turbines. As gas turbines have become larger cussion on burners is given in Chapter 14.
and more efficient, duct burner supplemental heat
input has increased correspondingly. Duct burn- 1.10.1 Competing Priorities
ers are suitable for a wide variety of direct-fired air
heating applications where the physical arrangement There have been many changes in the traditional designs
requires mounting inside a duct. that have been used in burners, primarily because of
Integral deaerator
Silencer Uses low temperature heat
Attenuates noise level to deaerate feed-water for
to meet government improved thermal efficiency
and site requirements
CO catalyst
Reduces carbon
monoxide in the
flue gas
SCR catalyst
Reduces nitrous
oxides in the
flue gas
Duct burner
Provides supplementary firing
of turbine exhaust to increase
unfired steam production
Diverter valve
Modulates steam production
in the bypass systems
FIGURE 1.10
Typical duct burner location. (Courtesy of Hamon Deltak.)
Introduction 11
FIGURE 1.11
Ecojet burner in an HRSG.
the recent interest in reducing pollutant emissions. In which can increase flame radiation. The actual impact
the past, the burner designer was primarily concerned of staging on the heat transfer from the flame is highly
with efficiently combusting the fuel and transferring the dependent on the actual burner design.6
energy to a heat load. New and increasingly more strin- In the past, the challenge for the burner designer was
gent environmental regulations have added the need to to maximize the mixing between the fuel and the oxi-
consider the pollutant emissions produced by the burner dizer to ensure complete combustion. If the fuel was
(see Chapter 10). In many cases, reducing pollutant difficult to burn, as in the case of low heating value
emissions and maximizing combustion efficiency are at fuels such as waste liquid fuels or process gases from
odds with each other. For example, a well-accepted tech- chemical production, the task could be very challeng-
nique for reducing NOx emissions is known as staging, ing. Now, the burner designer must balance the mixing
where the primary flame zone is deficient of either fuel of the fuel and the oxidizer to maximize combustion
or oxidizer.4 The balance of the fuel or oxidizer may be efficiency while simultaneously minimizing all types
injected into the burner in a secondary flame zone or, in of pollutant emissions. This is no easy task as, for
a more extreme case, may be injected somewhere else in example, NOx and CO emissions often go in opposite
the combustion chamber. Staging reduces the peak tem- directions. When CO is low, NOx may be high and
peratures in the primary flame zone and also alters the vice versa. Modern burners must be environmentally
chemistry in a way that reduces NOx emissions because friendly, while simultaneously efficiently transferring
fuel-rich or fuel-lean zones are less conducive to NOx heat to the load.
formation than near stoichiometric zones.5 NOx emis-
sions increase rapidly with the exhaust product temper-
1.10.2 Design Factors
ature. Since thermal NOx is exponentially dependent on
the gas temperature, even small reductions in the peak There are many types of burner designs that exist due to
flame temperature can dramatically reduce NOx emis- the wide variety of fuels, oxidizers, combustion chamber
sions. However, lower flame temperatures often reduce geometries, environmental regulations, thermal input
the radiant heat transfer from the flame since radiation sizes, and heat transfer requirements. Additionally,
is dependent on the fourth power of the absolute tem- heat transfer requirements include, for example, flame
perature of the gases. Another potential problem with temperature, flame momentum, and heat distribution.
staging is that it may increase CO emissions, which is Garg7 lists the following burner specifications that are
an indication of incomplete combustion and reduced needed to properly choose a burner for a given applica-
combustion efficiency. However, it is also possible that tion: burner type, heat release and turndown, air supply
staged combustion may produce soot in the flame, (natural draft, forced draft, or balanced draft), excess air
level, fuel composition(s), firing position, flame dimen- In some cases, the burner may have more than one
sions, ignition type, atomization media for liquid fuel type of fuel. An example is shown in Ref. [12]. Dual-
firing, noise, NOx emission rate, and whether waste gas fuel burners are designed to operate typically on either
firing will be used.7 Some of these design factors are gaseous or liquid fuels. These burners are used, usually
briefly considered next. for economic reasons, where the customer may need
to switch between a gaseous fuel like natural gas and
a liquid fuel like oil. These burners normally operate
1.10.2.1 Fuel
on one fuel or the other, and sometimes on both fuels
Depending upon many factors, certain types of fuels are simultaneously. Another application where multiple
preferred for certain geographic locations due to cost fuels may be used is in waste incineration. One method
and availability considerations. Gaseous fuels, particu- of disposing of waste liquids contaminated with
larly natural gas, are commonly used in most industrial hydrocarbons is to combust them by direct injection
heating applications in the United States. In Europe, through a burner. The waste liquids are fed through
natural gas is also commonly used along with light fuel the burner, which is powered by a traditional fuel such
oil. In Asia and South America, heavy fuel oils are gen- as natural gas or oil. The waste liquids often have very
erally preferred although the use of gaseous fuels is on low heating values and are difficult to combust with-
the rise. out auxiliary fuel. This further complicates the burner
Fuels also vary depending on the application. For design where the waste liquid must be vaporized and
example, in incineration processes, waste fuels are com- combusted concurrently with the normal fuel used in
monly used either by themselves or with other fuels like the burner.
natural gas. In the petrochemical industry, fuel gases
often consist of a blend of several fuels, including gases
like hydrogen, methane, propane, butane, propylene, 1.10.2.2 Oxidizer
nitrogen, and carbon dioxide.8
The predominant oxidizer used in most industrial
The fuel choice has an important influence on the
heating processes is atmospheric air. This can present
heat transfer from a flame. In general, solid fuels like
challenges in some applications where highly accu-
coal and liquid fuels like oil produce very luminous
rate control is required due to the daily variations
flames, which contain soot particles that radiate like
in the barometric pressure and humidity of ambient
blackbodies to the heat load. Gaseous fuels like natu-
air. The combustion air is sometimes preheated to
ral gas often produce nonluminous flames because
increase the overall thermal efficiency of a process.
they burn so cleanly and completely without produc-
Combustion air is also sometimes blended with some
ing soot particles. A fuel like hydrogen is completely
of the products of combustion, a process usually
nonluminous because there is no carbon available to
referred to as flue gas recirculation (FGR). FGR is used
produce soot.
to both control boiler superheat or reheat and reduce
In cases where highly radiant flames are required,
NOx emissions.
a luminous flame is preferred. In cases where convec-
tion heat transfer is preferred, a nonluminous flame
may be preferred in order to minimize the possibility of
1.10.2.3 Custom-Engineered Solutions
contaminating the heat load with soot particles from a
luminous flame. Where natural gas is the preferred fuel From the early beginning of the company, Coen has been
and highly radiant flames are desired, new technologies a leader in innovation. New products were regularly
are being developed to produce more luminous flames. developed and introduced to the market place. These
These include processes like pyrolyzing the fuel in a products were the beginning of “custom-engineered
partial oxidation process,9 using plasma to produce soot solutions.” This was an informal process until 1978 when
in the fuel,10 and generally controlling the mixing of the Coen organized a standalone R&D department. Why?—
fuel and oxidizer to produce fuel-rich flame zones that because everyone was too busy working on projects to
generate soot particles.11 spare time to develop new ones. One of the co-editors of
Therefore, the fuel itself has a significant impact on this book (Steve) was hired that year into the new R&D
the heat transfer mechanisms between the flame and the department.
load. In most cases, the fuel choice is dictated by the end New products were released on the average of 3–4 per
user as part of the specifications for the system and is year, such as the dual zone burner kiln burner, biomass
not chosen by the burner designer. The designer must firing, electronic products, control systems, low NOx
make the best of whatever fuel has been selected. In burners, larger burners, and much more. Many new prod-
most cases, the burner design is optimized based on the ucts were, in fact, custom-engineered solutions to solve
choice for the fuel. unique customer applications. The existing test facility
Introduction 13
was slowly modernized and CFD was first utilized for a

duct burner project in 1985.
References
R&D grew and was subdivided into R&D and staff
engineering as the complexity of projects and applica- 1. A.H. Light, Efficient Oil Burning: A Graphic Description of
tions grew. This set the stage for a significant increase in Combustion, Coen Co., New York, 1920.
“custom-engineered solutions.” 2. The Coen Story, Coen Co. Inc., 1980s.
3. Manos, T., Hamworthy Combustion Engineering
At the time, the Clean Air Act was passed in the
Limited, Biddles Limited, Dorset, U.K., 2004.
United States and new environmental rules were being
4. J.L. Reese, G.L. Moilanen, R. Borkowicz, C. Baukal,
proposed, requiring reduced emissions. Alternative D. Czerniak, and R. Batten, State-of-the-art of NOx emis-
fuels and more efficient unit operation were becom- sion control technology, in Proceedings of International
ing a customer priority. Coen was continuously being Joint Power Generation Conference., Phoenix, AZ, October
asked by its customers, “How can I solve this?,” thus 3–5, 1994, ASME Paper 94-JPGC-EC-15.
the phrase was coined: “custom-engineered solu- 5. C. Baukal, Industrial Combustion Pollution and Control,
tions.” Since then, Coen has conducted hundreds of Marcel Dekker, New York, 2004.
studies resulting in custom solutions to customer 6. C. Baukal (ed.), Industrial Burners Handbook, CRC Press,
problems. Boca Raton, FL, 2004.
What is a custom-engineered solution? It is generally 7. A. Garg, Better burner specifications, Hydrocarbon Pro-
cessing, 68, (8), 71–72, 1989.
as follows:
8. T. Dark, J. Ackland, and J. White, Fuels, in The John Zink
Combustion Handbook, C. Baukal (ed.), CRC Press, Boca
1. Coen is approached by a customer with a com-
Raton, FL, 2001, Chapter 5.
bustion problem of some kind. 9. M.L. Joshi, M.E. Tester, G.C. Neff, and S.K. Panahi, Flame
2. Coen does not have an “off-the-shelf” solution. particle seeding with oxygen enrichment for NOx reduc-
3. The company conducts a study of the prob- tion and increased efficiency, Glass, 68, (6), 212–213, 1990.
10. R. Ruiz, and J.C. Hilliard, Luminosity enhancement of
lem and produces a proposal to solve the
natural gas flames, in Proceedings of 1989 International Gas
problem.
Research Conference, T.L. Cramer (ed.), Govt. Institutes,
Rockville, MD, pp. 1345–1353, 1990.
The study may involve a site visit, measurements, and 11. A.G. Slavejkov, T.M. Gosling, and R.E. Knorr, Low-NOx
possibly a CFD study. In some cases, a scale burner/ staged combustion device for controlled radiative heat-
system may be constructed and tested. A report is ing in high temperature furnaces, U.S. patent 5,611,682,
issued with a quote for equipment to solve the problem. March 18, 1997.
Historically, all of these custom-engineered solutions 12. API Publication 535, Burner for Fired Heaters in General
have been successful and arrived at an economical solu- Refinery Services, 1st edn., American Petroleum Institute,
tion where none existed. Washington, DC, July 1995.
2
Engineering Fundamentals
Stephen B. Londerville
CONTENTS
2.1 Introduction...................................................................................................................................................................... 15
2.2 Time, Length, and Mass.................................................................................................................................................. 15
2.2.1 English Units........................................................................................................................................................ 16
2.2.2 SI Units.................................................................................................................................................................. 16
2.2.3 Absolute English and SI Units........................................................................................................................... 16
2.2.4 Dimensional Analysis......................................................................................................................................... 17
2.2.5 Some Simple Derived Units................................................................................................................................ 17
2.2.5.1 Velocity................................................................................................................................................... 18
2.2.5.2 Volume.................................................................................................................................................... 18
2.2.5.3 Density.................................................................................................................................................... 18
2.2.5.4 Specific Weight...................................................................................................................................... 18
2.2.5.5 Pressure.................................................................................................................................................. 18
2.3 Work and Energy............................................................................................................................................................. 19
2.3.1 Power..................................................................................................................................................................... 19
2.4 Temperature...................................................................................................................................................................... 19
2.5 Enthalpy and Specific Heat............................................................................................................................................ 20
2.6 Viscosity of Gas and Liquid........................................................................................................................................... 21
2.7 Bulk Modulus................................................................................................................................................................... 22
2.8 Slang Units and Notations.............................................................................................................................................. 23
2.9 Conclusion........................................................................................................................................................................ 23
References................................................................................................................................................................................... 23
the incident cost NASA well over $600 million.1 Despite

the infinite complexity in the engineering of the Orbiter,
2.1 Introduction
it was a simple conversion error that was ultimately the
The NASA Mars Climate Orbiter case is a monumental source of failure. This holds true for all engineers as it is
example of the significance of units and conversions for easy to overlook such basic fundamentals.
practicing engineers. The Mars Climate Orbiter was set
to land on the surface of Mars on September 23, 1999, to
study the Martian climate, atmosphere, and surface. The
orbiter trajectory was planned to be 226 km (140 miles)
above the surface so that the gravitational acceleration
of Mars could pull the Orbiter through the atmosphere
2.2 Time, Length, and Mass
at safe speeds. Instead, the trajectory was only 57 km Since the existence of early civilization, setting up a
(35 miles) and the orbiter was disintegrated by the atmo- standard for weights and measures was vital for trade
spheric stresses. The magnitude of error was off by a and construction. The earliest establishment of the mod-
factor of 4.45, the exact conversion from Newtons to lb- ern universal measurement system was set forth by the
force. The trajectory was programmed for force inputs Magna Carta of 1215, which proposed that “there shall
in metric Newtons; however, the crew at the controls be one unit of measure throughout the realm.” Later,
was entering trajectory data in imperial units of lb- the growing development of the sciences during the
force. Fortunately, the Orbiter was unmanned; however, eighteenth century created a need for a more extensive
15
and universal measurement system and thus ushered TABLE 2.1

the creation of the original metric system, which was Common English Units
drafted in France throughout the 1790s.2 Length Feet
A set of units describes and quantifies the physical Mass Pound-mass
properties of the universe. Properties such as time, Time Second
distance, velocity, and power are examples of units. Temperature Rankine
Many of these quantities can be related through physi-
cal laws as will be exemplified throughout this chapter. TABLE 2.2
Fundamental units, however, are units that describe
Common SI Units
physical quantities from which all other units can be
derived. There are seven fundamental units; however, Length Meter
strictly speaking, all units can be derived with three of Mass Kilogram
these fundamental units: Time Second
Temperature Kelvin
1. Mass
2. Length System of Units and is abbreviated from French:
Systeme international d’unites. Originally, the system
3. Time was developed around the meter and is used in most
countries today.3 The common SI units are shown in
One early establishment of mass was the kilogram, which Table 2.2.
was defined as 1 kg to equal the mass of 1000 cm3 of water.
Length was originally defined in the metric system
to be one ten-millionth of the distance from the Earth’s 2.2.3 Absolute English and SI Units
equator to the North Pole. This measurement was very
challenging to reproduce when necessary, so in 1889 a It is important to highlight the differences between
precision bar that consisted of 90% platinum and 10% weight and mass. While mass is the measure of the
iridium was created and marked so that it would pre- amount of matter in an object, weight is the force of an
cisely represent 1 m at 0°C (32°F). Authorized metrolo- object due to gravity. An absolute unit is a unit that does
gists were allowed to travel to the International Bureau not include the gravitational acceleration. In the English
of Weights and Measurements to measure and mark system the absolute unit of mass is the pound-mass as
their own bars for regional prototypes. The meter was opposed to pound. Similarly, in the SI system the abso-
then redefined more precisely in 1960 as 1,650,763.73 lute unit of mass is simply the kilogram or kilogram-
wavelengths emitted by krypton-86 in a vacuum.2 mass while weight is measured in units of Newtons
From classic antiquity, the day was divided into 12 h or kilogram-force. By definition, weight and mass are
of daylight and 12 h of night. During the medieval related by the following:
period, the minute was introduced as the 60th part of an g
w=m (2.1)
hour and the second was introduced as the 60th part of gc
a minute. Today, a second is related to the radiation of a
specific quantum transition in cesium-133.2 where
w is weight
m is mass (absolute unit of mass)
2.2.1 English Units g is gravitational acceleration
The English unit system is a product of the early devel- gc is proportionality constant
opments of standardization in medieval England. It is
also commonly called the system of imperial units. This Applying the previous equation to Newton’s second
system of units was officially declared in 1824 by the law yields
British Weights and Measures Act and was later refined dv 1 1
and reduced until 1959. The unit system is still used F = m× × = ma × (2.2)
dt g c gc
by England and much of its former empire.3 Table 2.1
shows the basic units of this system. where
F is force
v is velocity
2.2.2 SI Units t is time
The SI unit system is the modern form of the metric sys- a is acceleration
tem created in 1960. It is often called the International gc is proportionality constant
Engineering Fundamentals 17
Thus in English units yields TABLE 2.3

Prefixes and Abbreviations
 ft  1  lbf s 2 
(
lbf = m l b m ) × 32.17  2  × 
.17  l bm ft
 Prefix Abbreviation Power
 s  32   Pico P 10−12
gc
Nano N 10−9
lbm ft Micro M 10−6
g c = 32.17 Milli M 10−3
lbf s 2
Centi C 10−2
For SI, then Kilo K 103
Mega M 106
• Kilogram-mass
Giga G 109
 m  1  s2 N  Tera T 1012
N =m kg ( m ) × 1  2  × 
 s  1  k g m m


gc
 quantities. SI prefixes are commonly used in both SI and

English systems and are shown in Table 2.3.
kg mm
gc = 1
N ⋅ s2 2.2.4 Dimensional Analysis
• Kilogram-force As discussed earlier in this section, all units can be derived

using three fundamental units: time, mass, and length. Much
kg s  2  like what has already been done throughout this chapter,
) × 9.81 ms  × 9
1
 
kg f = m k g( m
 
.81  k g m 
2
f
dimensional analysis is a powerful method engineers and
gc
 m  scientists can use to balance units in all conversions.

The most common application of dimensional analysis
kg m m is the grid method, which is a systematic way to carry and
g c = 9.81
kg f s 2 cancel units to the final desired unit. For example, take the

power required for a bicycle to travel at a velocity of 20 mph
(32 km/h). An engineer estimates that the force required
Example 2.1
to move against wind and rolling resistance of the tires is
Find the weight of a 20 kg mass on earth in 4 lbf. Applying this to the equation P = FV, we have
Newtons, kilogram-force, and pound-force.
1.0 m / s 1.0 N
m
1 s2 N P = F × V=4 lbf × 20 mph × ×
N = 20k g × 9.81 2 × = 196.12 N 2.237 mph 0.2248 lbf
m
s 1k g m
m
N⋅m
m 1kg f s 2 = 119.25 = 119.25 W
kg f = 20k g × 9.81 × = 20kg f s
m
s2 9.81 kgm m
The unit of watt has been thus derived from the equa-
For English units, convert kilogram-mass to tion of power and the conversion of given units.
pound-mass:
Example 2.2
1 lbm
20 kg × = 44.06 lbm
m .4539 kg Convert 50 mph into kph (km/h):
m
miles 5, 280 ft 0.3048 m 1 km km

ft 1lbf s 2 50 × × × = 80.47
lbf = 44.06l b m × 32.174 × = 44.06lbf h 1 miles 1 ft 1000 m h
s2 32.174 lb m ft
From this A list of conversion factors can be found in Table 2.6 at

the end of this chapter.
4.45lbf = 1 N
196.12 N = 20 kg f 2.2.5 Some Simple Derived Units

As mentioned earlier in this chapter, all units can be
∴ 9.806 N = 1 kg f
derived using three fundamental units. In this section,
Numerical prefixes help to reduce the amount of a practical approach to the derivation of simple units is
zeros shown before and after decimals in numerical offered.
2.2.5.1 Velocity Example 2.5
Velocity is defined as the rate of change in the position Find the density of water in pound-mass per
of an object. Velocity can also include direction, making cubic feet if the density is one gram per cubic
centimeter.
it a vector quantity. In some texts the vector quantity is
considered velocity while the scalar quantity is consid-
ered speed. g kg lbm 2.54 3 cm 3
1 3
× × ×
cm 1000 g 0.4530 kg in 3
∆length (2.3)
Velocity =
∆time 123 in 3 lbm
× = 62.43 3
Common units of velocity include ft 3
ft
2.2.5.4 Specific Weight
ft miles m km
s h s h
Specific weight is weight per unit volume. That is, while
density is the mass per unit volume, specific weight
accounts for gravitational force.
Example 2.3 g
γ=ρ (2.5)
Convert 100 feet per second into kilometers per gc
hour.
where
ft 0.3048 m km 3600 s km ρ is density
100 × × × = 109.8 g is gravitational acceleration
s 1 ft 1000 m h h
gc is proportionality constant
Common units of specific weight are

2.2.5.2 Volume
The 3D space occupied by a substance is volume. kg f lbf N
Common units of volume include m3 ft 3 m3

m 3 , ft 3 , cm 3 , L , gal
Example 2.6
If specific weight of air at standard temperature
Example 2.4 and pressure (STP) is
f lb
, convert to
Newton per cubic meter. .076
How many gallons are in a cubic meter? 3
ft
lb f 4.448 N ft 3 N
.076 × × = 11.94 3
ft 3 gal ft 3 lb f .3048 m
3 3
m
1m 3 × × = 264gal
.3048 3 m 3 .1337 ft 3
2.2.5.5 Pressure
2.2.5.3 Density Pressure occurs when a force is applied to a surface. It is
Density is the mass per unit volume. Something that defined as the applied force per unit area.
would be considered very dense could be a bar of gold; Force
conversely, something that has relatively low density Pressure = (2.6)
would be cork. Area
mass Common units of pressure include

ρ= (2.4)
volume
lbf N lb kg f
= psi = pascal(Pa) 2f
Common units for density are in.2 m2 ft cm 2
kg lbm lbm kg
There are also some commonly used terms to denote
m3 ft 3 in.3 L
pressures that hold special significance for scientist and
engineers, many of which will be further explained in As work can be both positive and negative, engineers
Chapter 8. have developed a formal sign convention for thermo-
dynamic systems. Heat transfer to a system and work
N
bar = 10 5 atm = 14.7 psig = 760 torr done by a system are positive while heat transfer from
m2 a system and work done on a system are considered
negative.
Example 2.7
A container is pressurized at 2 atm. What is the
pressure of the container in kPa? 2.3.1 Power
1.01 × 10 5 N/m 2 kPa Power is the rate of change in work or energy transfer.
2 atm × × = 202kPa Burning 1 kg of liquid fuel oil releases far more energy
1 atm 1000 Pa
than burning 1 kg of dynamite, but because 1 kg of
dynamite releases energy far quicker than oil, the explo-
sion of dynamite produces more power. Instantaneous
power is defined as
2.3 Work and Energy
s
ds
Work is an energy interaction between a system and its
surroundings. In a thermodynamic system, work can ∫ F ⋅ dt
o (2.8)
be any energy crossing the boundary of a closed sys-
tem that is not heat. Work can take many forms such
as shaft, spring, or electrical. Mechanical work is clas-
Average power can be calculated using the formula
sically defined as the dot product of force and length.
s s
∫o
F ⋅ ds =
∫ F cos θ ds
o (2.7) Power =
∆Work
∆Time
(2.9)
where
Common units for power are
F is force
s is length 
θ is angle between F and ds J N⋅m ft ⋅ lbf
= = Watt(w) 1 horsepower (hp) = 550
s s s
It is important to recognize that work can be zero if
force is applied perpendicular to the displacement vec-
tor. Work and energy are commonly expressed in the Example 2.9
following units:
Convert Btu into kilowatts.
kg ⋅ m 2
80 × 106
= N ⋅ m = Joule(J) Btu h
s2

Example 2.8 6 Btu 1J 1h

80 × 10 × ×
If a 5 kg box is lifted vertically 20 ft by a crane, h −4 3600 s
what is the work in N · m and ft · lbf. 9.478 × 10 Btu
m
5 kg × 9.81 = 49 N
s2 1W ⋅ s 1 kW
× × = 23 , 446 kW
 .3048 m  3
Σ Fs cos θ = 49 N ×  20 ft ×  = 298 N ⋅ m J 10 W
2.4 Temperature
 ft 

Temperature is considered by some to be a fundamental
Converting to ft · lbf:
unit; however, more stringently, it can be defined as the
lbf ft measure of energy per particle per degree of freedom,
298 N ⋅ m × × = 220 ft ⋅ lbf which can then be expressed in terms of length, mass,
4.44 N .3048 m
and time.
Temperature is commonly measured in Fahrenheit Example 2.10

(English) and Celsius (SI). The relationship between the Convert 600°F to Celsius, Kelvin, and Rankine
two measures will be derived. By first noting the freez-
ing and boiling points of water at sea level.
°R = °600 + 459.67 = 1056.97°R
• H2O freezes at 0°C and 32°F 5
°C = (600 − 32) = 315.5°C
• H2O boils at 100°C and 212°F 9
K = 315.5°C + 273.15 = 588.7 K
Plotting Fahrenheit against Celsius using these
boundaries displays a linear relationship (see Figure 2.1).
Finding the equation of the line in Figure 2.1 will yield
the relationship between the two units.
y = mx + b 2.5 Enthalpy and Specific Heat

∆y 212 − 32 180 9 Enthalpy and specific heat are very commonly used
m= = = =
∆x 100 100 5 thermodynamic properties that will be explained fur-
ther in Chapter 3. Enthalpy (h) is the measure of total
b = 32
energy content of a substance in a thermodynamic sys-
Yielding the relationship tem and is discussed in depth in Chapter 3.
9 energy
°F = ∞C + 32 h= (2.13)
5 mass
(2.10)
5
∞C = ( ∞F − 32) Common units for enthalpy are
9
Absolute temperature is based upon the kinetics of gas Btu kJ
molecules and scales the zero to lowest attainable tem- lbm kg
perature where gas molecules exhibit extremely minimal
amounts of movements. Units for absolute temperature
are Rankine (English) and Kelvin (SI). Absolute tempera- Specific heat, sometimes called heat capacity, is a
tures are commonly used in ideal gas calculations. parameter that relates enthalpy to temperature. It is the
energy required to raise the temperature of a unit mass
Rankine :† °R = °F + 459.67 (2.11) of a substance by one degree.
Kelvin : †K = °C + 273.15 (2.12) energy

Cp = (2.14)
∆temp(Unit Mass)
Common units of specific heat

212
Btu kJ Kcal
lbm °F kg∞k g∞C
°F
Example 2.11
Water has a specific heat of 1cal .
C p =
32 Convert Cp to units of
Btu and J g∞C
lb ∞F kg∞C
cal 4.186 J mBtu 1000 g
1 × × ×
g ∞C 1 cal 1.055 × 10 3 J kg
0 °C 100
0.454 kg 1° C∆ 1.00 Btu

FIGURE 2.1 × × =
Celsius and Fahrenheit relation. lbm 1.8°F∆ lbm °F
It is important to note here that In liquids, additional intermolecular forces exist

between molecules contributing to additional shear
cal Btu stress. Thus, the viscosities of liquids are several orders
1 =1
g∞C lbm ∞F of magnitude higher than that of gases. Further, in liq-
uids, viscosity decreases with an increase in tempera-
For the second conversion ture, while, in gases, viscosity decreases with decreasing
temperatures.4
4.186 J 1000 g
Units of viscosity are often given in
cal J
1 × × = 4186
g ∞C 1 cal kg kg∞C lbf ⋅ s lbm kg N⋅s
= 2 = Pa ⋅ s
ft 2 ft ⋅ s m⋅s m
In fluid flow calculations, viscosity very often appears
divided by density. It is used so often that engineers have
made it a fluid property and named it kinematic viscosity.
2.6 Viscosity of Gas and Liquid
µ
Viscosity, sometimes called dynamic viscosity, is the υ=
measure of a fluid’s resistance to shear deformation (see ρ
Chapter 8). For example, crude oil will pour more slowly
than water from an identical beaker with the same The units are commonly given in
angle. The slow flow of oil implies a low rate of strain.
ft 2 m2
Thus oil has a higher resistance to shear deformation s s
than water and correspondingly, it is said to have higher
viscosity than water.
The term stoke is often used to represents m2/s divided
Consider the fluid flow over a flat plate. At the bound-
by 10−4, likewise centistokes is m2/s divided by 10−6.
ary surface of the plate a fluid has the same velocity as
The petroleum industry sometimes measures kine-
the surface. That is to say that in the case of a stationary
matic viscosity in Saybolt Seconds Universal (SSU) and
surface, velocity of the fluid is zero at the surface (see
is defined as follows:
Figure 2.2). It is also important to note that in this case,
both liquids and gases are considered fluids. y = 4.6243(Centistokes) + 66.687

Absolute viscosity is defined as
Example 2.12
Shear Stress Force/Area
µ= ⇒
dV/dy Velocity/Length The viscosity of water at normal condition is
lbf ⋅ s . Find the viscosity of water at
Force × Time 3.732 × 10 −5
=
Areaa
(2.15) ft 2 in
normal condition . N⋅s
m2
lb f ⋅ s 4.448 N 10.763 ft 2 N⋅s
3.732 × 10 −5 × × = 1.786 × 10 −3 2
ft 2 lb f 1m 2 m
V
The viscosities of some common fluids are presented in
Table 2.4.
dy
y TABLE 2.4
Dynamic Viscosity of Common Fluids
dV
Referenced at Normal Conditions
Absolute Viscosity
Fluid at 32°C mPa-s lbf-s/ft2
Air 0.0186 0.3886
Stationary solid wall Methane 0.0112 0.2340
Carbon dioxide 0.0151 0.3154
FIGURE 2.2
Water 0.7970 16.6493
Velocity distribution next to a boundary.
Example 2.13 TABLE 2.5

Convert 2 GPM of fuel oil #2 to lbm/h to Btu/h, Samples of the Two Decimal
where the density and heating value of oil is 57.4 Notations
lbm/ft3 and 16,791 Btu/lbm, respectively. Standard Continental European
1.125 1,125
gal 1 ft 3 60 min lb lb
2 × × × 57.4 m3 = 920.8 m 1,000,000.25 1 000 000,25
min 7.48 gal 1h ft h
Btu Btu
× 16, 791 = 15.46 × 106
lbm h Liquid water is considered an incompressible fluid and
gases like air are highly compressible fluids. The mea-
sure of compressibility is defined by the bulk modulus,
sometimes called bulk modulus of elasticity.
dp Pressure Change
Ev = − =− (2.16)
2.7 Bulk Modulus dV / V Volume Change Fraction
Fluids can often be categorized as either compressible or
incompressible. Compressible fluids will significantly where
decrease in volume under an applied pressure, while dp is differential change in pressure
the volumes of incompressible fluids generally undergo dV is differential change in volume
very miniscule changes with large applied pressures. V is volume of fluid
TABLE 2.6
Conversion Factors
Length L L 1  m = 3.281  ft = 1.094  yd = 39.37  in. = km/1000
1  ft =.3048  m = 12  in. = mile/5280  = km/3281
1  mm = m/1000  = in./25.4  = 39.37 mile
Velocity V L/T 1  m/s = 3.6 km/h = 3.281 ft/s = 2.237 mph = 1.944 knots
1 ft/s =.03048 m/s = 0.6818 mph = 1.097 km/h =.5925 knots
Mass m M 1  kg = 2.205  lbm = 1000  g = slug/14.59
1 lbm = kg/2.205  = slug/32.17-453.6 g = 16 oz
Density ρ M/L3 1000 kg/m3 = 62.43 lbm/ft3 = 8.345 lbm/gal
Force F ML/T2 1 lbf = 4.448  N = 32.17  lbm·ft/s2
1  N = kg · m/s2 = 0.2248 lbf
Pressure P M/LT2 1  Pa = N/m2 =.00001 bar =.007501 torr
1 atm = 101.3  kPa = 2116 psf = 1.013 bar = 14.70  psi = 760 torr
1  psi = atm/14.70  = 6.895  kPa = 51.71 torr
Volume ρ L3 1 m3 = 35.31 ft3  = 1000  L = 264.2  gal
1 ft3 = 0.02832 m3  = 28.32  L = 7.481  gal
1  gal = 231 in.3 = barrell (petroleum)/42 = 3.785 L = 0.003785 m3
Volume flow rate L3/T 1 m3/s = 35.31  ft3  = 2119  cfm = 264.2  gal/s = 15850  gal/m
1  cfs = 1  ft3/s = 28.32  L/s = 7.481  gal/s = 448.8  gal/m
Mass flow rate Q M/T 1  kg/s = 2.205  lbm/s =.06852 slug/s
Energy and work E,W ML2/T2 1  J = kg · m2/s2  = N · m = 0.7376  ft·lbf =.0009478 Btu =.2388 cal
Power P,Ė ML2/T3 1  W = J/s = N · m/s = kg·m2/s3 = 0.001341 hp
= 0.7376 ft ·lbf/s =.0009478 Btu/s
1  hp = 0.7457  kW = 550  ft·lbf/s = 33,000  ft · lbf/min = 2544  Btu/h
Angular speed ω 1/T 1 rad/s = 9.549 rpm =.1591 rev/s
Viscosity μ M/LT2 1 Pa·s = kg/m · s = N · s/m2 = 0.02089 lbf·s/ft2 = 0.6720 lbm/ft · s
Kinematic viscosity ν L2/T 1 m2/s = 10.76  ft2/s
Temperature T — K = °C + 273.15
°C = (°F − 32)/1.8
°R = °F + 459.67
°F = 1.8°C + 32
The units for bulk modulus are Measures, this notation is used mostly by continental
European countries (see Table 2.5).
Unit Area
Ev =
Force
Commonly used units include

2.9 Conclusion
m2 in.2
= Pa −1 atm −1
Units and conversions (Table 2.6) are vital knowledge to
N lbf
all aspects of engineering and is the most important tool
for deriving equations and identifying mistakes. In this
chapter the origins and distinctions of the English and
Metric systems have been reviewed, commonly used units
have been derived, and dimensional analysis has been
2.8 Slang Units and Notations applied. It is common for engineers to look up conversions
in derived charts; however, as exemplified throughout the
In practice, engineers, technicians, designers, and chapter, engineers are strongly recommended to follow
operators often find it convenient to abbreviate com- through the m echanics of units and conversions to develop
monly used units. The term “pounds” is often used to a deeper understanding of engineering principles.
denote the units of pressure pounds per square inch
(psi). Further it is also common to use the “#” symbol to
denote pound-weight.
In comparing the energy content of various grades of
natural gas and other fuels, the units MMB and MKB References
are often used. Although these units are not valid
1. NASA, Mars climate orbiter mishap investigation board
through the governing body of metrology and are tech-
phase I report, Press release, November 10, 1999.
nically incorrect, MMB or MMBtu represent million 2. H. Arthur Klein, The World of Measurements: Masterpieces,
Btu burned per hour, while MKB denotes thousand Btu Mysteries and Muddles of Metrology, Simon and Schuster,
burned per hour. New York, 1974.
The use of a comma instead of a decimal point to 3. Marks’ Standard Handbook for Mechanical Engineers,
represent a decimal marker is an area of possible con- McGraw Hill, New York, 2006.
fusion. Officially declared valid in the 10th resolu- 4. C. Crowe et al. (ed.), Engineering Fluid Mechanics, 9th
tion of the 2003 General Conference of Weights and edn., John Wiley & Sons Inc., Hoboken, NJ, 2009.
3
Combustion Fundamentals
Stephen B. Londerville, Joseph Colannino, and Charles E. Baukal, Jr.
CONTENTS
3.1 Introduction...................................................................................................................................................................... 26
3.2 Uses for Combustion....................................................................................................................................................... 26
3.3 Brief Overview of Combustion Equipment and Heat Transfer................................................................................ 26
3.4 Chemical Combustion Fundamentals.......................................................................................................................... 27
3.4.1 States of Matter..................................................................................................................................................... 27
3.4.2 Physical Properties of Matter............................................................................................................................. 27
3.4.3 Chemical Structure.............................................................................................................................................. 28
3.4.4 Periodic Table....................................................................................................................................................... 28
3.4.5 Equations and Avogadro’s Number.................................................................................................................. 28
3.5 Gaseous State.................................................................................................................................................................... 30
3.5.1 Kinetic-Molecular Theory.................................................................................................................................. 30
3.5.2 Gas Laws............................................................................................................................................................... 30
3.5.3 Standard and Normal Air.................................................................................................................................. 32
3.5.4 Properties of Air................................................................................................................................................... 32
3.5.5 Humidity............................................................................................................................................................... 32
3.5.6 Psychrometric Chart............................................................................................................................................ 33
3.5.7 Dalton’s Law of Partial Pressures, Saturation, and Superheated Vapor...................................................... 33
3.6 Oxidation Reduction Equations..................................................................................................................................... 33
3.6.1 Redox Reactions of Gaseous Fuels and Excess Air......................................................................................... 34
3.6.2 Flue Gas................................................................................................................................................................. 34
3.7 Air-to-Fuel Ratio.............................................................................................................................................................. 35
3.7.1 Air-to-Fuel Mixture Ratio................................................................................................................................... 39
3.7.2 Air-to-Fuel Mass Ratio........................................................................................................................................ 40
3.7.3 Turbine Exhaust Gas........................................................................................................................................... 41
3.8 Chemical Thermodynamics........................................................................................................................................... 43
3.8.1 Enthalpy, Entropy, and Heat Capacity.............................................................................................................. 43
3.8.2 Heat of Combustion............................................................................................................................................. 43
3.8.3 Adiabatic Flame Temperature............................................................................................................................ 44
3.8.4 Dissociation.......................................................................................................................................................... 45
3.9 Practical Liquid Fuels...................................................................................................................................................... 46
3.10 Combustion Kinetics....................................................................................................................................................... 47
3.10.1 Thermal NOx Formation..................................................................................................................................... 48
3.10.2 Prompt NOx Formation....................................................................................................................................... 48
3.10.3 Fuel-Bound NOx................................................................................................................................................... 49
3.11 Flame Properties.............................................................................................................................................................. 49
3.11.1 Flame Temperature.............................................................................................................................................. 49
3.11.2 Available Heat...................................................................................................................................................... 52
3.11.3 Minimum Ignition Energy................................................................................................................................. 53
3.11.4 Flammability Limits............................................................................................................................................ 53
3.11.5 Flame Speeds........................................................................................................................................................ 56
3.12 Substoichiometric Combustion...................................................................................................................................... 57
3.12.1 Equilibrium and Thermodynamics.................................................................................................................. 57
3.12.2 Substoichiometric Combustion Revisited........................................................................................................ 57
25
3.13 General Discussion.......................................................................................................................................................... 58

3.13.1 Air Preheat Effects............................................................................................................................................... 58
3.13.2 Fuel Blend Effects................................................................................................................................................. 58
3.14 Emissions.......................................................................................................................................................................... 68
3.15 Quick Sizing..................................................................................................................................................................... 68
3.15.1 Finding Saturated Humidity.............................................................................................................................. 68
3.15.2 Stoichiometric Combustion of Air Simplified................................................................................................. 69
3.15.3 Density of Low-Pressure Gases......................................................................................................................... 69
References................................................................................................................................................................................... 69
Radiant heat transfer requires a line-of-sight to the

flame. Only this heat transfer mechanism can operate in
3.1 Introduction
a vacuum. For example, the earth receives essentially all
Combustion is the controlled release of heat from the its heat from the sun through this mechanism. Convection
chemical reaction between a fuel and an oxidizer. The requires the bulk movement of a hot fluid. In a boiler,
fuels in the refining, petrochemical, and power gen- hot combustion gases transfer heat to the outer tube wall
eration industries are almost exclusively hydrocarbons. via convection. Convection occurs naturally by means
Hydrocarbons are composed of only hydrogen (H) and of buoyancy differences between hot and cool fluids—
carbon (C) in their molecular structure. Natural gas and termed natural convection—or by motive devices such as
fuel oil are examples of hydrocarbon fuels. Practical fans or blowers creating high velocities—termed forced
fuels containing small contaminants such as nitrogen convection. Heat transfers from the outer to inner tube
(N2) and oxygen (O2) are described later in this chapter wall by conduction—the predominant heat transfer mode
and in other chapters of this book such as Chapter 4. through metals. Inside the tube, convection is the pre-
dominant mode of heat transfer to the inside fluid. A more
complete discussion of heat transfer is given in Chapter 7.
Most large boilers have water in the tubes and fire
3.2 Uses for Combustion outside—called water-tube boilers (see Figure 3.1). Fire-
tube boilers put the fire and hot gases in tubes sur-
Combustion is used either directly or indirectly to pro-
rounded by water. This system is applicable to smaller,
duce virtually every product in common use. Combustion
unattended boilers.
processes produce and refine fuel, generate electricity,
Process heaters are akin to water-tube boilers, but
prepare foods and pharmaceuticals, and transport goods.
with some very important differences. First, process
Fire has transformed humankind and separated it from
heaters contain a process fluid in the tubes, rather than
the beasts, illuminated nations, and safeguarded genera-
water. The process fluid is usually a hydrocarbon, for
tions. It has been used in war and peace, to tear down
example, crude oil. Process heaters come in two main
and build up; it is both feared and respected. It is a most
varieties: vertical cylindrical (VC) and cabin style. VCs
powerful tool and worthy of study and understanding.
comprise a cylindrical flame zone surrounded by pro-
cess tubes. Cabin-style heaters are rectangular with wall
and roof tubes. The radiant section comprises the space
3.3 Brief Overview of Combustion surrounded by tubes having a direct view of the flame.
Most process heaters also have a convective section com-
Equipment and Heat Transfer
prised of overhead tubes that cannot directly view the
In the process industries, combustion powers gas tur- flame. Convective tubes receive their heat from the direct
bines, process heaters, reactors, and boilers. The burner contact of the combustion gases. The transition from the
combusts fuel and generates products of combustion radiant to convective sections is known as the bridgewall.
and heat. A firebox contains the flame envelope. In Reactors such as cracking furnaces and reforming fur-
boilers, the fire heats water in the tubes to boiling. The naces are more extreme versions of process heaters. Here,
steam rises to a steam drum that separates the liquid the process fluid undergoes chemical transformations to a
and vapor phases, returning water to the tubes and different substance. For example, in an ethylene cracking
passing steam. The steam may be further heated in a furnace, liquid or gas feedstock transforms to ethylene
superheater. Superheaters raise the temperature of the (C2H4), an intermediate in the production of polyethylene
steam above the boiling point, using either radiant and/or and other plastics. There are many specialized types of
convective heat transfer mechanisms. reactors using combustion as the heat source.
Combustion Fundamentals 27
FIGURE 3.1
“O” type package boiler. (Courtesy of Victory Energy Operations, LLC.)
expand to fill a closed container. Gaseous matter is the

most compressible state with a substantially lower bulk
3.4 Chemical Combustion Fundamentals
modulus than liquids.1
Chemistry is the fundamental backbone of combustion.
Advanced concepts in combustion are founded upon 3.4.2 Physical Properties of Matter
the application of chemistry. It is therefore important
to review and highlight fundamental concepts that will Mass is the measure of the amount of matter within
play a key role in later topics. a substance. Mass differs from weight as mass is an
invariable quantity and does not depend on the accel-
eration of gravity. The International System of Units (SI)
3.4.1 States of Matter unit for mass is the kilogram (kg) while imperial units
are measured in pound mass (lbm). See Chapter 2 for a
Matter is the term for the substance that all physical
discussion about units.
objects consist of. Matter includes atoms and other par-
Volume is the space that a body of matter occupies. The
ticles that have mass and occupy volume. Classically,
SI unit for volume is cubic-meter (m3) and cubic-feet (ft3)
matter exists in three distinct states: solid, liquid, and
in imperial units. Density is defined as the mass per unit
gaseous vapor.
volume of a substance. Density is normally represented by
In solid state matter, molecules are packed closely
the Greek letter ρ (rho). Density is commonly expressed in
together. The forces between these molecules are strong
(kg/m3) for SI units and (lbm/ft3) for imperial units:
enough to create an object that is rigid and occupies a
definite shape and volume. As a result, particles in solid
state matter are not free to move and can only vibrate. Mass
Density = (3.1)
Solid matter is the most difficult to compress and gener- Volume
ally has a very high bulk modulus.
In liquid state matter, molecules flow easily despite Specific gravity is an important parameter that is used
strong intermolecular force between molecules. Liquids to compare many parameters and is defined as the ratio of
take shape of the container and will remain in an open the density of a fluid to the density of water. In the case of
container. Liquid matter is difficult to compress, but still gases, specific gravity is the ratio of the density of the gas
has a considerably lower bulk modulus than solids. to the density of air.1 The density of air and water used for
In gaseous state matter, molecules move freely with specific gravity are referenced at standard air conditions,
little interactions except during collisions. Gases which are discussed in a later section of this chapter.
3.4.3 Chemical Structure nonmetallic elements with one another. Carbon diox-
ide is an example of covalent bonding as oxygen atoms
Atoms are the almost infinitesimally small building blocks
share their excess electrons with carbon.2
of matter. The early experiments of Ernest Rutherford
Bond energy is the energy change that accompanies
and Niels Bohr depict the atom with a densely packed
the formation of a chemical bond. In bond formation,
center, called a nucleus, surrounded by a large cloud of
energy is either released (exothermic) or absorbed (endo-
negatively charged electrons. The nucleus is composed
thermic). The amount of energy released or absorbed
of electrically neutral neutrons and positively charged
determines the bond strength; the larger the energy
protons. Atoms with different quantities of neutrons,
difference the more stable the bond. Compounds with
protons, and electrons make up different elements.2
weak bond energy’s tend do readily dissolve within a
The first category of substances is an element. An ele-
solution whereas compounds with stronger bond ener-
ment is a pure substance that cannot be decomposed
gy’s may not be readily soluble.3
chemically. Currently 116 elements are known, with
familiar examples being carbon (see Figure 3.2), nitro-
gen, and oxygen. Each element is assigned a unique 3.4.4 Periodic Table
atomic number which is equal to the number of protons
Every element is represented by one or two unique
found in the nucleus of the atom.
letters that often abbreviate the full element name in
The second category of substances is a molecule.
English, Latin, or German, for example, the symbol C
Molecules are the smallest collection of chemically bound
represents carbon, H hydrogen, and Ca calcium.
atoms that define a substance, for example, nitrogen, which
A chemical formula is a symbol or group of symbols
exists as the gaseous molecule N2 at room conditions.
that represent the elements and their respective quanti-
Physical and chemical properties of an atom are pri-
ties which make up the composition of a substance. The
marily determined by the energy associated with the
formula for water is H2O, which specifies two atoms of
movement and number of electrons, especially pertain-
hydrogen and one atom of oxygen.
ing to the outer most electrons of an atom called valance
In the early 1869, Dmitri Mendeleev developed the
electrons. The exchange of electrons between atoms is
periodic table through chemical observations to predict
the result of chemical bonding.
trends in and group regular patterns of chemical behav-
Ionic bonding holds two or more atoms together
ior. Consequently, the periodic table organizes elements
through electrostatic forces that exist between ions of
by increasing atomic number in the horizontal rows and
opposite charge. Ionic bonding most commonly occurs
by similar properties in the vertical columns.3
between a metallic element (cation) and a nonmetallic
The periodic table (see Figure 3.3) can be separated into
element (anion). Table salt is a familiar example of an
two main categories—metals and non-metals. With the
ionic bond, where the positively charged sodium bonds
exception of hydrogen, all metals are located to the left and
with the negatively charged chlorine.
in the middle of the table. These elements are prone to hav-
In covalent bonding, two or more atoms are bonded
ing fewer valance electrons than protons, giving the ele-
together from sharing electrons. Such bonding
ment a positive charge also known as a cation. Conversely,
most commonly occurs between the interactions of
non-metals are located to the right of the table. Non-metals
are prone to having more valance electrons than protons,
–
6 protons
giving the element a negative charge known as an anion.
6 neutrons Elements on the 8A column, often called noble gases,
include He, Ne, Ar, Kr, Xe, and Rn, which have completely
–
Electrons filled their respective valance electron orbital, making
their chemical reactivity very low. Elements in the 1A col-
– umn have relatively empty valance electrons and are thus
–
very chemically reactive in comparison to noble gases.
3.4.5 Equations and Avogadro’s Number

– Chemical reactions can be represented concisely
through chemical equations:
2H 2 + O 2 → 2H 2O
–
FIGURE 3.2 In a chemical equation, the starting substances, or reac-

Carbon atom with six protons, neutrons, and electrons. tants, are placed on the left side of the arrow and to the
Combustion Fundamentals
FIGURE 3.3
29
Periodic table. (From Leigh, G.J. (ed.), Nomenclature of Inorganic Chemistry, Blackwell Scientific Publications, Oxford, U.K., 1990; Chemical and Engineering News, 63(5), 27, 1985.)
right of the arrow are the substances produced by the 3.5.1 Kinetic-Molecular Theory
reaction called the products.
The kinetic-molecular theory was developed in the
From the law of conservation of mass, atoms are nei-
nineteenth century to describe the motion of gaseous
ther created nor destroyed in a chemical reaction, thus a
molecules and provides the foundation for gas laws.
chemical equation must have an equal amount of items
This theory can be summarized in three postulates:
on both the reactant and product side. Once the formula
of the reactants and products are known, the equation
can be balanced by determining the coefficients that • Gases are composed of molecules which occupy
provide the equal number of each atom on the reactant a volume that is far smaller than the total vol-
and product sides2: ume occupied by a gas. Molecules are relatively
far apart.
CH 4 + O 2 → CO 2 +H 2O • Gas molecules are in continuous motion mov-
Step 1: Unbalanced equation ing in straight lines, often colliding with one
another. Pressure is formed when gas molecules
collide with the walls of the container.
CH 4 + O 2 → CO 2 + 2H 2O
• The kinetic energy of gas molecules increases
Step 2: Match H on product side as temperature rises. Molecules travel at higher
velocities with increasing pressure.3
CH 4 + 2O 2 → CO 2 + 2H 2O
Step 3: Match O2 on reactant side
3.5.2 Gas Laws
The weight of an atom is infinitesimally small; for From the kinetic-molecular theory, early experiments
example, a hydrogen atom weighs 1.67 × 10−24 g (3.68 × with gases reveal four variables are needed to define
10−27 lbm). Expressing weight in this way poses an incon- the physical state of a gas: temperature, pressure,
venience to scientists and engineers. Thus, atomic weight volume, and the amount of gas expressed in moles.
is introduced as the relative weight of atoms in different Boyle’s law states that for a fixed amount of gas at a
elements which is proportional to their actual weight. fixed temperature, pressure and volume are inversely
Molecular weight uses atomic weight as the mass of an proportional:
element of a molecule. For example, the element oxygen
has an atomic weight of 16 atomic mass units (amu) and PV = Constant (3.2)

thus a molecular weight of 16 grams per mole (g/mol).
Likewise, imperial units may be used as well, H2O has an where
atomic weight of 18 amu [(2 × 1) + 16] and thus a molecu- P is the pressure of the gas, psia or kPa
lar weight of 18 pounds per mole (lb/mol). V is the volume of the gas, ft3 or m3
The number 6.023 × 1023 was proposed by Jean
Perrin in honor of Avogadro, who found that the Charles’s law states that for a fixed amount of gas
volume of gas is proportional to the number of atoms. at a fixed pressure, the volume and temperature are
Avogadro’s number is the exact number of carbon proportional:
atoms in 12 g of the isotope carbon-12. Further 6.023 ×
1023 is now used to represent the exact number of V
= Constant (3.3)
atoms in 1 gmol of any molecule.2 With Avogadro’s T
number, one can calculate the weight of a single atom
of any element. where
T is the absolute temperature, °R or °K
V is the volume of the gas, ft3 or m3
The ideal gas law is the culmination of all gas laws as

3.5 Gaseous State it is derived from combining Boyle’s law, Charles’s law,
and Avogadro’s law. The law applies for typical com-
Combustion is chemical reaction that occurs between bustion reactions and relates the pressure, volume, and
gaseous air and fuel. Further, many fuels both exist and number of moles:
are ignited in their gaseous state. The study of gases
will help to predict vital parameters of air and fuels. PV = nRT (3.4)
where Finally, multiply the density by the volume to

P is the pressure of the gas, psia or Pa obtain the weight of gas, m = ρV = 0.126  lbm/ft3 ×
V is the volume of the gas, ft3 or m3 1000 ft3 = 126 lb. Furthermore, using the ideal
n is the number of moles gas law can determine the molar density for
any gas:
R is the gas constant =
10.73 psia-ft3/lb mol-°R or
8.314 kJ/kgmol/K PV = nRT

T is the absolute temperature, °R or K
where
Degrees Rankine (°R) are defined as the number of n is the molar density, lb mol/ft3
Fahrenheit degrees above absolute zero, the coldest pos- R is the gas constant = 1545 ft-lbf/lb mol-°R
sible theoretical temperature. Equation 3.4 shows that
gas volume and moles are directly proportional. 2117 psfa lb mol
n= = 0.00258
Another useful form of the ideal gas law is  ft-lb f  ft 3
1545   530 °R
 lb mol-°R 
PM = ρRT (3.5) Inverting the value yields
ft 3
where = 387
lb mol
ρ is the density of the gas, lbm/ft3 or kg/m3
M is the molecular weight of the gas, lb/lb mol or for any gas at STP, independent of molecular weight
g/mol (MW).
Another helpful equation derived from the ideal gas
Also the ideal gas law can provide solutions in terms law (used to relate with MW) can be used to calculate
of mass by replacing mass with number of moles (n) in density of any gas relating to temperature and pressure:
Equation 3.4 and replacing the gas constant (R) with the
corresponding constant for mass. An example best rein- PV = mRT
forces these points.
m P
ρ= =
Example 3.1 V RT
Given a 1000 ft3 vessel contains methane at with
30 psig at 70°F. How many lb mol of methane does
ft-lb f lb mol
the vessel contain? What is the gas density? How R = 1545
much does the gas weigh? The solutions follow. lb mol°R(MW)lbm
At standard temperature and pressure (STP, see
next section for more explanation), the pressure At STP, the density of air is 0.075 lbm/ft3 and MW is 28.85:
will be 14.7 psia or 2117 lb/ft2 and absolute tem-
perature 529.7°R (or rounding to yield 530°R). 2117 psfa
0.075 =
From Equation 3.4,  1545   ft-lb f 
  530°R
28.85   lbm -°R 
PV (30 + 14.7 )[psia] × 1000 [ft 3]
n= = Since
RT  psia ft 3 
10.73   ( 459.7 + 70)[°R]
 lb mol°R  R P MW
PV = m T and ρ= and
= 7.86 lb mol MW RT

P MW
R= = Constant
From Equation 3.3 ρT
PM (30 + 14.7 )[psia] × 16.05[lb lb mol] Then, the density of any gas can be related to another
ρ= = via
RT  psia ft 3 
10.73   ( 459.7 + 70)[°R ]
 lb mol°R  MW2 P2 T1
ρ2 = ρ1 (3.6)
= 0.126 lb/ft 3 MW1 P1 T2
Example 3.2 60°F (15.6°C) = Air @ 14.7 psia (1 atm) at 60°F

For methane (CH4) at 100°F and 10 psig, what is Density: 0.075 lb/ft3 (1.201 kg/m3)
the gas density?
From Equation 3.6, using air at STP, the gas den- 32°F (0°C) = Air @ 14.7 psia (1 atm) and 32°F (0°C)
sity will be Density: 0.080 lb/ft3 (1.281 kg/m3)
MWmethane P2 530 Atmospheric pressure decreases with elevation and can

ρmethane = 0.075
28.85 14.7 T2 be corrected using the following formula1:
16 (10 + 14.7 psi ) × 530°R g/αR
= 0.075  T − α( z − z o ) 
28.85 × 14.7 psi × ( 460 + 100 ) °R p = po  o 
(3.7)
 To 
lb
= 0.066 where
ft 3
p is the atmospheric pressure
Alternate units of R po is the known reference atmospheric pressure
To is the known reference temperature
8.314 J K −1 mol −1 z is the elevation
zo is the known reference elevation
8.314 × 10 −5 m 3 atm K −1 mol −1 α is the lapse rate = 5.87 × 10−3 K/m
g is the acceleration of gravity
1545 ft-lb f -lb mol −1 °R −1 R is the specific gas constant = 1716 ft/slug × °R
10.731 ft 3 psi °R −1 lb mol −1

3.5.4 Properties of Air
1.986 Btu-lb mol °R −1 −1
As air is a key component of combustion, the composi-
tion of air will be very important in future calculations to
82.057 cm 3 atm K −1 mol −1 predict chemical reactions. According to research from
the National Oceanic and Atmospheric Administration,6
62.363 L Torr K −1 mol −1 the average composition of air is as shown in Figure 3.4.
Table 3.1 lists the density of dry air using 79% nitrogen
and 21% oxygen as well as average composition which is
3.5.3 Standard and Normal Air sometimes used. The molecular weight of average air is
Air pressure and density are standardized at the weight 28.966 while the 79% nitrogen and 21% oxygen assump-
of the earth’s atmosphere exerted on a surface at sea tion has the molecular weight of 28.842. For most engi-
level. There are two commonly used standards for air neering calculations these are small differences.
pressure and density: normal temperature and pressure
(NTP) and standard temperature and pressure (STP).4 3.5.5 Humidity
STP is defined by IUPAC (International Union of Pure In this chapter, the calculations for dry air have been
and Applied Chemistry) as referring to 0°C (273.15 K, idealized. However, the amount of water vapor in the
32°F) and 105 Pa. Alternately, STP is defined in Imperial air is important in many applications such as fan sizing,
and U.S. system of units as air at 60°F (520°R) and
14.696 psia (15.6°C, 1 atm). In the United States, NTP is
sometimes referenced to 1 atm and 32°F (0°C). To fur-
Nitrogen: 78.084%
ther complicate the situation, there are other standards
Oxygen: 20.948%
such as SATP, ISA, and ICAO. It is best to be clear on
Argon: 0.934%
actual temperature and pressure used as a reference
Carbon dioxide: 0.031%
and to state what conditions are being used for “STP” or
Neon, helium, methane, sulfur
“NTP.” Further, it is important to note that the two stan- dioxide, hydrogen, and other
dards also apply for other gases such as gaseous fuels.5 minor gases: 0.003%
Many times the properties for air are based on the
composition of 78% nitrogen, 21% oxygen, and 1% argon.
For combustion calculations, dry air is often defined as
FIGURE 3.4
21% O2 and 79% N2. This composition for dry air is used Composition of air by volume. (Modified from McQuiston, C. and
throughout this chapter for computations. For actual siz- Parker, J.D., Heating, Ventilating, and Air Conditioning, John Wiley &
ing of equipment, Ar and moisture should be considered: Sons, New York, 1982.)
TABLE 3.1 gas mixture is the sum of the partial pressure for each
Density Comparison of Average Air and 79% N2 21% individual gas molecule:
O2 Assumption
n
Density of Air
lb/ft3 kg/m3
∑P
PTotal =
i=1
i (3.11)
Temp (°F) Average 79/21 Average 79/21 Example: PTotal = PN 2 + PO2 + PCO2
0 0.0805 0.0802 1.290 1.206
60 0.0763 0.0761 1.222 1.217 (Example of Dalton’s law of partial pressures for nitro-
70 0.0748 0.0745 1.1982 1.194 gen, oxygen, and carbon dioxide.)
Saturation is the process in which liquid boils into
flame temperature, and air heaters/dryers. Three types of
a vapor phase. Saturation temperature and pressure
humidity exist: absolute, relative, and specific humidity.
are the conditions required for saturation to occur.
Absolute humidity is the mass of water per unit vol-
Saturation temperatures and saturated pressures vary
ume of air:
for each liquid, but are always directly related: as satu-
Mass vapor ration pressure increases for a liquid so does saturation
Absolute humidity = (3.8) temperature. Thus for a given saturation temperature
Volumenet
there is a corresponding saturation pressure.
Relative humidity is the ratio of moisture in the air at a In superheated vapors, the temperature exceeds the
given temperature and vapor pressure to the maximum saturation temperature at corresponding saturation
amount of moisture air can hold at that temperature and pressure. All molecules of a superheated vapor are com-
saturation pressure. Relative humidity ranges from 0 for pletely in the vapor phase.
completely dry air to 100 for saturated air:
Pv
ϕ= (3.9)
Psat 3.6 Oxidation Reduction Equations
Specific humidity is the ratio of the mass of vapor to Combustion is a chemical reaction in which fuel under-
mass of dry air: goes oxidation. Oxidation and reduction, often called
“Redox,” are chemical reactions that transfer electrons
Mass vapor between reactants. In an oxidation, an atom becomes
Specific humidity = (3.10)
Massdry-air more positively charged by losing an electron:
3.5.6 Psychrometric Chart Li(s) → Li + ( aq) + e −
When working with gas–vapor mixtures, calculating Na(s) → Na + ( aq) + e −

the physical properties can become tedious. Thus, engi-
neers have tabulated air–vapor properties graphically Ca(s) → Ca 2 + ( aq) + 2e −

in a complex chart generally referred to as the psychro-
metric chart. The chart relates six parameters: dry bulb (Examples of oxidation reactants)
temperature, wet bulb temperature, relative humidity, Conversely, reduction occurs when an atom gains an
humidity ratio, enthalpy, and specific volume. If any two electron and is more negatively charged:
of the mentioned independent parameters are known,
the four other properties can be looked up through the 0 0 +1 −1
H 2 + Cl 2 → 2 H Cl
chart. See Section 13.5 for a practical method of finding
saturated humidity through steam tables. (Example of a redox equation where H2 is oxidized by
Cl2 to form HCl.)
Oxidation numbers are assigned for good book-
3.5.7 Dalton’s Law of Partial Pressures,
keeping of electrons gained by reduction and lost by
Saturation, and Superheated Vapor
oxidation. In the earlier example, hydrogen’s oxida-
Gases are usually composed of a mixture of molecules. tion number increased from 0 to +1 as it is oxidized.
Partial pressure is the pressure that each molecule would Similarly, chlorine’s oxidation number decreases from 0
exert if it occupied the same volume alone at the same to −1 because it is reduced.
temperature; it is also representative of the volume frac- The method of balancing redox reactions is defined
tion of an enclosed component. The total pressure of the as follows:
1. CH4 + O2 → CO2 + H2O This introduces the concept of excess air. Excess air is
2. CH4 + O2 → CO2 + 2H2O an added amount of air to the burner system which is in
excess of the amount required for perfect stoichiometric
3. CH4 + 2O2 → CO2 + 2H2O
combustion. Excess air is intentionally added to provide
a. Balance C a more thorough mixture for more complete combustion
b. Balance H and control of combustion chamber temperatures or to
c. Balance required O2 produce hot air. Excess air, represented by ɛ, is introduced
into the chemical reaction as a multiplier of air (predomi-
3.6.1 Redox Reactions of Gaseous nantly O2 and N2):
Fuels and Excess Air %Excess Air
ε= + 1 (3.12)
As discussed earlier, fuels in their purest form are com- 100
posed of hydrogen and carbon, called hydrocarbons. CH 4 + 2ε(O 2 + 3.76N 2 )
Very few fuels originally contain oxygen and nitrogen,
but some compounds do have the tendency to pick up → CO 2 + 2H 2O + 2(ε − 1)O 2 + 2ε 3.76N 2

oxygen, nitrogen, and sometimes sulfur. Fuels can be
categorized into gaseous fuels or liquid fuels and are The aforementioned equation shows two important
discussed extensively in the fuels chapter; however in chemical features of complete combustion: no carbon
this section, the chemical reactions for gaseous hydro- monoxide (CO) and some unreacted oxygen appear in
carbons will be developed7–9 (see Table 3.2). the combustion products. To account for any hydrocarbon
The following is an introduction of a redox example fuel, variables A and B are introduced as moles of carbon
of methane (CH4) combustion. Assume dry air is com- and moles of hydrogen, respectively. Equation 3.13 gives a
posed of 21% oxygen and 79% nitrogen, thus yield- generic equation for gaseous hydrocarbon fuels with air:
ing 21 mol O2 and 79 mol N2. To simplify this fraction, B

divide the number of moles of nitrogen with oxygen: C AHB + ε  A +  (O 2 + 3.76N 2 )
 4
79 mol N 2
= 3.76 B  B  B
21 mol O 2 → ACO 2 + H 2O + (ε − 1)  A +  O 2 + ε  A +  3.76N 2
2  4  4
showing 3.76 mol of N2 for every 1 mol of O2 (3.13)
where A and B are defined as
CH 4 + O 2 + 3.76N 2 → CO 2 + H 2O + 3.76N 2
C AHB
the unbalanced equation of methane combustion
Further, for gases, if nitrogen and oxygen are intro-
CH 4 + 2O 2 + ( 2 ) 3.76N 2 → CO 2 + 2H 2O + ( 2 ) 3.76N 2 duced, Equation 3.14 may be generalized:

 B
Balancing the equation of methane combustion C AHBN C O D + ε  A +  ( O 2 + 3.76N 2 )
The products of chemical combustion reactions are  4
called flue gas, as they are the gaseous exhaust that exits B  B D
→ ACO 2 + H 2O + (ε − 1)  A + +  O 2
through the chimney, or flue. The aforementioned equa- 2  4 2
tion is theoretical in that it presumes that all the oxygen B C

and fuel react and that nitrogen does not. Actually, trace + ε  A + +  3.76N 2 (3.14)

 4 2
amounts of nitrogen will react with oxygen to form
nitrogen oxides (NOx). Further, in industrial practice,
perfect mixing cannot be achieved. It is actually more 3.6.2 Flue Gas
cost-effective to ensure complete combustion with the In situ analyzers measure the flue gas species in the
addition of air above the theoretical requirements. actual hot wet environment. In contrast, extractive ana-
lyzers remove the flue gas, condense the water, and mea-
TABLE 3.2 sure the concentration of the flue gas species in the dry
gas. Therefore, two sets of equations are needed for wet
Common Gaseous Fuels
and dry measurements extracted from Equation 3.13.
CH4 Methane Total wet products
C2H6 Ethane
 B  B B
C3H8 Propane
TWP = A + (ε − 1)  A +  + ε  A +  3.76 + (3.15)
C4H10 Butane  4  4 2
Total dry products Example 3.3

For the combustion of methane (CH4), find the air-
 B  B
TDP = A + (ε − 1)  A +  + ε  A +  3.76 (3.16) to-fuel ratio by volume and weight as well as flue
 4  4 gas-to-fuel by weight when there is 20% excess air.
Using the general equation (3.13) for gaseous
 B  B fuels results are as follows:
(ε − 1)  A +  (ε − 1)  A + 
 4  4  (3.17)
f O2 ,wet = f O2 ,dry =  B
TWP TDP C AHB + ε  A +  ( O2 + 3.76 N 2 )
 4
A A B B
f CO2 , wet = f CO2 ,dry = (3.18) → ACO 2 +

H 2 O + ( ε − 1)  A +
TWP TDP   O2
2 4
 B  B  B
εA+  εA+  + ε A +  3.76 N 2
 4
 4  4
f N2 ,wet = f N2 ,dry = (3.19)
TWP TDP
The hydrocarbon CH4 and 20% excess air yields
B 1
f H2 O,wet = (3.20) ε = 1 + 20% = 1.20
2 TWP
Now solve using the equations presented earlier.
where Air-fuel volume
f is the mole or volume fraction of the subscripted species
0 < f < 1 and the subscripts wet or dry refer to in situ or  A  B
extractive measurements, respectively   = 4.76ε  A +  = 11.424
F v 4

Because of the strong relationship between oxygen and
The specific gravity of methane is 0.554.
excess air, the excess oxygen can be used as a measure
Solving for air-fuel weight yields
of excess air (see Figure 3.5a through f).
 A  A 1 lbmair
  =   × = 20.62
F w F v SGfuel lbmfuel

3.7 Air-to-Fuel Ratio Lastly, the flue gas-to-fuel ratio can be found:
The air-to-fuel ratio is an important parameter that will
 FG   A lbm FG
dictate many later combustion calculations (see Table 3.3).   =   + 1 = 21.62
The air-to-fuel ratio will be found by both volume and F w  F w lbmfuel

weight. The general form of air-to-fuel ratios will be pre-
sented and then an example calculation will be provided. Further the molecular weight of the flue gas in
Air-to-fuel ratio by volume can be found in the general general form, can be determined by
form of hydrocarbon gaseous fuels, Equation 3.13, through
MWFG
 A  B
  = 4.76ε  A +  (3.21) B B B
F V 4
2
(18) + ( ε − 1)  A + 4  32 + ε  A + 4  3.76 ( 28)
A 44 +
=
B  B  B
where CAHB is the fuel. A + + ( ε − 1)  A +  + ε  A +  3.76
To find air-fuel weight, use the molecular weight: 2  4  4
(3.24)
 A  A MWair  A  1
  =   × =  × (3.22)
F w F v MWfuel  F v SGfuel

The molecular weight is introduced as a ratio of
the actual weight of the fuel:
Flue gas-fuel weight can also be calculated in general
form using
MWfuel = 12A + B
 FG   A  Fuel   A
  =   +   =   + 1 (3.23) It is important to note that these equations assume
F w F w Fuel w F w
air is composed of 21% O2 and 79% N2.
20% 90%
18%
Dry N2
85%
16%
H2O
14%
80%
Vol.% species (except N2)
12%
CO2 Wet N2
Vol.% N2
10% 75%
8%
70%
6%
O2
4%
65%
2%
Vol.% dry species > Vol.% wet species in all cases
0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(a) Percent excess air
20% 90%
18%
Dry N2
85%
16%
H 2O
14%
80%
12%
CO2 Wet N2 Vol.% N2
10% 75%
8%
70%
6%
O2
4%
65%
2%
0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(b) Percent excess air
FIGURE 3.5
Species concentration versus excess air for the following fuels: (a) CH4, (b) natural gas. (From Baukal, C.E. (ed.), The John Zink Combustion
Handbook, CRC Press, Boca Raton, FL, 2001.)
20% 90%
18%
Dry N2
85%
16%
H2O
14%
80%
12%
CO2 Wet N2
Vol.% N2
10% 75%
8%
70%
6%
O2
4%
65%
2%
0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(c) Percent excess air
20% 90%
18%
Dry N2
85%
16%
H2O
14%
80%
CO2
12%
Wet N2
Vol.% N2
10% 75%
8%
70%
6%
O2
4%
65%
2%

0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(d) Percent excess air
FIGURE 3.5 (continued)

Species concentration versus excess air for the following fuels: (c) simulated refinery gas (25% H2, 50% CH4, 25% C3H8), (d) C3H8. (From Baukal, C.E.
(ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
20% 90%
18%
85%
16%
Dry N2
14%
CO2 80%
12%
Wet N2
Vol.% N2
10% 75%
H2O
8%
O2 70%
6%
4%
65%
2%
0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(e) Percent excess air
20% 90%
18%
85%
16%
Dry N2
14%
CO2 80%
12%
Wet N2 Vol.% N2
10% 75%
H2O
8%
70%
6%
O2
4%
65%
2%
0% 60%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
(f ) Percent excess air

Species concentration versus excess air for the following fuels: (e) No. 2 oil, and (f) No. 6 oil. (From Baukal, C.E. (ed.), The John Zink Combustion
TABLE 3.3
Combustion Data for Hydrocarbons
Higher
Heating Theor. Adiabatic
Value Air/Fuel Flame Ignition Flammability
(Vapor), Ratio, Max Flame Temp Temp Flash Limits (in Air)
−1
Hydrocarbon Formula Btu lb m by Mass Speed (ft s−1) (in Air) (°F) (in Air) (°F) Point (°F) (% by Volume)
Paraffins or alkanes
Methane CH4 23,875 17.195 1.1 3484 1301 Gas 5.0 15.0
Ethane C2H6 22,323 15.899 1.3 3540 968–1166 Gas 3.0 12.5
Propane C3H8 21,669 15.246 1.3 3573 871 Gas 2.1 10.1
n-Butane C4H10 21,321 14.984 1.2 3583 761 −76 1.86 8.41
iso-Butane C4H10 21,271 14.984 1.2 3583 864 −117 1.80 8.44
n-Pentane C5H12 21,095 15.323 1.3 4050 588 <−40 1.40 7.80
iso-Pentane C5H12 21,047 15.323 1.2 4055 788 <−60 1.32 9.16
Neopentane C5H12 20,978 15.323 1.1 4060 842 Gas 1.38 7.22
n-Hexane C6H14 20,966 15.238 1.3 4030 478 −7 1.25 7.0
Neohexane C6H14 20,931 15.238 1.2 4055 797 −54 1.19 7.58
n-Heptane C7H16 20,854 15.141 1.3 3985 433 25 1.00 6.00
Triptane C7H16 20,824 15.151 1.2 4035 849 — 1.08 6.69
n-Octane C8H18 20,796 15.093 — — 428 56 0.95 3.20
iso-Octane C8H18 20,770 15.093 1.1 — 837 10 0.79 5.94
Olefins or alkenes
Ethylene C2H4 21,636 14.807 2.2 4250 914 Gas 2.75 28.6
Propylene C3H6 21,048 14.807 1.4 4090 856 Gas 2.00 11.1
Butylene C4H8 20,854 14.807 1.4 4030 829 Gas 1.98 9.65
iso-Butene C4H8 20,737 14.807 1.2 869 Gas 1.8 9.0
n-Pentene C5H10 20,720 14.807 1.4 4165 569 1.65 7.70
Aromatics
Benzene C6H6 18,184 13.297 1.3 4110 1044 12 1.35 6.65
Toluene C7H8 18,501 13.503 1.2 4050 997 40 1.27 6.75
p-Xylene C8H10 18,663 13.663 4010 867 63 1.00 6.00
Other hydrocarbons
Acetylene C2H2 21,502 13.297 4.6 4770 763–824 Gas 2.50 81
Naphthalene C10H8 17,303 12.932 4100 959 174 0.90 5.9
Sources: Gray, D.E. (ed.), American Institute of Physics Handbook, 2nd edn., McGraw-Hill Book Company, New York, 1963; Perry, R.H. et al.
(eds.), Chemical Engineers’ Handbook, 4th edn., McGraw-Hill Book Company, New York, 1963; Weast, R.C. (ed.), Handbook of Chemistry
and Physics, 53rd edn., The Chemical Rubber Company, Cleveland, OH, 1972; gives the heat of combustion of 500 organic com-
pounds; Steere, N.V. (ed.), Handbook of Laboratory Safety, 2nd edn., The Chemical Rubber Company, Cleveland, OH, 1971; Physical
Measurements in Gas Dynamics and Combustion, Princeton University Press, 1954.
3.7.1 Air-to-Fuel Mixture Ratio Estimate the hydrocarbon formula by calculating

the moles of carbon and hydrogen:
Gases are not always isotropic; fuels often contain a
mixture of gaseous compounds. The following is a sam-
mole C = 50%(1) + 30%(2) = 1.1 = A
ple solution for a given gas mixture.
Example 3.4 mole H = 50%( 4) + 30%(6) + 20%(2) = 4.2 = B
Find the air-to-fuel ratio by weight of a gas mix- C AHB = C1.1H 4.2
ture composed of
50% CH 4 The air-fuel by volume Equation 3.21 yields
30% C 2H6 20% Excess air
 A  4.2 
  = 4.76(1.2)  1.1 +  = 12.28
20% H 2 F v 4 

Further, the molecular weight of the fuel can be 0.1083 mol O 2 1 mol Air 28.85 g Air
found as follows: × ×
g Fuel 0.21 mol O 2 1 mol Air
14.87 g Air
MWfuel = 12 A + B = 12(1.1) + 4.2 = 17.4 =
1 g Fuel

17.4
Specific gravity (SGfuel ) = = 0.6
28.842
 A
  = 14.87
F WT
 A 12.28
  = = 20.4
F WT 0.6

Many solid and liquid fuels cannot be defined
with CAHB. In such cases, use the general equa-
3.7.2 Air-to-Fuel Mass Ratio tion for hydrocarbon combustion along with
the fuel’s mole ratio as defined in the following
Solid and liquid fuel compositions are often given on a example.
mass basis. The following is an example of calculating
the air-to-fuel ratio when given the mass of C and H2.
Example 3.6
Example 3.5 Given the same fuel as Example 3.4 with a mass
composed of 85% carbon and 15% hydrogen,
Mass g – Mole/g – Fuel Mole O2 ϕ = 1 find the air-fuel ratio by weight and the chemi-
cal equation with no excess air using the new
85% C 0.0708 0.0708
mole ratio:
15% H2 0.075 0.0375
0.1083
 H
  =x
C mole
Convert to mole/gram fuel:
0.85 gC gmol g × mol C

× = 0.0708  x
g Fuel 12 gC g Fuel CH x + ε  1 +  ( O 2 + 3.76N 2 )
 4
x  x  x
0.15 gH 2 g × mol g × mol H 2 → CO 2 + H 2 O + ( ε − 1)  1 +  O 2 + ε  1 +  3.76N 2
× = 0.075 2  4  4
g Fuel 2 g H2 g Fuel

Because there is no excess air
Developing a relation between weight of C and
moles of O2
ε=1
C + O 2 → CO 2
Convert mass percentage to mole fraction:
0.0708 mol C 1 mol O 2 0.0708 mol O 2

× =
g Fuel 1 mol C g Fuel 0.15 gH 12 gC 1 mol H mol H
× × = 2.11 =x
0.85 gC 1 mol C 1 gH mol C

Developing a relation between weight of H2 and
moles of O2
Finding the molecular weight of the fuel
1
H2 + O 2 → H2O
2 MWfuel = 12 + x = 14.11

1
0.075 mol H 2 2 mol O 2 0.0375 mol O 2  A  x  4.76 mol Air mol Fuel
× = = ε1 +  × ×
g Fuel 1 molH 2 g Fuel  
F WT  4 mol Fuel (12 + x ) gFuel
28.85 gAir
Combining the two relationships the required × = 14.87
mol Air
amount of O2 per mass of fuel is found:
This yields the same answer as Example 3.5: TEG

Note that one cannot find  
A
with this form Normalize
 F  Vol
12% O2 = 0.12 mol O2 1 mol O2
of the oxidation equation. However, the volume of
4% H2O 0.04 H2O 0.33 mol H2O
flue gas products, molecular weight, O2, etc. can
5% CO2 0.05 CO2 0.416 mol CO2
be found.
79% N2 0.79 N2 6.58 mol N2
Applying the multiplier with each molecule

3.7.3 Turbine Exhaust Gas normalizes the quantities and provides the stoi-
chiometric coefficients as shown in the following:
In turbine exhaust gases (TEG), the products are deliv-
ered to the burner at elevated temperatures, (1200°F or  B
650°C), which in turn reduces the oxygen content in air C AHB + ε  A +  ( O 2 + 0.33H 2 O + 0.416CO 2 + 6.58N 2 )
 4
to roughly 12%. Thus, in cases like this as well as other
oxidizer streams, O2, with N2, CO2, and H2O must be B  B
→ ACO 2 + H 2 O + ( ε − 1)  A +  O 2
normalized. This method is best illustrated through an 2  4
example.
 B
+ ε  A +  (0.33H 2 O + 0.416CO 2 + 6.58N 2 )
 4

Example 3.7
ε = 1 + 300% = 4
Given a mole percentage composition of a turbine
exhaust gas, find the turbine exhaust gas to fuel
Solving for TEG by weight
by weight and the percentage of wet oxygen of
CO2 with an excess air of 300%.
 TEG   B
  = ε  A +  (1 + 0.33 + 0.416 + 6.58 )
12% O2 F V 4
4% H2O
5% CO2  1
79% N2 = 4  1 +  (8.326 ) = 66.6
 1
The general equation is
MWTEG = 0.12(32) + 0.04(18) + 0.05( 44) + 0.79(28)
= 28.88
 B  
C AHB + ε  A +   O 2 +3.
76N 2 
 4   air   TEG   TEG   MWTEG 
  = 
F  W  F  V  MWF 
B  B  B
→ ACO 2 + H 2 O + ( ε − 1)  A +  O 2 + ε  A +  3.76N 2
2  4  4 28.88
= 66.6 × = 120.2
16
The products can be rewritten as
Note that specific gravity should be used in this
 B
C AHB + ε  A +  ( O 2 + 3.76N 2 ) case because TEG ≠ air. Manipulating the general
 4 equation for chemical combustion for O2 yields
B  B    B
→ ACO 2 + H2O + ε  A +   O
2 +3
76N 2  −  A +  O 2
. % O 2 wet
2  4  air  4
 
used O 2
B
( ε − 1)  A + 4 
As developed in Equation 3.13 from air with = × 100
B  B  B
21% O2 and 79% N2, a similar process is done A+ + ( ε − 1)  A +  + ε  A +  (0.33 + 0.416 + 6.58 )
using TEG by normalizing the TEG stream to a 2  4   4
mole of O2 and the balance of N2, H2O, and CO2
follows. For example, 3 ( 2) 600
= =
 1  1 + 2 + 3 ( 2) + 4 ( 2)(8.326 ) 75.61
12% O 2 → 0.12 mol O 2  = 1 mol O 2
 0.12  = 7.93% wet
 1 
Thus, in this case, the multiplier is  = 8.33 TEG examples are typically given in lbm/h or kg/h
 0.12 
basis and are sometimes best solved with a table.
Example 3.8 klbm h lbm

MWFG = 1003.35 × = 28.24
Given the combustion of methane yielding h 35.53 klbmol mol FG

106 lbm/h TEG at 1000°F with a fuel input of 80 ×
106 Btu/h HHV. Find the molecular weight of TEG,  TEG  10 6
the emission of O2, H2O, and CO2 in lbm/h, TEG to   = = 298.5
F WT 3350
fuel by weight %O2 and ɛ.
The given TEG flow breakdown
3.33
% O 2 wet = × 100 = 9.37%
lbm/h klb mol/h 35.53
lbm/h TEG lbm/h Products TEG
120 k O2 − 13,400 = 106.6 k O2 3.75 Finding ε
40 k H2O + 7,537.5 = 47.54 k 2.22
H2O 10 6 lbTEG
For ε = 1 → 13, 400 lbm O 2 ×
50 k CO2 + 9,212 = 59.21 k CO2 1.136 120 klb O 2
790 k N2 — = 790 k N2 28.2
= 111.6 klb TEG
106 lbm/h 1003.35 k 35.306
10 6 10 6
MWTEG = = 28.32 ε= = 8.96
35, 306 111.6 K

Example 3.9
The combustion reaction consumes O2 and
forms H2O and CO2: The same example given earlier can be solved using
the general formula of hydrocarbon combustion.
CH 4 + 2O 2 → CO 2 + 2H 2 O TEG pph klb mol/h Normalized O2 = 1

120 k O2 3.75 1
Btu lbm CH 4 3, 350 lbm CH 4 40 k H2O 2.22 0.592
80 × 10 6
× =
h 23, 875 Btu h 50 k CO2 1.136 0.303

790 k N2 28.2 7.52
lbm CH 4 lbmol CH 4 2 lb molO 2 32 lbm O 2

− O 2: 3, 350 × × ×  B
h 16 lbm CH 4 lbmol CH 4 lbmol O 2 C AHB + ε  A +  ( O 2 + 0.592H 2 O + 0.303CO 2 + 7.52N 2 )
 4
lbm O 2
= 13, 400
h B  B
→ ACO 2 + H 2 O + ( ε − 1)  A +  O 2
2  4
lbm CH 4 lbmol CH 4 2 lbmol H 2 O  B
+ H 2 O: 3350 × × + ε  A +  (0.592H 2 O + 0.303CO 2 + 7.52N 2 )
h 16 lbm CH 4 lbmol CH 4  4

18 lbm H 2 O lb H O
× = 7537 m 2 MW TEG = 28.32
lbmol H 2 O h

 4
ε  1 +  (1 + 0.592 + 0.303 + 7.52) 28.32
lbm CH 4 lbmol CH 4 1 lbmol CO 2  TEG   4
CO 2: 3350 × ×   =
h 16 lbm CH 4 1 lbmolCH 4 F  WT 16

44 lbm CO 2 lb CO 2 TEG = 10 6 lb/h
× = 9212 m
lbmol CO 2 h
Fuel = 3350 lb/h

.  4
m FG k pph klbm/h
10 6
ε 1 +  (9.415) 28.32
 4
O2 106.6 3.33 = → ε = 8.96
3350 16
H2O 47.54 2.64
CO2 59.21 1.345
The method used in Example 3.8 yields the
N2 790 28.21
same results as Example 3.9. Either is correct and
1003.35 35.53

it is a user preference.
industrial combustion conditions. However, H2O can

exist as either a liquid or a vapor, depending on how
3.8 Chemical Thermodynamics
much heat is extracted from the process. If so much
The amount of heating energy released in a chemical heat is extracted that the H2O condenses, then the com-
reaction can be calculated through thermodynamics. bustion yields its HHV. If water is released from the
stack as a vapor, then combustion yields the LHV. The
3.8.1 Enthalpy, Entropy, and Heat Capacity process industry usually uses the LHV. Boiler and tur-
bine calculations usually use the HHV. However, either
How hot can a flame be? First, there is a difference measure can be used in combustion calculations as
between heat (Q) and temperature (T). Heat is energy in long as one is consistent.
transit. When a body absorbs heat, it stores it as another For the purpose of combustion, the concern is with
form of energy, increasing the body’s temperature and the changes in the energy of a system as opposed to
expanding it. That is, the material uses some of the energy of a particular state; because of this, any state
thermal energy to raise the temperature and some of may be chosen as the initial state and assigned a zero
the energy to expand the body against the atmosphere. value of enthalpy for that substance. However, when
The same amount of heat absorbed in different mate- the process involves chemical change, the composi-
rials will yield different temperature increases and tion of the system is no longer the same as it was in
expansions. the beginning of the process. It is thus necessary to
For example, 100 Btu of heat will raise the temper- choose a reference state for all substances which has
ature of 1 lb of water by 100°F and expand the vol- been defined by engineers to be at 77°F (25°C) and 1
ume of that water by approximately 2.2%. The same atm (101.3 kPa), known as the standard reference state.
100 Btu of heat absorbed by 1 lb of air will increase Heat of formation is the enthalpy of a substance at the
the temperature by 400°F and expand the volume by standard reference state.10
approximately 62%. The total energy used to raise the By definition, these heating values can be related by
temperature and increase the volume is called enthalpy
(H), classically defined as the measure of total energy
mH2 O
content of a substance in a thermodynamic system.10 HHV − LHV = ( hvap ) (3.26)
Enthalpy relates to temperature by a quantity known mfuel
as the isobaric heat capacity, Cp. Heat capacity is the
amount of heat required to raise the temperature of 1 where
unit mass of a substance by one degree of temperature. mH2 O is the mass of H2O
The change in internal energy or enthalpy for an ideal mfuel is the mass of fuel
gas during a process can be determined by integrating hvap is the enthalpy of vaporization of H2O at the speci-
Equation 3.25: fied temperature
∫
h2 − h1 = Cp (T ) dT
1
(3.25) Example 3.10
Here, an example of calculations for enthalpy of
combustion, HHV, and LHV of methane gas is
From this equation, it is important to note that Cp varies presented:
as a function of temperature.
CH 4g + 2O 2g → CO 2g + 2H 2 O liq + h

3.8.2 Heat of Combustion
In addition to the conservation of mass, energy is also Heat of formation at standard reference state
conserved in a combustion reaction. One measure
of the chemical energy of a fuel is the heat of combus-
CH 4 g = −17.88 kcal/gmol
tion. Table 3.3 gives heats of combustion for some typi-
cal fuels on a HHV mass basis.11 Heat of combustion
is reported as either net heating value (lower heating CO 2g = −94.051 kcal/gmol

value, LHV) or gross heating value (higher heating
value, HHV). To understand the difference, reconsider H 2 O liq = −68.315 kcal/gmol
the chemical equation for methane combustion. When
methane burns, it produces two products: CO2 and
H 2 O g = −57.796 kcal/gmol
H2O. The CO2 will remain a gas under all conceivable
For H 2 O liq To find enthalpy at vaporization

mH2O
HHV − LHV = ( hvap )
mfuel

∆h = ∑ (products) − ∑ (reactants)
Btu
hvap = 970
∆h = (1 × −94.051) + ( 2 × −68.315) − [ −17.88 ] lbm

kcal
= −212.8 CH 4 + 2 O2 → CO 2 + 2H 2 O
gmol

Negative enthalpy denotes exothermic reactions: mol H 2 O lbm 1 mol CH 4

2 × 18 ×
mol CH 4 mol H 2 O 16 lbm
kcal mH2O
∴ CH 4 g + 2O 2g → CO 2g + 2H 2 O liq + 212 = 2.25 mass ratio =
gmol mfuel

kcal Converting
212 = HHV since H 2 O is condensed.
gmol kcal Btu
LHV = 191.76 ⇒ 21, 589
gmol lbm

Conversely, for H2Og
3.8.3 Adiabatic Flame Temperature

∆h = ∑ (products) − ∑ (reactants) An adiabatic flame is an idealized situation where there
is no heat loss to the surroundings during combustion,
∆h = (1 × −94.051) + ( 2 × −57.796 ) − [ −17.88 ] that is, to say that all generated heat is applied to the
products. Once the reactants and their states are speci-
kcal fied, the enthalpy of the reactants can easily be deter-
= −191.76
gmol mined through tables. However, the calculations of the

enthalpy of the products are not as simple because the
kcal specific heat varies as a function of temperature. To be
∴ CH 4 g + 2O 2g → CO 2g + 2H 2 O g + 191.76
gmol able to predict the enthalpy of products, engineers have
created tables that provide the enthalpy of combustion
products given varying temperatures of adiabatic flame.
kcal
191.76 = LHV since H 2 O is not condensed Through these tables, engineers can very closely pre-
gmol
dict the adiabatic flame temperature by interpolating
enthalpy solutions.10
Btu
Conversion to
lbm
Example 3.11
kcal C J Btu The following example will find adiabatic flame
212 × 1, 000 4.186 × 9.47 × 10 −4 temperature through the enthalpy solutions
gmol kcal C J
using standard gas tables.
1 gmol g Btu
× × 454 = 23,868 Given
16 g lb m lbm

CH4 at 20% EA
Finding density of methane
Find adiabatic flame temperature.
lb 16 lb From the general form,
ρCH4 = .076 m3 × = .042 m3
Sft 28.85 Sft
 B
C AHB + ε  A +  ( O 2 + 3.76N 2 )
Now one can find Btu per unit volume:  4
Btu lb Btu B  B  B
23, 868 × .042 m3 = 1006 → ACO 2 + H 2 O + ( ε − 1)  A +  O 2 + ε  A +  3.76N 2
lbm Sft Sft 3 2  4  4
The products are as follows: Next interpolate again for enthalpy from interpolated
tempAD:
From the Gas Tables CO2 H2O
ACO2 = 1 lb mol CO2 h (70°F) = 3968.3 Btu/lb 3240°F 44,860 3240°F 36,765
mol 3247°F 44,961 3247°F 36,851
B/2H2O = 2 lb mol H2O h (70°F) = 4202.2 3260°F 45,151 3260°F 37,011
B O2 N2
(ε − 1)  A + 4  O 2 = 0.4 lb mol O2 h (70°F) = 3676.1
3240°F 30,077.5 3240°F 28,735.1
3247°F 30,140 3247°F 29,145
 B
ε  A +  3.76N 2 = 9.04 lb mol N2 h (70°F) = 3680.8 3260°F 30,258 3260°F 29,907.5
 4
Interpolating once more for tempAD from the calculated

By definition LHV.
Guess
Heating value TempAD LHV
∑ Moles of product × (∑ h ∑h ) 3200°F 338,242

= products − reactants X 344,000

3247°F 347,312
In practical application, LHV is used when the H2O  344, 000 − 338, 242 
product of the reaction exists in vapor phase, while  347 , 312 − 338, 242  47 + 3, 200 = 3, 231°F
HHV is used when the H2O product is in a liquid
phase. This provides a fairly good approximation.
In most cases, the H2O product is in gaseous phase, Note that the actual flame temperature will be much
and from the previous example LHV in the combustion cooler than this, because heat will transfer from the flame
of methane was calculated and now the units of lb mole to the surroundings via convection and radiation. This
need to be converted: example was used to show the hand calculation of adia-
batic flame temperatures. In practice, solutions would be
Btu lbm kBtu generated using a curve fit of enthalpy versus tempera-
LHV = 21, 589 × 16 = 344 ture. This can be done using a simple computer solution.
lbm lb mol lb mol

3.8.4 Dissociation
Now, set up the heating value equation: At elevated temperatures, bonds that hold gas mole-
cules together weaken. These gas molecules are said to
344, 000 = 1 ( hCO2 − 3, 968.3 ) + 2 ( hH2 O − 4, 202.2) dissociate and form new molecules from their respec-
tive constituents. Dissociation is a reversible process
+ 0.4 ( hO2 − 3, 676.1) + 9.04 ( hN2 − 3, 680.8 ) and when undergoing dissociation, gases are heteroge-
neous in composition; normally only 1%–2% of a given
gas dissociates under typical combustion temperatures:
Looking up enthalpy from temperature in gas tables,
one can then calculate the LHV solutions tabulated 2CO 2 ←
→ 2CO + O 2
next:
Dissociation of carbon dioxide gas
Calculated 2H 2O ←
→ H 3 O1+ + OH1−
TempAD CO2 H2O O2 hO2 N2 hN2 LHV Result
3,500°F 48,646.80 39,988 32,440 30,982 338,242 Too low Dissociation of water
3,200°F 44,280 36,274 29,715 28,390 374,557 Too high
The measure of a gas’s stability against dissociation is
represented by the equilibrium constant, Kp; the smaller
Interpolating temperature from enthalpy yields the value of the constant, the more stable the gas.
Given the chemical equation
temp AD = 3247 ⇒ next guess aA + bB ↔ cC + dD

c d
Kp =
( pC ) ( pD ) (3.27)
a b 3.9 Practical Liquid Fuels
( pA ) ( pB )
As mentioned earlier, the topic of fuels will be exten-
sively covered in other chapters; the scope of this section
where is only to introduce the chemical properties of liquid
pA is the partial pressure of element A fuels. A parameter that can predict many properties of
pB is the partial pressure of element B a liquid fuel is API gravity. API gravity was devised by
pC is the partial pressure of element C the American Petroleum Institute as a standard mea-
pD is the partial pressure of element D sure of a fuel’s density and is defined as
p is the mixture pressure
Degree API gravity = 141.5
− 131.5
Specific gravity @ 60°F
The process of finding the equilibrium constant
(3.28)
through this equation can become very time-consum-
ing and involved when calculating for multiple partial The API gravity of liquid fuels holds many relation-
pressures. Fortunately, the equilibrium constant is a ships with several of the fuel’s properties. Some rela-
property that can be looked up in tables such as the tionships between properties and API gravity for liquid
Janaf Thermochemical Tables.12 In these charts, the Kp for fuels are given in the following (see Tables 3.46 and 3.513):
many gases is tabulated as a function of temperature.
It is also important to note that the values for equilib- • The lower a fuel’s API gravity, the heavier a fuel
rium constants are commonly given in base 10, that is, is in viscosity, the higher the carbon residue,
log10 Kp. and the heavier the weight.
Dissociation is an important factor to consider, as • Conversely, the higher a fuel’s API gravity, the
high percentages of dissociation will decrease the theo- lighter a fuel is in viscosity, the lower the carbon
retically attainable temperature of a gas. residue, and the lower the weight.
TABLE 3.4
Specific Gravity and Properties of Common Liquid Fuels
Grade Fuel No. 1 No. 2 No. 4 (Light) No. 4 No. 5 (Light) No. 5 (Heavy) No. 6
Specific gravity, 60/60°F 0.8499 0.8762 0.8762 — — — —
(degree API), max 35 min 30 min 30 max — — — —
Flash point °F min 100 100 100 130 130 130 140
Kinematic viscosity mm2/s
Pour point °F max 0 20 20 20 — — —
At 100°F min 1.4 2 2 5.8 >26.4 >65 —
Max 2.2 3.6 5.8 26.4 65 194 —
At 104°F min 1.3 1.9 — 5.5 >24 >58 —
Max 2.1 3.4 — 24 58 168 —
Saybolt viscosity
Universal at 100°F
Min — 32.6 32.6 45 >125 >300 >900
Max — 37.9 45 125 300 900 9000
Furol at 122°F
Min — — — — — 23 >45
Max — — — — — 40 300
Distillation temperature, °F
10% point max 420 — — — — — —
90% point min — 540 — — — — —
Max 550 640 — — — — —
Sulfur content, mass, max 0.5 0.5 — — — — —
Corrosion copper strip, max 3 3 — — — — —
Carbon residue, 10% b; %m, max 0.15 0.35 — — — — —
Water and sediment, % vol, max 0.05 0.05 0.5 0.5 1 1 2
Sources: Adapted from Schmidt, P., Fuel Oil Manual, 4th edn., Industrial Press, New York, 1985. With permission.
TABLE 3.5
Liquid Fuel Properties by API Gravity as well as Common Coals
Heating Values (Btu/gal) Heating Values (Btu/lbm)
Specific Percent BTU Fired/Cubic Foot
API Gravity Gravity Hydrogen Higher Lower Higher Lower of Air
0.000 1.076 8.940 158,610 149,470 17,684 16,665 98.32
2.000 1.060 9.152 157,380 148,330 17,815 16,791 98.67
4.000 1.044 9.364 156,110 147,250 17,935 16,918 99.00
6.000 1.029 9.577 154,860 146,340 18,055 17,062 99.29
8.000 1.014 9.789 153,570 145,260 18,165 17,182 99.55
10.000 1.000 10.002 152,280 144,000 18,270 17,277 99.78
12.000 0.986 10.214 151,000 142,720 18,372 17,365 99.99
14.000 0.973 10.426 149,720 141,440 18,470 17,449 100.17
16.000 0.959 10.639 148,440 140,160 18,564 17,529 100.33
18.000 0.946 10.851 147,160 138,790 18,654 17,593 100.47
20.000 0.934 11.064 145,880 137,510 18,739 17,663 100.59
22.000 0.922 11.276 144,600 136,230 18,821 17,731 100.69
24.000 0.910 11.488 143,420 135,040 18,908 17,804 100.77
26.000 0.898 11.701 142,140 133,760 18,982 17,863 100.83
28.000 0.887 11.913 140,950 132,580 19,063 17,931 100.88
30.000 0.876 12.126 139,660 131,300 19,124 17,979 100.91
32.000 0.865 12.338 138,490 130,120 19,200 18,040 100.94
34.000 0.855 12.550 137,310 128,940 19,269 18,094 100.95
36.000 0.845 12.763 136,130 127,560 19,334 18,117 100.95
38.000 0.835 12.975 134,960 126,570 19,396 18,191 100.95
40.000 0.825 13.188 133,760 125,390 19,450 18,233 100.94
Coal
Pittsburg #8 bituminous — — 95,166 — — — 97.400
Anthracite — — 90,233 — — — 96.530
Source: Adapted from Stultz, S.C. and Kitto, J.B., Steam: Its Generation and Use, 40th edn., Babcock and Wilcox Steam manual, Barberton,
OH, 1992.
Note: Air referenced at 70°F.
• The higher the API gravity, the greater the HHV perform very well. For first-order oxidation, the gen-
on a mass basis created when burning the fuel. eral expression is
• The higher the API gravity, the lower the unit
weight of the fuel. d (chemical )
= − K [ O 2 ][ chemical ] (3.29)
• The higher the API gravity, the higher the dt
hydrogen and the lower the carbon content. where
• API gravity can indicate the grade of a fuel.
 −E 
• As API gravity decreases, the rate of combustion  
K = Ae  RT 
also decreases while the flame length increases
(Schmidt9). and,
A is the pre-exponential factor/frequency factor in
appropriate units
R is the universal gas constant in appropriate units
T is the absolute temperature
3.10 Combustion Kinetics
E is the activation energy, usually listed in kcal/mol
Combustion systems operate with moderate to high t is time in seconds
temperatures and, as such, exact chemical kinet-
ics is very important for emissions prediction, they For perfectly stirred reactors well downstream of the
must be correct. Generally, first-order reaction rates initial mixing, integration of the first-order equation
results in a simple equation for constant temperature The utilization of gas kinetic data for emission forma-
and O2 mole fraction in a time step may be expressed as tion can be computed from detailed temperature and
species concentration in the flow field in many ways
such as the Rayleigh flux theorem. Simply stated in all
 Chemical final 
1−  = 1 − e − k (O2 )( ∆t)  (3.30) cases, one can post process thermal map data in some
 Chemical initial   discrete volume form or insert into a CFD code (see
Chapter 9) using the Rayleigh flux theorem as follows:
For utilization, and performance prediction, kinetic
cv cs
data can be utilized from literature such as Battelle ∂
Columbus Laboratories “Chemical aspects of after-
burner systems.”14 For instance for CO destruction,
∂t∫ ∫
nρdv = nρ (V ⋅ da ) (3.35)

several kinetic data are available such as those found
in Ref. [15]. where
n is chemical in mass units
 −25 , 000  2 t is time
d[CO] P 
(CO )(O 2 )0.5 (H2O )0.5 
−
RT 
= −1.8 × 107 e   ρ is density
dt RT  v is volume
(3.31) a is area
V is velocity vector
Most published CO rates involve H2O because CO
destruction requires the (OH)−1 radical to produce the where described in words, the formation of (n) through
reaction. the volume surface is equal to the integrated rate of
For HC and VOC incineration, several sources are formation over the control volume. It is then a simple
available, such as Ref. [16] where, in general extrapolation to extend this concept for even coarse vol-
umes as follows:
 12 , 200 
d(C aHb ) mol dn
∑ dt ρ∆v = nρ (V ⋅ a )
 
= −5.52 × 108 P −0.815Te  T 
(C aHb )0.5 (O 2 ) (3.36)
dt cm 2s
(3.32)
This method can be very useful for fully mixed down-
stream products even with coarse volumes. But one
3.10.1 Thermal NOx Formation must be careful with coarse volumes to be sure that the
temperature and concentrations are uniform.
Thermal NOx is formed at high temperatures when
molecular N2 and O2 dissociate and react to form NO.
For formation, rather than destruction such as NOx,
3.10.2 Prompt NOx Formation
the equations are similar such as the formation of
thermal NOx: Another NOx formation mechanism is prompt NOx.
This occurs at the flame front and is responsible for no
d( NO) − 
 E  more than 20 ppm NOx in refinery or natural-gas fueled
= 2 Ae  RT  (O)eq (N 2 ) (3.33) equipment. The mechanism can be summarized as
dt
and CH x + N 2 → HCN ↔ CN (not balanced)

K0
(O)eq = (O 2 )0eq.5 (3.34) HCN ↔ CN+O 2 → NO + CO + H (not balanced)
(RT )0.5
Both of these reactions are very fast and do not
One generally accepted practice to compute NO require high temperature. It would appear that one way
using Equations 3.33 and 3.34 is to assume O2 in to reduce NOx from the prompt mechanism would be
equilibrium with O and O2 concentration using the to dilute the HCN and CN species on the fuel side of
Westenburg results for K0 (see Ref. [17]) for O2 equilib- the combustion zone, or reduce the available oxygen.
rium and Zeldovich constants, A and E as measured The use of premix combustion is very effective for
by Bowman.18 prompt NOx reduction.
3.10.3 Fuel-Bound NOx TABLE 3.6
The fuel-bound NOx mechanism is similar to prompt Adiabatic Flame Temperatures

NOx and proceeds through the same HCN-CN chemis- Air
try. However, the fuel-bound mechanism differs in the Fuel °F °C
following ways:
H2 3807 2097
CH4 3542 1950
1. The fuel-bound mechanism requires nitrogen
C2H2 4104 2262
as part of the fuel molecule.
C2H4 3790 2088
2. At low fuel-nitrogen concentrations, all of the C2H6 3607 1986
bound nitrogen converts to NOx. C3H6 4725 2061
3. The fuel-bound mechanism can be responsible C3H8 3610 1988
for hundreds of ppm NOx, depending on the C4H10 3583 1973
amount of nitrogen bound in the fuel. CO 3826 2108
The first steps in the chemistry differ in that the inter-

mediates are formed directly from pyrolysis of the par- where slightly more O2 is needed to fully combust all
ent molecule. Ambient nitrogen is unimportant: the fuel. Nearly all industrial combustion applica-
tions are run at fuel-lean conditions to ensure that CO
CH x N y → HCN ↔ CN (not balanced) emissions are low. Therefore, depending on the actual

burner design, the flame temperature may be close to
The subsequent chemistry (oxidation pathways for its peak, a condition that is often desirable for maximiz-
HCN and CN) is identical to prompt NOx. ing heat transfer. One problem often encountered by
Reducing the available oxygen, reducing the nitrogen maximizing the flame temperature is that high flame
content in the fuel, or diluting the fuel species with an temperature maximizes NOx emissions. NOx increases
inert gas reduces NOx. approximately exponentially with gas temperature.
This has led to many design concepts for reducing the
peak flame temperature to minimize NOx emissions.19
Figure 3.7 shows how preheating the air in the com-
bustion of the three fuels shown dramatically increases
3.11 Flame Properties the adiabatic flame temperature. The increase is nearly
linear for the air preheat temperature range shown.
The flame temperature is a critical variable in deter- Air preheating is commonly done to both increase the
mining the heat transfer, as is shown in Chapter 7. This overall system efficiency (which will be graphically
section shows how the adiabatic flame temperature shown later) and to increase the flame temperature,
is affected by the fuel composition, the e quivalence especially for higher temperature heating and melting
ratio, and the air and fuel preheat temperatures. processes like melting metal or glass. Figure 3.8 shows
As previously mentioned, real flame temperatures are the effect of preheating the fuel on the adiabatic flame
not as high as the adiabatic flame temperature, but the temperature. Again, there is a nearly linear rise in the
trends are comparable and representative of actual flame temperature, but the magnitude of the increase
conditions. is much less than for air preheating. This is due to the
much larger mass of air compared to the mass of fuel
in the combustion process. Preheating the air to a given
3.11.1 Flame Temperature
temperature requires much more energy than preheat-
Table 3.6 shows the adiabatic flame temperature for coming the fuel to that same temperature, because of the dif-
mon hydrocarbon fuels combusted with air. Figure 3.6 ference in mass.
shows the adiabatic flame temperature as a function of Figure 3.9 shows how the flame temperature varies for
the equivalence ratio for three fuels: H2, CH4, and C3H8. fuel blends of CH4/H2 and CH4/N2. The flame tempera-
The peak temperature occurs at about stoichiometric ture increases as the H2 content in the blend increases.
conditions (ϕ = 1.0). In that case, there is just enough It is important to note that the increase is not linear; the
oxidizer to fully combust all the fuel. Any additional increase is more rapid at higher levels of H2. Because of
oxidizer absorbs sensible energy from the flame and the relatively high cost of H2 compared to CH4 and C3H8,
reduces the flame temperature. In most real flames, the it is not used in many industrial applications. However,
peak flame temperature often occurs at slightly fuel- high H2 fuels are often used in many of the hydrocarbon
lean conditions (ϕ < 1.0). This is due to imperfect mixing and petrochemical applications for fluid heating. Because
Equivalence ratio
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
3800 2100
3600 2000
3400 1900
1800
Adiabatic flame temperature (°C)

Adiabatic flame temperature (°F)
3200
1700
3000
1600
2800
1500
2600 1400
2400 1300
2200 1200
H2
2000 CH4 1100
1800 C3H8 1000
1600 900
800
1400
700
1200
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Equivalence ratio
FIGURE 3.6
Adiabatic flame temperature versus equivalence ratio for air/H2, air/CH4, and air/C3H8 flames, where the air and fuel are at ambient tempera-
ture and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Air preheat temperature (°C)
0 100 200 300 400 500 600 700 800 900 1000
4600
2500
4400
2400
4200
2300
4000 2200
3800 H2 2100
CH4
C3H8
2000
3600
1900
3400
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Air preheat temperature (°F)
FIGURE 3.7
Adiabatic flame temperature versus air preheat temperature for stoichiometric air/H2, air/CH4, and air/C3H8 flames, where the fuel is at ambi-
ent temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Fuel preheat temperature (°C)

0 100 200 300 400 500
4000 2200
3900 2150

H2
CH4 2100
3800
C3H8
2050
3700
2000
3600
1950
3500
0 100 200 300 400 500 600 700 800 900 1000
Fuel preheat temperature (°F)
FIGURE 3.8
Adiabatic flame temperature versus fuel preheat temperature for stoichiometric air/H2, air/CH4, and air/C3H8 flames, where the air is at ambi-
ent temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
4000 2200
2000
3500
1800

3000
1600
2500 1400
1200
2000
1000
1500 800
H2
N2 600
1000
400
500
200
0 0
0 10 20 30 40 50 60 70 80 90 100
H2 or N2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.9
Adiabatic flame temperature versus fuel blend (CH4/H2 and CH4/N2) composition for stoichiometric air/fuel flames, where the air and fuel are
at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
2450
4400 77°F (25°C)
2400
500°F (260°C)
4300
1000°F (538°C) 2350

Adiabatic flame temperature (°F) 1500°F (816°C)
4200
2300
4100 2250
4000 2200
3900 2150
3800 2100
2050
3700
2000
3600
1950
3500
0 10 20 30 40 50 60 70 80 90 100
H2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.10
Adiabatic flame temperature versus fuel blend (CH4/H2) composition and air preheat temperature for stoichiometric air/fuel flames, where
the fuel is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
such fuels are by-products of the chemical manufacturing Stack losses

process, their use is much less expensive than purchasing
H2 from an industrial gas supplier as well as being more 50%
cost-effective than purchasing other fuels. The graph also
Gross
shows that the adiabatic flame temperature decreases for
heating
N2/CH4 fuel blends as the N2 content increases. Again, the value
decrease is not linear and rapidly decelerates at higher of fuel 40% Load
N2 contents until no flame is present for a “fuel” having
100% N2. Figure 3.10 shows how preheating the combus-
10%
tion air for fuel blends of H2 and CH4 increases the adia-
batic flame temperature. However, the increase is not a Furnace losses
dramatic rise from pure CH4 to pure H2. Again, the change
in flame temperature with blend composition is nonlinear. FIGURE 3.11
Sample Sankey diagram showing distribution of energy in a
combustion system. (From Baukal, C.E. (ed.), The John Zink Combustion
3.11.2 Available Heat Handbook, CRC Press, Boca Raton, FL, 2001.)
The available heat in a process is defined as the gross
heating value of the fuel minus the energy carried sometimes graphically depicted using a Sankey dia-
out of the exhaust stack by the flue gases. This dif- gram. Figure 3.11 presents a very simplified Sankey
ference is the energy that is available to do work. diagram showing that only 40% of the energy goes to
However, some of that energy will be lost by conduc- the load in that example. The available heat for that
tion through the heater walls, by radiation through example is 50%, which includes the 40% to the load
openings, by air infiltration that will absorb sensible and the 10% lost to various sources. Figure 3.12 shows
energy, as well as by other types of energy losses the calculated available heat for three different fuels
that are dependent on the burner and heater designs as a function of the exhaust or flue gas temperature. As
and by the process operations. The accounting of the expected, there is a rapid decrease in available heat as
distribution for where energy goes in a process is the exhaust gas temperature increases. This indicates
Gas temperature (°C)

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
90 90
80 80
70 70
60 60
Available heat (%)
Available heat (%)

50 50
40 40
30 H2 30
CH4
20 C3H8 20
10 10
0 0
0 400 800 1200 1600 2000 2400 2800 3200 3600 4000
Gas temperature (°F)
FIGURE 3.12
Available heat versus gas temperature for stoichiometric air/H2, air/CH4, and air/C3H8 flames, where the air and fuel are at ambient tempera-
ture and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
that more and more energy is being carried out of the consumed. The horizontal axis shows the progress
exhaust instead of being transferred to the load as the of the reaction. At the upper left, the diagram shows
exhaust temperature increases. At the adiabatic flame that the fuel/air mixture has a high potential energy.
temperature for each fuel, there is no available heat as At the lower right, it is noted that the products of
all the energy was carried out in the exhaust. Figure combustion have relatively little remaining chemical
3.13 shows that the available heat increases with the energy. Because energy must be conserved, the dif-
air preheat temperature, which simply indicates that ference between the upper and lower energy levels
energy was recovered in the process and was used must be the amount of heat that the combustion reac-
to preheat the combustion air. Figure 3.14 shows that tion liberates. Note, however, that the energy diagram
preheating the fuel increases the efficiency, but to a does not slope monotonically along the reaction coor-
much lesser extent than air preheating. The mass of dinate, but contains a hump. This hump is the mini-
air is much greater than the mass of fuel, so preheat- mum ignition energy.
ing the fuel is less effective than preheating the air if What the diagram says is that fuel and air com-
the preheat temperature is the same. prising a very high chemical energy may exist in a
metastable state, until one introduces a spark or flame
3.11.3 Minimum Ignition Energy of sufficient energy. Once the system reaches the
minimum ignition energy, the reaction will be self-
Ignition energy graphs usually have the vertical axis
sustaining until the reaction consumes enough of the
as the relative energy of the fuel mixture (see Figure
reactants. At that point, the reaction cannot liberate
3.15). The reactants start from an initial state. If the
enough heat to supply the minimum ignition energy
minimum ignition energy is supplied, the reactant
and the flame goes out.
bonds will rupture, producing intermediate species
such as CH3, H, O, etc. Such species are extremely
3.11.4 Flammability Limits
reactive and recombine to form the final products, CO2
and H2O. Since the net heat release is greater than the Suppose that fuel and air are not provided in stoi-
minimum ignition energy, the reaction, once started, chiometric proportions, but have a great excess of
will continue until virtually all of the reactants are fuel or air. Will the flame continue to propagate if the

0 100 200 300 400 500 600 700 800 900 1000
90 90
85 85
H2
80 CH4 80
C3H8
75 75
Available heat (%)
Available heat (%)

70 70
65 65
60 60
55 55
50 50
45 45
0 200 400 600 800 1000 1200 1400 1600 1800 2000
FIGURE 3.13
Available heat versus air preheat temperature for stoichiometric air/H2, air/CH4, and air/C3H8 flames at an exhaust gas temperature of 2000°F
(1100°C), where the fuel is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca
Raton, FL, 2001.)

0 50 100 150 200 250 300 350 400 450 500
6 6
5 5
H2
Increase in available heat (%)
Increase in available heat (%)
4 CH4 4
C3H8
3 3
2 2
1 1
0 0
0 100 200 300 400 500 600 700 800 900 1000
FIGURE 3.14
Available heat versus fuel preheat temperature for stoichiometric air/H 2, air/CH4, and air/C3H8 flames at an exhaust gas temperature of
2000°F (1100°C), where the air is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
Press, Boca Raton, FL, 2001.)
reciprocal flammability limits weighted by their mole

fractions.
CH3, H, O, ...
Energy coordinate
Minimum ignition energy 100

CH4, O2 Limit of flammability = (3.37)
a b c ...
+ + +
Initial state A B C
Net heat release

Final state
where
CO2, H2O a,b,c… are the percentage represented by the respec-
Reaction coordinate tive gas combination in the mixture
A,B,C… are the respective flammability limits of gas
FIGURE 3.15 combination from flammability tables
Graphical representation of ignition and heat release. (From Baukal,
C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton,
FL, 2001.)
Example 3.12
ignition source is removed? That depends on whether The following will present a practical example
the fuel/air mixture has enough chemical energy to of calculating upper and lower flammability
exceed the minimum ignition energy. If not, the flame limits:
Given: A natural gas is composed of the flam-
will extinguish. This leads to a lower and upper flam-
mable gases 79% CH4, 17% C2H6, and the inert
mability limit. The limits of flammability define the gases 3% N2, 1% CO2.
limiting chemical composition in which the gas mixture Problem: Find upper and lower limits of flam-
will no longer ignite and continue to produce flames. mability of the gas mixture.
The lower limit (fuel lean) of flammability represents
the smallest ratio of fuel that, when mixed with air, can
• The gas must first be dissected into combina-
maintain a flame without the input of an external heat tions of flammable gas alone and flammable gas
source. Past the upper limit (fuel-rich), the large ratio of with inert gas. In this case there are four com-
gas begins to act as a diluent so that flames cannot be binations: CH4, CH4 + N2, C2H6, and C2H6 + CO2.
self-sustaining.4,20 A list of the upper and lower limits of • Choose a ratio of inert gas to combustible gas by
flammability for common gases is available in Table 3.3. allocating a percentage of “pure” combustible gas
For gas mixtures, one can use Le Chatelier’s rule to to the combinations of combustible and inert gas.
estimate flammability limits for gas mixtures. Because • Totaling up the percentage of dissectioned gas
this is only an estimate, one must confirm the flam- gives the percentage of mixture of each respective
mability limit of the actual mixture. Such experiments gas combination.
are relatively inexpensive and many third parties exist • With the ratio of inert to combustible gas of each
that can perform this kind of analysis. combination now known, use the flammabil-
Le Chatelier’s rule states that the flammability limit ity tables to find the upper and lower limits of
of a mixture is equal to the reciprocal of the sum of flammability.
Limits of
Ratio of Inert to Dissection% Flammability
Combustible Gas Combustible Gas CH4 CH4 N2 CO2 Total% Lower Upper
CH4 1 3 — 3 — 6 10.5 23
CH4 0 76 — — — 76 5 15
C2H6 1 — 1 — 1 2 6.5 18.5
C2H6 0 — 16 — — 16 3 12.5
Total 79 17 3 1 100
With the percentage of each respective gas combina- speed, the flame front will travel backward (called
tions and their limits of flammability, calculate the fuel’s burnback or flashback).
limits of flammability using Le Chatelier’s equation. Typical burners operate with fuel flows in excess of
the laminar flame speed. To avoid liftoff, several devices
Lower limit of flammability are used. Consider premix burners first. Fuel flows
across an orifice into the throat of a venturi. The venturi
100 100  F is designed to entrain air near the stoichiometric ratio.
= = = 4.7  
a b c d 6 76 2 16  A v Gradual flow passages are used to avoid turbulence,
+ + + + + + and hot gases are recirculated back to the burner. The
A B C D 10.5 5 6.5 3
fuel–air mixture is supplied at velocities above the lami-
nar flame speed. As the fuel jet issues from the burner,
Upper limit of flammability the velocity slows considerably. The flame front estab-
lishes where the flame and gas velocities are equal. The
100 100  F
= = = 14.9   sudden expansion from the burner avoids liftoff as the
a b c d 6 76 2 16  A v
+ + + + + + velocity rapidly slows. The high fuel–air velocity avoids
A B C D 23 15 18.5 12.5
burnback. Sudden expansions of this type are used as
flame holders because they stabilize the flame front and
The range of flammability narrows when pressures keep it from moving forward or backward.
fall below atmospheric. The increase of temperature Another concept used in premix burners is quench
tends to widen the range of flammability. The lower distance, the distance needed to remove sufficient heat
limit of flammability can be predicted at different tem- from the flame to extinguish it. Here, burner slots or
peratures using the Burgess–Wheeler law: orifices have a finite thickness that exceeds the quench
distance. Because the burner is cooler than the flame,
if the flame does begin to burnback, the heavy metal
 3.14(T − 298)  will remove sufficient heat and cool the flame below its
LT = L298 1 − (3.38)
 L298 (LHV)  minimum ignition energy. Without this feature, a flame
that finds its way into a premix burner could flashback.
where With flashback, the combustion occurs in the burner,
LT is the adjusted lower limit of flammability rather than at the flame holder. Sustained burnback will
L298 is the lower limit of flammability destroy the burner in a short time.
LHV is the lower heating value Diffusion burners supply fuel with no premix cham-
T is the temperature to be adjusted in Kelvins ber. The fuel meets the air outside the fuel nozzle. With
diffusion burners, flashback is not an issue because the
fuel alone cannot support combustion (i.e., the upper
3.11.5 Flame Speeds
flammability limit is exceeded). However, liftoff is still
The reaction between fuel and air can only occur at a a concern. If the flame lifts off the burner, it may travel
finite speed. That finite speed depends on the speed of to a place beyond the flammability limits and extin-
the reaction (chemical) and the amount of turbulence guish. Under certain conditions, the flame can repeat-
in the flame (physical). If the flame has a lot of turbu- edly liftoff and re-establish. This behavior is dangerous
lence, hot pockets of gas recirculate and the mixture because the fuel may burn incompletely during one
burns faster. To first focus on the chemical part, sup- part of the cycle and reignite later, causing an explo-
pose a long tube is filled with a flammable mixture. sion. The cycle of liftoff and burnback can occur many
If one end of the tube is ignited, the flame front will times a second, causing rumble or vibration. Such
move along the tube at a precise velocity. A flame that rumble can be a sign of dangerous instabilities.
has no turbulence is a laminar flame. Accordingly, the Modern burners are designed to give high heat
flame speed of a laminar flame is known as the lami- release in short distances. This necessitates fuel veloc-
nar flame speed and is a function of the kinetics of the ities that greatly exceed the laminar flame speed. To
combustion reaction. Under standard conditions, this stabilize such flames, various flame holders are used.
is a function of the fuel chemistry alone. Now suppose For example, an ignition ledge on a burner is a type
that instead of a stationary fuel mixture with a moving of flame holder known as a bluff body. Even if the air
flame front, the fuel is moved. If the fuel is metered flows by the ledge at very high speed, the air speed
exactly at its flame speed, the flame front will remain very close to the ledge will be very slow. The flame
stationary. If the fuel is metered faster than the flame will then establish very near the ignition ledge and be
speed, the flame front will move forward (called liftoff quite stable even over a wide range of firing rates. The
or blowoff). If the fuel is metered slower than its flame burner tile itself is designed with a sudden expansion
into the furnace, which also acts as a flame holder The double-headed arrow indicates that the reaction
because the gas velocity decreases rapidly just after proceeds in both directions simultaneously. When the
the expansion. rate of the forward reaction equals that of the reverse,
the process is in dynamic equilibrium. Equilibrium is
characterized by the following relation:
K = [CO 2 ][H 2 ] [CO][H 2 O] (3.42)

3.12 Substoichiometric Combustion
where the brackets denote wet volume concentrations of
The concept of excess air presumes air in addition to the enclosed species. For substoichiometric combustion,
that required for combustion. However, if one does not it will be useful to define the following quantities: α =
provide enough air, combustion may still continue, gen- [H2]/[H2O], β = [CO]/[CO2], then K = α/β.
erating large quantities of CO and combustibles. This
is referred to as substoichiometric combustion. Process
heaters and boilers should NEVER be operated in this
3.12.2 Substoichiometric Combustion Revisited
mode. Suddenly adding air to such a hot mixture could
result in explosion. Because substoichiometric combus- Now that equilibrium and the water gas shift reaction
tion may have DEADLY consequences, it is useful to have been defined, one can define substoichiometric
consider the process, observe its features, and normally combustion. Solving the mass balance for C, H, and oxy-
avoid it. Further, substoichiometric combustion is an gen, in turn for α and β, and using the relation K = α/β,
important mechanism in the development and use of low one obtains the following equations:
NOx burners when the substoichiometric products may
be important in the design. The stoichiometric ratio, Φ, is
a fuel to air ratio. It has the following relationship with ɛ: 1 2+β x  79 
CH x + +  O2 + N2 
2  1 + β 2(1 + βK )   21 
1
Φ= (3.39)
(1 + ε)  1   β  1 x 
→ CO 2 +  CO +  H2O
 1 + β   1 + β  2  1 + βK 
1 − Φ (3.40)
ε=
Φ 1  βKx   79   1   2 + β x 
+ H2 +      + N2
Equation 3.41 shows a modified form of the general 2  1 + βK   21   2   1 + β 2(1 + βK ) 

equation for hydrocarbon combustion. (3.43)
CH x + a  O 2 + 79 21 N 2  Now, by combining equations for the left side of the rela-

tion, one knows that a must have the following expression:
→ b CO+(1 − b) CO 2 + ( x − c) 2H 2
+ c 2H 2O + a 79 21 N 2 (3.41) 1 x 1 2 + β x 
a= 1+  =  +  (3.44)
Φ 4 2  1 + β 2(1 + βK ) 
where a, x are specified and b, c are unknown having the
One could solve for Φ and substitute into Equation 3.43.
relation 2a = 2 − b + c/2.
However, the equation is quadratic and complicated. An
The reader should note that the formulation neglects
easier solution is to solve for both Φ and the desired spe-
soot. Turns21 has pointed out that using an equilibrium
cies using a parametric relation in β. Equation 3.45 gives
calculation with the water gas shift reaction arrives at a
the relation for Φ:
good approximation for substoichiometric species. This
is adequate for investigating the general features of ( x + 4)(1 + β)(1 + βK )
substoichiometric combustion. Φ= (3.45)
2(2 + β)(1 + βK ) + x(1 + β)
3.12.1 Equilibrium and Thermodynamics Solving for the species as a function of β gives the
following:
The chemical equation gives the water gas shift reaction
as follows: x 79  1 β x 
TWP = 1 + +  + + 
CO + H 2O ↔ CO 2 + H 2 2 21  1 + β 2(1 + β) 4(1 + βK ) 

(O2 + 3.76N2) in proportions to make the mixture flam-

1  βKx  79  1 β x 
TDP = 1 + + + + mable. A spark is then initiated to ignite the mixture.
2  1 + βKx  21  1 + β 2(1 + β) 4(1 + βK ) 
  
The right box represents the process an infinite time

later to ensure all the reactions have gone to comple-
f O2 ,wet = 0 f O2 ,dry = 0 (3.46) tion (i.e., reached equilibrium). In reality, most com-
bustion reactions are completed in only a fraction of a
second. Many species are then present after the reac-
1  1  1  1  tion is completed. The exact composition depends on
f CO2 , wet = f CO2 ,dry = (3.47)
TWP  1 + β  TDP  1 + β  the ratio of the fuel to air. For example, if not enough
air is present, then CO will be generated. If sufficient
1  β  1  β  (3.48) air is present, then little or no CO will be present. This
f CO,wet = f CO,dry = is illustrated in Figure 3.18 which shows the predicted
TWP  1 + β  TDP  1 + β 
species for the adiabatic equilibrium combustion of
1  x  methane and air as a function of the stoichiometry.
f H2 O,wet =   (3.49) For methane, the stoichiometric O2:CH4 ratio for theo-
2 TWP  1 + βK 
retically perfect combustion is 2.0. Stoichiometries less
than 2.0 are fuel-rich, as insufficient oxygen is present
1  βKx  1  βKx 
f H2 , wet = f H2 ,dry = to fully combust the fuel. Stoichiometries greater than
2 TWP  1 + βK  2 TDP  1 + βK  2.0 are fuel-lean, as excess oxygen is present. Figure 3.18
(3.50) shows that the exhaust product composition is highly
dependent on the ratio of the fuel to the oxidizer.
79 1 2+β x 
f N2 , wet = +
21 2 TWP  1 + β 2(1 + βK ) 

3.13.1 Air Preheat Effects
(3.51)
79 1  2 + β x  Figure 3.19 shows the major species for the predicted
f N2 ,dry =  + 
21 2 TDP  1 + β 2(1 + βK )  exhaust gas composition for the stoichiometric combus-
tion of methane with preheated air. There is almost no
change up to temperatures of about 1000°F (540°C), and
Combining the excess air and substoichiometric equa-
only a relatively small change at higher temperatures.
tions, one can construct a graph of species concentra-
Figure 3.20 shows the predicted minor species in the
tions versus Φ, as shown in Figure 3.16a through f for
exhaust gas for the same reaction of ambient tempera-
various fuels on a wet and dry basis. In particular, the
ture methane with preheated air. This graph shows that
substoichiometric portion of the graphs use K = 0.19,
there is a dramatic increase in all the minor species as
which corresponds to a temperature of ∼2200°F (1100°C).
the air preheat temperature increases. This is due to
As Turns21 has pointed out for propane, this gives excel-
chemical dissociation. Figure 3.21 shows the predicted
lent agreement with rigorous equilibrium calculations.
major species in the exhaust products for the combustion
Note that one can generate considerable CO and H2
of preheated methane with ambient air. There is very
from substoichiometric combustion. If air is suddenly
little change in the species concentration with fuel pre-
admitted to such a hot mixture, explosion is likely.
heat. Note that higher fuel preheat temperatures present
safety problems because of the auto-ignition temperature
of methane, which is approximately 1200°F (650°C) in air.
Figure 3.22 also shows that the predicted minor species
concentrations increase with fuel preheat temperature.
3.13 General Discussion
3.13.2 Fuel Blend Effects
In this section, the concepts discussed so far are
applied to combustion in general. Figure 3.17 shows Fuel blends are particularly important in many of the
a schematic of an adiabatic equilibrium process. The hydrocarbon and petrochemical industries. Figure 3.23
boxes represent perfectly insulated enclosures, which shows the predicted major species for the combustion of
do not exist in reality, but are useful for illustrating the air with fuel blends consisting of H2 and CH4. CO2 and N2
concept. The boxes are filled with a combustible mix- decline and H2O increases as the H2 content in the fuel
ture of a fuel and oxidizer, in this case, methane and increases. It is important to note that the species concen-
air, respectively. The left box represents the process trations are not linear functions of the blend composi-
at the time just before a spark is applied to ignite the tion, where the change occurs more rapidly at higher H2
mixture. The only species in the box are CH4 and air compositions. Figure 3.24 is a similar plot of the predicted
20% 90%
88%
18%
86%
16% 84%
H2O Dry N2
82%
14%
80%
12% 78%
CO2
Vol.% N2
76%
10%
74%
CO
8% 72%
Wet N2 70%
6%
68%
O2
4% 66%
H2 64%
2%
Vol.% dry > Vol.% wet 62%
species in all cases
0% 60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(a) Φ
20% 90%
88%
18%
86%
16% 84%
Dry N2
H2O 82%
14%
80%
12% 78%
CO2 Vol.% N2
76%
10%
74%
CO
8% 72%
Wet N2 70%
6%
68%
O2
4% 66%
H2 64%
2%
0% 60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(b) Φ
FIGURE 3.16
Species concentration versus stoichiometric ratio for the following fuels: (a) CH4, (b) natural gas. (From Baukal, C.E. (ed.), The John Zink
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
20% 90%
88%
18%
86%
16% 84%
Dry N2
H2O 82%
14%
80%
12% 78%
CO2
Vol.% N2
76%
10%
74%
CO
8% 72%
Wet N2 70%
6%
68%
O2
4% 66%
H2 64%
2%
0% 60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(c) Φ
20% 90%
Dry N2 88%
18%
86%
16% 84%
H2O 82%
14%
80%
CO2
12% 78%
Vol.% N2
76%
10%
CO 74%
8% 72%
Wet N2 70%
6%
68%
O2
4% 66%
64%
2% H2
0% 60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(d) Φ

Species concentration versus stoichiometric ratio for the following fuels: (c) simulated refinery gas (25% H2, 50% CH4, 25% C3H8), (d) C3H8.
(From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
20% 90%
88%
18%
Dry N2 86%
16% 84%
82%
14%
CO2 80%
12% 78%
Vol.% N2
CO 76%
10%
74%
H2O
8% 72%
Wet N2
70%
6% O2
68%
4% H2 66%
64%
2%
0% 60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(e) Φ
20% 90%
88%
18%
86%
16% 84%
Dry N2
CO 82%
14%
CO2 80%
12% 78%
76% Vol.% N2
10%
74%
H2O
8% 72%
70%
6%
Wet N2 68%
4% O2
66%
H2
64%
2%
0%
60%
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
(f) Φ

Species concentration versus stoichiometric ratio for the following fuels: (e) No. 2 oil, and (f) No. 6 oil. (From Baukal, C.E. (ed.), The John Zink
CO2 , CO, . . .
CH4
H2O, OH, H, . . .
O2 Q=O Q=O
N2, NO, NO2 , . . .
N2
O2, O, . . .
Time = O Time = ∞
T = Ambient T = Flame temperature
P = Atmosphere P = Atmospheric
FIGURE 3.17
Adiabatic equilibrium reaction process. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
O2: CH4 stoichiometry

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
80 80
75 75
70 70
65 65
CO
60 CO2 60
Species concentration (vol.%)

55 H2 55
50 H2O 50
45 N2 45
40 O2 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
O2: CH4 stoichiometry
FIGURE 3.18
Adiabatic equilibrium calculations for the predicted gas composition as a function of the O2:CH4 stoichiometry for air/CH4 flames, where the air
and CH4 are at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
minor species as functions of the H2/CH4 fuel blend. This Real combustion processes are not adiabatic, as the whole
graph also shows strong nonlinearities as the H2 content intent is to transfer heat from the flame to some type of
increases. Figure 3.25 shows the predicted major species load. The amount of heat lost from the process determines
for the combustion of air with fuel blends consisting of the temperature of the exhaust gases. The higher the heat
an inert (N2) and CH4. At the extreme of 100% N2, there is losses from the flame, the lower the exhaust gas tempera-
no fuel left in the “fuel blend” and no combustion takes ture. Figure 3.27 shows the predicted major species for the
place. There is a rapid change in the species concentra- combustion of air and methane as a function of the exhaust
tions as the N2 content increases. Figure 3.26 shows the gas temperature. The peak temperature is the adiabatic
predicted minor species for the combustion of N2/CH4 flame temperature. There is relatively little change in the
fuel blends. This graph also shows a rapid decline in the major species concentration as a function of temperature.
species concentration, in this case for the minor species. Figure 3.28 shows the predicted minor species for the

0 100 200 300 400 500 600 700 800 900 1000
75 75
70 70
65 65
60 60
55 55

CO2
50 50
H2O
45 N2 45
40 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 200 400 600 800 1000 1200 1400 1600 1800 2000
FIGURE 3.19
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the major species as a function of the air preheat
temperature for air/CH4 flames, where the CH4 is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion

0 100 200 300 400 500 600 700 800 900 1000
3.5 3.5
CO
3.0 H 3.0
H2
NO
2.5 2.5
O
OH
O2
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
FIGURE 3.20
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the minor species as a function of the air preheat
temperature for air/CH4 flames, where the CH4 is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion

0 50 100 150 200 250 300 350 400 450 500
75 75
70 70
65 65
60 60
55 55

50 50
CO2
45 H2O 45
40 N2 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 100 200 300 400 500 600 700 800 900 1000
FIGURE 3.21
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the major species as a function of the fuel preheat
temperature for air/CH4 flames, where the air is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion

0 50 100 150 200 250 300 350 400 450 500
1.1 1.1
1.0 1.0
0.9 CO 0.9
H2
0.8 NO 0.8
OH
O2
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0.0 0.0
0 100 200 300 400 500 600 700 800 900 1000
FIGURE 3.22
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the minor species as a function of the fuel preheat tem-
perature for air/CH4 flames, where the air is at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook,
CRC Press, Boca Raton, FL, 2001.)
70 70
60 60

50 CO2 50
H2O
N2
40 40
30 30
20 20
10 10
0 0
0 10 20 30 40 50 60 70 80 90 100
H2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.23
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the major species as a function of the fuel blend (H2 + CH4)
composition for air/fuel flames, where the air and fuel are at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion
1.6 1.6
CO
H
1.4 H2 1.4
NO
1.2 OH 1.2
O2 Species concentration (vol.%)
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
0 10 20 30 40 50 60 70 80 90 100
H2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.24
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the minor species as a function of the fuel blend (H2 + CH4)
100 100
90 90
80 80

70 70
60 60
CO2
50 H2O 50
N2
40 40
30 30
20 20
10 10
0 0
0 10 20 30 40 50 60 70 80 90 100
N2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.25
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the major species as a function of the fuel blend (N2 + CH4)
0.9 0.9
0.8 CO 0.8
H2
0.7 NO 0.7
OH
0.6 O2 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0.0 0.0
0 10 20 30 40 50 60 70 80 90 100
N2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 3.26
Adiabatic equilibrium stoichiometric calculations for the predicted gas composition of the minor species as a function of the fuel blend (N2 + CH4)

0 200 400 600 800 1000 1200 1400 1600 1800
75 75
70 70
65 65
60 60
55 55

50 50
CO2
45 H2O 45
40 N2 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 400 800 1200 1600 2000 2400 2800 3200 3600
FIGURE 3.27
Equilibrium calculations for the predicted gas composition of the major species as a function of the combustion product temperature for air/CH4
flames, where the air and fuel are at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
Boca Raton, FL, 2001.)

0 200 400 600 800 1000 1200 1400 1600 1800
0.9 0.9
0.8 CO 0.8
H2
0.7 NO 0.7
OH
O2 0.6
0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0.0 0.0
0 400 800 1200 1600 2000 2400 2800 3200 3600
FIGURE 3.28
Equilibrium calculations for the predicted gas composition of the minor species as a function of the combustion product temperature for air/CH4
flames, where the air and fuel are at ambient temperature and pressure. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
Boca Raton, FL, 2001.)
combustion of air and methane as a function of the exhaust general chemical equation for fuel oxidation mentioned
gas temperature. The concentrations are essentially zero in Example 3.6 and repeated above, compute
up to temperatures of about 2000°F (1100°C) and rapidly
increase up to the adiabatic flame temperature. N ppmvd lb mol Chem lb mol  FG 
× ×  
106 lb mol FG lb mol  F 
1 − lb mol Fuel MWchem lbm lb Fuel

× × × m
( MWF ) lb Fuel lb mol Chem HHV
3.14 Emissions
106 Btu lb m
Common examples of gaseous emissions are NOx, CO, × =
million 106 Btu
and SOx. The concentration of each pollutant is mea-
sured at the stack in dry parts per million volume
(ppmvd). As ppmvd varies with the amount of O2 in the where N ppmvd = ppmv of chemical x
stack, it is necessary to establish a reference percentage
of O2, for example, 100 ppm NOx at 3% O2.
In order to establish the percentage of dry O2, use the
 FG   x  x
excess air measured in the stack:   = 1 + ( ε − 1)  1 +  + ε  1 +  (3.54)
F dry,vol  4 4
 x
CH x + ε  1 +  ( O 2 + 3.76 N 2 )
 4 % Excess air
Substituting φ = + 1 and simplifying gives
x  x  x 100
→ CO 2 + H 2O + ( ε − 1)  1 +  O 2 + ε  1 +  3.76 N 2
2  4 4
N ppmvdMW c   x   4.76 EA 
1 + 1 +   + 3.76 
From the products of the general form of hydrocarbon HHV (12 + x )   4   100 
reactions mentioned earlier obtain
lb m N
= (3.55)
106 Btu
x
(ε − 1)  1 +  100
4
O 2 %dry = (3.52)
 x  x This equation is a general relationship between a prod-
1 + ( ε − 1)  1 +  + ε  1 +  3.76
 4  4 uct species in dry PPM by volume and lbm 106 Btu for any
where chemical compound:
 H
  = “x”
C mole Btu
HHV = and
lbm
% Excess air
Substituting φ = +1
100 MWc = Molecular weight of “N” chemical EA

= % Excess air
The general relation for EA and O2 dry for any HC fuel is
1
O 2 %dry = (3.53)
 
 1  1 3.15 Quick Sizing
 + 3.76 + 0.0476
  1 + x   EA 3.15.1 Finding Saturated Humidity
  4 
A shortcut is to use steam property tables of saturated
pressure versus temperature to find the saturated
It is often necessary to calculate mass emission rate humidity in the air. Water vaporizes at 100°C (212°F)
lbs 
per Btu produced  6 from ppmvd. Using the which, according to the steam tables, yields a satura-
 10 Btu  tion pressure of 101.42 kPa (14.7 psi). To get saturated
humidity, then simply take the ratio of Psat at any At STP conditions
temperature to 101.32 kPa in order to get saturation
humidity: TSTP = 530°R
PSTP = 14.6 psia

Psat @ Temp
Saturated humidity = (3.56) lbm
Psat @ Boiling temp ρair STP = 0.075
ft 3
Example 3.13 18
SGH2O = = 0.621
For example, suppose a room of 35°C, then look- 28.97
ing up the properties of steam at 35°C in the steam
tables yields Applying Equation 3.57
At 35° C → Psat of 5.63 kPa.

 40 psia   760°R  lbm
And Psat at boiling = 101.42 kPa at boiling ρsteam = (0.075)(0.621)     = 0.0884 3
 14.7 psia  530°R ft

5.63 kPa
Saturated humidity = = 5.55%
101.42 kPa From standard steam tables, the exact density = 0.0906 lbm ft 3
yielding a small 2.4% difference.
Thus a room at 35°C yields a saturated humidity of

5.55%.
3.15.2 Stoichiometric Combustion of Air Simplified References

In finding air-to-fuel ratio by volume it is generally 1. C. Crowe et al. (ed.), Engineering Fluid Mechanics, 9th
accepted to assume that all gaseous paraffinic hydro- edn., John Wiley & Sons Inc., Hoboken, NJ, 2009.
carbons consume a STP ft3 (70°F) of air for every 106 Btu 2. W. Nebergall, General Chemistry, 4th edn., D.C. Heath
(HHV) fired (106 Btu/ft 3air ). For the case of liquid fuels, and Company, Lexington, MA, 1972.
it is also accepted that the reaction consumes a STP ft3 of 3. T. Brown et al., Chemistry: The Central Science, 10th edn.,
Pearson Education, Upper Saddle River, NJ, 2006.
air for every 100 Btu fired (100 Btu/ft 3air ). Similarly, coal
4. F. Vandaveer, Gas Engineers Handbook, 1st edn., Industrial
reactions, on average, consume a cubic foot of air for Press, New York, 1965.
every 98 Btu fired (98 Btu/ft 3air ) . The actual Btu/ft 3air for 5. Y. Cengel, Heat and Mass Transfer: A Practical Approach,
coal and oil can be found in Table 4.6. For gaseous fuels 3rd edn., McGraw Hill, Willard, OH, 2007.
the reader can compute using the methods outlined in 6. C. McQuiston and J.D. Parker, Heating, Ventilating, and
Section 3.7 or computed from Table 3.3. Air Conditioning, John Wiley & Sons, New York, 1982.
7. R. Bolz et al., Handbook of Tables for Applied Engineering
Science, 2nd edn., CRC Press, Cleveland, OH, 1973.
3.15.3 Density of Low-Pressure Gases
8. F. Huang, Engineering Thermodynamics: Fundamentals and
A practical shortcut to find the density of a gas at low Applications, Macmillan Publishing Co., New York, 1976.
pressures is to use the following equation: 9. P. Schmidt, Fuel Oil Manual, 4th edn., Industrial Press,
New York, 1985.
 MWgas   Pgas   Tgas  10. Y. Cengel et al., Thermodynamics: An Engineering Approach,
Pressure = (0.075)  (3.57) 6th edn., McGraw Hill, Willard, OH, 2008.
 MWair   Pair STP   Tair STP 
11. F. Kreith (ed.), The CRC Handbook of Mechanical
Specific gravity Engineering, CRC Press, Boca Raton, FL, 1998.
12. D.R. Stull and Prophet, H., JANAF Thermochemical Tables,
Example 3.14 2nd edn., National Standards Reference Data System,
Washington, DC, 1971.
Find the density of steam at 300°F and 35.3 psig
13. S.C. Stultz and Kitto, J.B., Steam: Its Generation and Use,
Convert temperature to Rankine and gauge pres-
40th edn., Babcock & Wilcox Company, Barberton, OH,
sure to absolute.
1992.
14. R.H. Barnes, M.J. Saxton, R.E. Barrett, and A. Levy,
Tgas = 300 + 460 = 760°R Chemical Aspects of Afterburner Systems, Battelle
Columbus Laboratories, IPA-600/7-79-096, NTIS publi-
Pgas = 25.3 + 14.7 = 40 psia cation PB298465, April 1979.

15. G.C. Williams, Hottel, H.C., and Morgan, A.C., The com- 18. C.T. Bowman, Kinetics of pollution formation and
bustion of methane in a jet-mixed reactor, 12th Symposium destruction in combustion, Progress in Energy and
(International) on Combustion, The Combustion Institute, Combustion Science, 1:33–45, 1975.
Pittsburgh, PA, 1969. 19. J.L. Reese et al., State-of-the-art of NOx emission con-
16. V.S. Engleman, Bartok, W., Longwell, J.P., and Edelman, trol technology, Proc. International Joint Power Generation
R.B., Experimental and theoretical studies of NOx for- Conference, Phoenix, AZ, October 3–5, 1994.
mation in a jet stirred combustor, 14th Symposium 20. M. Zabethakis, Flammability characteristics of combus-
(International) on Combustion, The Combustion Institute, tible gases and vapors, U.S. Department of Commerce,
Pittsburgh, PA, 1973. Springfield, CA, 1965.
17. A.E. Westenberg, Turbulence modeling for CFD, 21. S.R. Turns, An Introduction to Combustion, McGraw-Hill,
Combustion Science and Technology, 4:59–67, 1971. New York, 1996.
4
Fuels
John Ackland, Jeff White, Richard T. Waibel, and Stephen B. Londerville
CONTENTS
4.1 Gaseous Fuels................................................................................................................................................................... 72
4.1.1 Introduction.......................................................................................................................................................... 72
4.1.2 Natural Gas........................................................................................................................................................... 72
4.1.3 Liquified Petroleum Gas..................................................................................................................................... 73
4.1.4 Refinery Gases...................................................................................................................................................... 73
4.1.5 Combustible Waste Gas Streams....................................................................................................................... 75
4.1.5.1 PSA Tail Gas........................................................................................................................................... 75
4.1.6 Flexicoking Waste Gas........................................................................................................................................ 76
4.1.7 Physical Properties of Gaseous Fuels............................................................................................................... 77
4.2 Gas Property Calculations.............................................................................................................................................. 77
4.2.1 Molecular Weight................................................................................................................................................. 77
4.2.2 Lower and Higher Heating Values.................................................................................................................... 82
4.2.3 Specific Heat Capacity......................................................................................................................................... 82
4.2.4 Flammability Limits............................................................................................................................................ 83
4.2.5 Burning Velocity.................................................................................................................................................. 84
4.2.6 Viscosity................................................................................................................................................................ 84
4.2.7 Derived Quantities.............................................................................................................................................. 84
4.2.7.1 Partial Pressure..................................................................................................................................... 84
4.2.7.2 Adiabatic Flame Temperature............................................................................................................. 85
4.2.7.3 Other Special Cases.............................................................................................................................. 85
4.3 Liquid Fuels...................................................................................................................................................................... 85
4.3.1 Production, Refining, and Chemistry............................................................................................................... 85
4.3.2 Oils......................................................................................................................................................................... 88
4.3.2.1 Light Oils................................................................................................................................................ 88
4.3.2.2 Heavy Oils............................................................................................................................................. 89
4.3.2.3 Residual Oils.......................................................................................................................................... 89
4.3.3 Liquid Naphtha.................................................................................................................................................... 90
4.3.4 Physical Properties of Liquid Fuels................................................................................................................... 90
4.3.4.1 Flash Point.............................................................................................................................................. 90
4.3.4.2 Pour Point............................................................................................................................................... 91
4.3.4.3 Distillation.............................................................................................................................................. 91
4.3.4.4 Viscosity................................................................................................................................................. 91
4.3.4.5 Density, Gravity, Specific Volume, and Specific Weight.................................................................. 91
4.3.4.6 Heat Capacity (Specific Heat).............................................................................................................. 93
References................................................................................................................................................................................... 94
71
TABLE 4.1
4.1 Gaseous Fuels Example Pipeline-Quality Natural Gas
Minimum Maximum
4.1.1 Introduction
Major and minor components (vol%)
The term “gaseous fuel” refers to any combustible fuel Methane 75%
that exists in the gaseous state under normal tempera- Ethane — 10.0%
tures and pressures. Gaseous fuels are typically com- Propane — 5.0%
posed of a wide range of chemical compounds. Low Butane — 2.00%
boiling point hydrocarbons (both paraffins and olefins), Pentane and heavier — 5.00%
hydrogen, carbon monoxide, and inert gases (nitrogen Nitrogen and other inerts — 3%–4%
and carbon dioxide) are among the many chemical Carbon dioxide — 3%–4%
constituents of common gaseous fuels. The purpose of Trace components
this section is to introduce many of the common fuel Hydrogen sulfide — 0.25–1.0 grains/100 scf
gas mixtures used as fuel in the hydrocarbon and pet- Mercaptan sulfur — 0.25–1.0 grains/100 scf
rochemical industries. Commonly occurring waste gas Total sulfur — 5–20 grains/100 scf
mixtures in flare systems are also described. Water vapor — 7.0 lb/mmcf
Oxygen — 0.2–1.0 ppmv
Other characteristics
4.1.2 Natural Gas Heating value, 950 1150
Btu/scf-gross saturated
Natural gas is a gaseous fossil fuel that is formed natu-
rally beneath the Earth and is typically found with or Source: Adapted from Gas Processors and Suppliers Association,
near crude oil reservoirs. According to the U.S. Energy GPSA Engineering Data Book, Vol. I, 10th edn., Tulsa, OK, 1987.
With permission.
Information Administration, in 2009, proven natural gas
Liquids: Free of liquid water and hydrocarbons at delivery tempera-
reserves in the United States totaled approximately 7.717 × ture and pressure.
1012 m3 (2.725 × 1014 ft3).1 Total natural gas consumption in Solids: Free of particulates in amounts deleterious to transmission
the United States in 2009 was estimated at 23.344 quadril- and utilization equipment.
lion Btu, 24.7% of the total U.S. energy consumption.2
Failure to remove the water vapor from raw natural
Natural gas consists of a fluctuating range of low boil-
gas prior to introduction to the pipeline network will
ing point hydrocarbons. Methane is the primary chemi-
result in increased corrosion rates, formation of solid
cal component, and can be present in amounts ranging
hydrate compounds that can restrict or interrupt gas
from 70% to 99.6% by volume. Ethane can be present in
flow, and freezing of valves and regulators during cold
amounts ranging from 2% to 16% by volume. Carbon
weather conditions.5 Techniques for the dehydration of
dioxide, nitrogen, hydrogen, oxygen, propane, butane,
natural gases include the following:
and heavier hydrocarbons are also typically present
in the fuel analysis.3 The exact analysis usually varies 1. Absorption with liquid desiccants: Glycols
somewhat depending on the source of the gas and on (typically triethylene glycol) are used to absorb
any heating value adjustments or supplementation. water vapor via countercurrent-flow, packed-
Natural gas quality specifications have historically bed absorption columns.6
been negotiated in individual contracts between the
2. Adsorption with solid desiccants: Water vapor
natural gas producer and the purchaser or pipeline
is adsorbed onto a bed of inorganic porous solid
company. Specification parameters often include upper
material (silica gel, alumina, molecular sieves,
and lower limits for heating value, chemical composi-
etc.).6,7
tion, contaminants, water content, and hydrocarbon
dew point. Table 4.1 outlines general specifications for 3. Dehydration with calcium chloride: Solid anhy-
pipeline-quality natural gas, as provided by the Gas drous calcium chloride (CaCl2) absorbs water
Processors Suppliers Association (GPSA).4 Typical com- from the wet natural gas and forms various
mercial natural gas compositions, listed by production calcium chloride hydrates (CaCl2 × H2O). These
region, are contained in Table 4.2. hydrates are removed from the natural gas
In addition to the primary combustible and inert stream as a calcium chloride brine solution.6
chemical components discussed earlier, raw natural gas 4. Refrigeration: A refrigeration coil is used to cool
can also contain undesirable amounts of water, hydrogen and condense water vapor from the wet natu-
sulfide, and/or carbon dioxide. Before the raw natural ral gas stream. Separation of the liquid phase
gas can be deposited into a pipeline transmission net- is accomplished via a two-phase, vapor/liquid
work, these undesirable components must be removed. separation drum.5
Fuels 73
TABLE 4.2
Commercial Natural Gas Components and Typical Ranges of Composition
Sample Gas Compositions by Production Region (vol%)
Fuel Gas Tulsa, North
Component OK Alaska Algeria Netherlands Kuwait Libya Sea Alabama Ohio Missouri Pennsylvania
CH4 93 100 87 81 87 70 94 90 94 84 83
C2H6 3 — 9 3 9 15 3 5 3 7 16
C3H8 1 — 3 <1 2 10 1 — <1 — —
C4H10 <1 — 1 <1 1 4 <1 — <1 — —
C5 and — — — — — — — — — — —
higher
CO2 1 — — 1 2 — <1 — 1 1 —
N2 2 1 <1 14 1 1 2 5 1 8 1
O2 — — — — — — — — <1 — —
H2 — — — — — — — — <1 — —
Total 100 100 100 100 100 100 100 100 100 100 100
Source: Adapted from Reed, R.J., North American Combustion Handbook, Vol. 1, North American Mfg. Co., Cleveland, OH, 1986. (Courtesy of
Fives North American Combustion, Inc.)
Hydrogen sulfide must be removed from the raw nat- conditions. LPG is vaporized for use as a fuel. The pri-
ural gas stream due to air pollution considerations and mary chemical components of LPG are propane, propyl-
corrosion hazards. The hydrogen sulfide content of com- ene, normal butane, isobutane, and butylene.8 The GPSA
mercial natural gas rarely exceeds 1.0 grains/100 std. ft3 Engineering Data Book contains industry standard prod-
(0.023 g/m3). A majority of pipeline companies respond- uct specifications for commercial propane (predomi-
ing to a 1994 poll limited hydrogen sulfide concen- nantly propane and/or propylene), commercial butane
trations to less than 0.3 g/100 std. ft3 (0.007 g/m3).8 In (predominantly butane and/or butylene), and commer-
addition, carbon dioxide is often removed from the raw cial butane–propane mixtures.4 LPG produced via the
gas because the inert component weakens the overall separation of heavier hydrocarbons from natural gas is
heating value of the gas stream.5 There are numerous mainly paraffinic, containing primarily propane, normal
commercial processes (chemical reaction, absorption, butane, and isobutane. LPG derived from oil-refinery gas
and adsorption) for the removal of acidic components may contain varying small amounts of olefinic hydrocar-
(H2S and CO2) from raw natural gas streams. The GPSA bons such as propylene and butylene.8
Engineering Data Book discusses many of these hydro- Most of the LPG used in the United States consists
carbon treatment processes in detail.6 Hydrogen sulfide primarily of propane.9 Due to their relatively high boil-
removed from the raw gas is generally converted to ele- ing point, LPG mixtures containing high concentra-
mental sulfur via the Claus process.6 tions of normal butane (boiling point = 31°F or −1°C at
After the necessary purification processes have been atmospheric pressure) or isobutane (boiling point = 11°F
completed, the commercial-grade natural gas is com- or −12°C at atmospheric pressure) are preferred for use
pressed to approximately 1000 psig (6.9 MPag) and is in warm climates. Conversely, LPG mixtures containing
introduced to a natural gas pipeline distribution net- high concentrations of propane (boiling point = −44°F
work.5 The gas is recompressed along the path to the or −42°C) are typically preferred for use in cold climates.5
consumer as necessary. Operating pressure at an indi- LPG is often used in the hydrocarbon or petrochemical
vidual natural gas burner located at a process furnace industry as a fuel gas supplement or as a standby/start-
inside a petrochemical or hydrocarbon processing facil- up fuel. However, due to its value as both a common pet-
ity is reduced to an operating pressure range that typi- rochemical feedstock and a marketable commodity, LPG
cally varies between 5 and 30 psig (34 and 207 kPag), is not typically preferred as a primary processing fuel.10
depending on the furnace’s heating requirements and
the individual burner’s design specifications.
4.1.4 Refinery Gases
Although commercial natural gas and LPG are often
4.1.3 Liquified Petroleum Gas
used as fuels in processing plants, internally generated
Liquified petroleum gas (LPG) is the general term used refinery fuel gases serve as the primary fuel component
to describe a hydrocarbon that is stored as a liquid under for most refineries, petrochemical plants, and hydrocar-
moderate pressure, but is a gas under normal atmospheric bon facilities. It is not usual for a process unit to produce
its own fuel supply. Often, fuel gas streams from vari- to ensure a well-mixed, homogeneous fuel gas mixture.
ous processing units are delivered to a common mixing However, static mixers are often impractical in the pet-
point within the plant, before the new gas mixture is rochemical and hydrocarbon processing industries,
returned to the processing units as refinery gas. Refinery typically due to pressure drop limitations of the refin-
fuel gases contain an extremely wide variety of chemi- ery fuel gas system.
cal constituents, including paraffins, o lefins, diolefins, Another problem often associated with the com-
aromatics, mercaptans, organic sulfides, ammonia, bustion of refinery fuel gases is the presence of liquid
hydrogen sulfide, carbon monoxide, carbon dioxide, etc. hydrocarbons in the refinery fuel gas stream, which
Because plants must operate in a manner best suited to can accelerate the coking and plugging rates of down-
maximize profit, the individual fuel gas streams origi- stream gas burner components. Sources of unwanted
nating at each process unit will vary in composition and liquid hydrocarbons in refinery fuel gas streams include
quantity, depending on numerous economic and tech- condensation of heavier fuel gas components (C5 and
nical factors.11 Table 4.3 contains typical chemical com- higher) due to natural cooling of the fuel gas stream,
positions of fuel gas streams originating from various liquid entrainment into absorber or fractionator over-
process units within a petroleum refinery.8 head gas streams, and lubrication oil contamination of
It is very important that the refinery fuel gas leaving the fuel gas stream. Potential solutions for the problems
the common mixing point is a homogenous mixture of associated with these liquid hydrocarbons include liq-
the fuel gas streams supplied. If the individual fuel gas uid extraction of the heavier chemical components (C5
supply streams vary significantly in calorific value, and and heavier) and filtration/coalescence of liquid compo-
if the supply streams are not combined in a homoge- nents from the gas stream. In addition, increasing the
neous manner, the calorific value of the nonhomoge- velocity of the flowing gas through burner components
neous refinery fuel gas mixture will also vary widely (tips, risers, etc.) has been proven to cool the hardware
and often instantaneously. Unless the gas burners and and inhibit the cracking reactions that eventually lead
control systems at each processing furnace have been to plugging and coking.
designed to accommodate instantaneous changes in Wet fuel gas can introduce problems in cooler cli-
fuel gas calorific value, the process will likely be impos- mates associated with the condensation and subsequent
sible to control. All of the combustion performance freezing of water vapor inside the fuel gas system. If the
parameters—including burner stability, emissions con- water vapor reaches the dew point in a cold atmospheric
trol, heat transfer efficiency, and heat flux—will suffer environment, there is danger of frost stoppage, freezing,
as a result of the nonhomogeneous fuel mixture.10 Static or bursting of lines—a considerable fire safety hazard
mixers are often used in various segments of industry that merits serious thought. Options to combat water
TABLE 4.3
Composition of a Typical Refinery Gas
Refinery Fuel Gas Source (Dry Gas)
Fuel Gas Cracked Gas Coking Gas Reforming Gas Combined Refinery Combined Refinery
Component (%) (%) (%) FCC Gas (%) Gas—Sample 1 (%) Gas—Sample 2 (%)
CH4 65 40 28 32 36 53
C2H4 3 3 7 7 5 2
C2H6 16 21 28 9 18 19
C3H6 2 1 3 15 8 6
C3H8 7 24 22 25 20 14
C4H8 1 — — — — —
C4H10 3 7 7 0 2 1
C5 and higher 1 — — — — —
H2 3 4 5 6 3 3
CO — — — — — —
CO2 — — — — — —
N2 — — — 7 8 3
H2O — — — — — —
O2 — — — — — —
H2S — — — — — —
Total 100 100 100 100 100 100
Source: Adapted from Nelson, W.L., Petroleum Refining Engineering, 3rd edn., Marcel Dekker, New York, 1949.
Fuels 75
present in the fuel gas system include dehydration sys- TABLE 4.4
tems (as discussed in Section 4.2.1) and steam/electric Typical Composition of Steam Reforming/PSA Tail Gas
tracing of refinery fuel gas lines.10
Fuel Gas Component PSA Tail Gas Composition (vol%)
CH4 17
4.1.5 Combustible Waste Gas Streams H2O <1
H2 28
The quantity and variety of combustible waste gas
CO2 44
streams in the hydrocarbon and petrochemical indus-
CO 10
tries is virtually unlimited. Many of these waste gas
N2 <1
streams are relatively high in inert concentration,
with large amounts of nitrogen and carbon diox- Total 100
ide often present. As a result, these waste fuels are Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook,
often low in heat content, with lower heating values CRC Press, Boca Raton, FL, 2001.
(LHVs) in the range of 400–800 Btu/scf (0.42–0.84 MJ/
PSA is a cyclic process that uses beds of solid adsor-
Nm3). For these reasons, waste fuels are not usually
bent to remove impurities such as carbon dioxide, carbon
compressed into the main refinery fuel gas system.
monoxide, methane, and nitrogen from the hydrogen
Two of the most widely used combustible waste gas
production stream. A simplified process flow diagram
fuels, pressure swing adsorption (PSA) tail gas and
of a typical steam reforming hydrogen production unit
Flexicoking gas, are discussed in detail in the follow-
using PSA is shown in Figure 4.1.
ing sections.
The steam reforming process is conducted in four
stages8,9,12,13:
4.1.5.1 PSA Tail Gas
1. Feedstock preparation: Feedstock (light hydro-
PSA tail gas is a low-pressure, low-Btu fuel gas produced carbons such as methane, propane, butane, and
as a by-product of a PSA process, a key purification com- light liquid naphtha) at approximately 450 psig
ponent in the steam reforming hydrogen production (31 barg) is preheated and purified to remove
process. Table 4.4 contains the approximate composition reformer catalyst poisons such as halogens and
of a typical PSA tail gas fuel stream. sulfur-containing compounds.
Feed
Steam Steam
generation
Reforming
Purification
Shift
Steam
generation
Cooling
99.9% Vol. H2
PSA
FIGURE 4.1
Simplified process flow diagram for hydrogen reforming/PSA. (Adapted from Meyers, R.A., Handbook of Petroleum Refining Processes, 2nd edn.,
McGraw-Hill, New York, Chapter 6.2, 1997.)
2. Reforming: The purified feedstock is reacted with TABLE 4.5

steam to form carbon monoxide and hydrogen: Typical Composition of Flexicoking Waste Gas
Flexicoking Waste Gas
 m
C nH m + nH 2O 1500° F and Ni Catalyst
→  n +  H 2 + nCO Fuel Gas
Composition (by Volume)
 2 Component Sample 1 Sample 2
(4.1) CH4 1.0% 0.8%
H2 20.0% 21.0%
The reaction is endothermic and occurs within the
CO2 10.0% 10.5%
process tubes of a reformer furnace in the presence
CO 20.0% 18.6%
of nickel catalyst at approximately 1500°F (815°C). N2 45.0% 45.6%
3. Shift conversion: The water–gas shift reaction H2O 4.0% 3.5%
is employed to convert the carbon monoxide H2S 150 ppm 0
produced in the reforming step into additional COS 120 ppm 120 ppm
hydrogen and carbon dioxide: Total 100% 100%
CO + H 2O → H 2 + CO 2 (4.2) Source: Adapted from Meyers, R.A., Handbook of Petroleum
Refining Processes, 2nd edn., McGraw-Hill, New
The shift conversion step is exothermic and is York, 1997.
conducted at approximately 650°F (343°C) in the
presence of a chromium/iron oxide catalyst.
continuous fluidized-bed thermal cracking process
4.
Hydrogen purification/PSA: Following the
used in the conversion of heavy hydrocarbon feedstocks
shift conversion step, the hydrogen production
(typically heavy gas-oils from atmospheric and vacuum
stream enters the PSA portion of the process.
distillation) to various gaseous and liquid hydrocarbon
Adsorbent beds remove the impurities (carbon
products. Table 4.5 contains the approximate composi-
dioxide, carbon monoxide, methane, and nitro-
tion of two sample Flexicoking waste gas fuel streams.14
gen) and a small portion of the product. Typical
A simplified process flow diagram of the Flexicoking
hydrogen recovery is 80% or greater, with prod-
process is shown in Figure 4.2. In the Flexicoking process,
uct purity of approximately 99.9 vol%.
hot (500°F–700°F or 260°C–370°C) gas-oil is injected into
the reactor vessel containing hot, fluidized coke particles.
The PSA unit must be frequently regenerated via Thermal cracking reactions inside the reactor vessel pro-
depressurization of the adsorbent beds. When depres- duce fresh petroleum coke that is deposited as a thin film
surization occurs, PSA tail gas (sometimes referred to as on the surface of existing coke particles inside the reactor
PSA waste gas) is produced at a pressure of about 5 psig bed. Cracked vapor products exit the Flexicoking process
(0.35 kg/cm2g) or less. The PSA tail gas consists of the through the reactor vessel overhead stream for additional
impurities removed by the adsorbent beds, as well as downstream processing. Coke from the reactor vessel is
the hydrogen that is not recovered in the product stream. continuously injected into the top of a second fluidized
The tail gas serves as the primary fuel for the reformer vessel, the coke heater, where it is heated and recycled
furnace burners. Due to flame stability problems associ- to maintain a reactor-bed temperature of 950°F–1000°F
ated with firing the low-pressure, high-inert-concentra- (510°C–540°C). A portion of the coke fed into the top
tion (carbon dioxide and nitrogen) PSA tail gas alone, section of the coke heater is injected into the bottom of
the PSA tail gas is typically supplemented by a light a third fluidized vessel, the gasifier. Inside the gasifier,
refinery fuel gas. The PSA and refinery fuel gases are the coke is reacted with air and steam at approximately
fired in a dual-fuel burner specifically designed for the 1500°F–1800°F (820°C–980°C), producing a low-energy
steam reforming/PSA process. In this arrangement, the fuel gas, or Flexicoking gas, consisting primarily of nitro-
PSA and refinery fuel gases enter the combustion zone gen, hydrogen, carbon monoxide, and carbon dioxide.
through separate fuel connections and burner nozzles. The Flexicoking gas flows from the top of the gasifier
The dual-fuel burners are capable of firing the two fuel to the bottom of the heater, where it provides the heat
mixtures separately or simultaneously, with PSA gas never necessary to maintain the reactor-bed temperature and
providing more than 85 vol% of the total reformer fuel. helps fluidize the coke heater bed. The high-temperature
Flexicoking gas leaving the coke heater is used for high-
4.1.6 Flexicoking Waste Gas
pressure steam generation before entrained coke fines are
Flexicoking waste gas is a low-pressure, low-energy removed in a cyclone/venturi scrubber system. Because
fuel gas produced by petroleum refiners as a by-prod- the low-energy gas stream leaving the Flexicoking pro-
uct of the Exxon Flexicoking process. Flexicoking is a cess contains substantial concentrations of H2S (~150 ppm
Fuels 77
Tertiary Low Btu gas

cyclones to cleanup
To fractionator
Steam
generation
Scrubber
Start
Slurry
Dry fines Venturi
scrubber
Gasifier
Reactor
Heater
Steam
Air
blower
FIGURE 4.2
Simplified process flow diagram for flexicoking. (From Meyers, R.A., Handbook of Petroleum Refining Processes, 2nd edn., McGraw-Hill,
New York, Chapter 12.1, 1997.)
by volume), the gas must first be sent through a hydrogen Having assembled this information, the following for-
sulfide removal system before it can be burned as fuel.9,13 mulae are used to calculate molecular weight:
1
4.1.7 Physical Properties of Gaseous Fuels MW = ∑ MW × y =i i (4.3)
Tables 4.6 through 4.8 provide physical and combustion ∑ x /MW i i
property data for a large variety of common fuel gas

mixtures and their chemical components. where
MW is the molecular weight of the mixture
MWi is the molecular weight of component i
yi is the mole fraction of component i
xi is the mass fraction of component i
4.2 Gas Property Calculations
4.2.1 Molecular Weight Example 4.1
Molecular weight is the mass in grams of 1 g-mol of a Given the mass percentages of 10% H2O, 25% CO2,
and 65% N2. Convert the mass percent to a vol-
chemical compound. Avogadro’s number defines the
ume percent basis.
number of molecules in a gram-mole to be 6.02252 × Taking the mass percentage composition of each
1023, a fundamental constant. To determine the molecu- gas and molecular weight will give the moles of
lar weight of a mixture of gases, it is necessary to know each gas per gram. The number of moles of H2O
the molecular weight of each compound and the com- in 1 g of mix gas will be 0.1 (gramH2O/gramMix) ×
position of the gases in terms of mole or mass fractions. (1 mol/18 g) = 0.0055 mol/g.
78
TABLE 4.6
Volumetric Analysis of Typical Gaseous Fuel Mixtures
Natural Gas (%) LPG (%) Refinery Gases (Dry) (%) Waste Gases (%)
Refinery Refinery
Fuel Gas Cracked Coking Reforming FCC Gas Gas PSA Flexicoking
Component Tulsa Alaska Netherlands Algeria Propane Butane Gas Gas Gas Gas Sample 1 Sample 2 Gas Gas
CH4 93.4 99 81 87 — — 65 40 28 32 36 53 17 1
C2H4 — — — — — — 3 3 7 7 5 2 — —
C2H6 2.7 — 3 9 — — 16 21 28 9 18 19 — —
C3H6 — — — — — — 2 1 3 15 8 6 — —
C3H8 0.6 — 0.4 2.7 100 — 7 24 22 25 20 14 — —
C4H8 — — — — — 100 1 — — — — — — —
C4H10 0.2 — 0.1 1.1 — — 3 7 7 0 2 1 — —
C5 and — — — — — — 1 — — — — — — —
higher
H2 — — — — — — 3 4 5 6 3 3 28 21
CO — — — — — — — — — — — — 10 20
CO2 0.7 — 0.9 — — — — — — — — — 44 10
N2 2.4 1 14 0 — — — — — 7 8 3 <1 45
H2O — — — — — — — — — — — — <1 3
O2 — — — — — — — — — — — — — —
H2S — — — — — — — — — — — — — —
Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
The Coen & Hamworthy Combustion Handbook
Fuels
TABLE 4.7
Physical Constants of Typical Gaseous Fuel Mixtures
Natural Gas LPG Refinery Gases (Dry) Waste Gases
Refinery Refinery
Cracked Coking Reforming FCC Gas Gas PSA Flexicoking
Fuel Gas Component Tulsa Alaska Netherlands Algeria Propane Butane Gas Gas Gas Gas Sample 1 Sample 2 Gas Gas
Molecular weight 17.16 16.1 18.51 18.49 44.1 58.12 22.76 28.62 30.21 29.18 28.02 24.61 25.68 23.73
LHV (Btu/SCF) 913 905 799 1,025 2,316 3,010 1,247 1,542 1,622 1,459 1,389 1,297 263 131
HHV (Btu/SCF) 1,012 1,005 886 1,133 2,517 3,262 1,369 1,686 1,769 1,587 1,515 1,421 294 142
Specific gravity 0.59 0.56 0.64 0.64 1.53 1.1 0.79 0.99 1.05 1.01 0.97 0.85 0.89 0.82
(14.696 psia/60°F,
Air = 1.0)
Wobbe number, 1,318 1,343 1,108 1,416 2,035 3,110 1,540 1,694 1,726 1,579 1,538 1,541 312 157
HHV/(SG1/2)
Isentropic coefficient 1.30 1.31 1.31 1.28 1.13 1.10 1.24 1.19 1.19 1.20 1.21 1.23 1.33 1.38
(Cp/Cv)
Stoichiometric air 10,554 10,567 10,554 10,525 10,369 10,371 10,402 10,379 10,322 10,234 10,311 10,375 9,667 8,265
required
(SCF/MMBtu)
Stoichiometric air 805 806 805 803 791 791 794 792 787 781 787 792 738 630
required
(lbm/MMBtu)
Air required for 15% 12,138 12,152 12,138 12,104 11,925 11,926 11,962 11,936 11,870 11,769 11,858 11,931 11,117 9,505
excess air
(SCF/MMBtu)
Air required for 923 924 923 920 907 907 910 908 903 895 902 907 845 723
15% excess air
(lbm/MMBtu)
Volume of dry 10,983 10,956 11,141 10,953 10,962 10,996 10,890 10,909 10,871 10,847 10,911 10,904 11,722 13,517
combustion products
(SCF/MMBtu)
Weight of dry 865 862 876 863 870 874 861 864 862 860 864 862 985 1,103
combustion products
(lbm/MMBtu)
Volume of wet 13,257 13,258 13,415 13,163 12,788 12,757 12,935 12,862 12,771 12,689 12,821 12,902 14,198 15,585
combustion products
(SCF/MMBtu)
Weight of wet 973 971 984 968 957 958 958 957 952 948 864 957 1,102 1,201
combustion products
(lbm/MMBtu)
AFT (°F) 3,306 3,308 3,284 3,317 3,351 3,351 3,342 3,348 3,359 3,371 3,353 3,345 3,001 2,856
Note: All values calculated using 60°F fuel gas and 60°F, 50% relative humidity combustion air.
79
TABLE 4.8
Physical Constants of Typical Gaseous Fuel Mixture Components
Gas Density Ideal Gas, 14.696
Specific Heat Latent Heat of psia, 60°F Btu/scf Heating Value
Boiling Vapor Capacity, Cp Vaporization
Btu/ft
3 Btu/lbm
Point Pressure 60°F and 14.696 psia and Specific Gas Specific
Fuel Gas Chemical Molecular 14.696 100°F 14.696 psia Boiling Point Gravity Density Volume LHV HHV LHV HHV
No. Component Formula Weight psia (°F) (psia) (Btu/lbm/°F) (Btu/lbm) (Air = 1) (lbm/ft3) (ft3/lbm) (Net) (Gross) (Net) (Gross)
Paraffin (alkane) series (CnH2n+2)

1 Methane CH4 16.04 −258.69 — 0.5266 219.22 0.554 0.042 23.651 909 1010 21,495 23,875
2 Ethane C2H6 30.07 −127.48 — 0.4097 210.41 1.038 0.079 12.618 1619 1770 20,418 22,323
3 Propane C3H8 44.10 −43.67 190 0.3881 183.05 1.522 0.116 8.604 2315 2516 19,937 21,669
4 n-Butane C4H10 58.12 31.10 51.6 0.3867 165.65 2.007 0.153 6.528 3011 3262 19,679 21,321
5 Isobutane C4H10 58.12 10.90 72.2 0.3872 157.53 2.007 0.153 6.528 3000 3252 19,629 21,271
6 n-Pentane C5H12 72.15 96.92 15.57 0.3883 153.59 2.491 0.190 5.259 3707 4009 19,507 21,095
7 Isopentane C5H12 72.15 82.12 20.44 0.3827 147.13 2.491 0.190 5.259 3699 4001 19,459 21,047
8 Neopentane C5H12 72.15 49.10 35.9 0.3866 135.58 2.491 0.190 5.259 3683 3985 19,390 20,978
9 n-Hexane C6H14 86.18 155.72 4.956 0.3864 143.95 2.975 0.227 4.403 4404 4756 19,415 20,966
Napthene (cycloalkane) series (CnH2n)

10 Cyclopentane C5H10 70.13 120.60 9.917 0.2712 137.35 2.420 0.180 5.556 3513 3765 19,005 20,368
11 Cyclohexane C6H12 84.16 177.40 3.267 0.2901 153.25 2.910 0.220 5.545 4180 4482 18,849 20,211
Olefin series (CnH2n)

12 Ethene C2H4 28.05 −154.62 — — 207.57 0.969 0.074 13.525 1499 1600 20,275 21,636
(Ethylene)
13 Propene C3H6 42.08 −53.90 226.4 0.3541 188.18 1.453 0.111 9.017 2182 2333 19,687 21,048
(Propylene)
14 1-Butene C4H8 56.11 20.75 63.05 0.3548 167.94 1.937 0.148 6.762 2879 3080 19,493 20,854
(Butylene)
15 Isobutene C4H8 56.11 19.59 63.4 0.3701 169.48 1.937 0.148 6.762 2860 3068 19,376 20,737
16 1-Pentene C5H10 70.13 85.93 19.115 0.3635 154.46 2.421 0.185 5.410 3575 3827 19,359 20,720
Aromatic series (C0H2n-6)

17 Benzene C6H8 78.11 176.17 3.224 0.2429 169.31 2.697 0.206 4.857 3591 3742 17,451 18,184
18 Toluene C7H8 92.14 231.13 1.032 0.2598 154.84 3.181 0.243 4.118 4274 4475 17,672 18,501
19 o-Xylene C8H10 106.17 291.97 0.264 0.2914 149.1 3.665 0.280 3.574 4958 5210 17,734 18,633
20 m-Xylene C8H10 106.17 282.41 0.326 0.2782 147.2 3.665 0.280 3.574 4956 5208 17,734 18,633
21 p-Xylene C8H10 106.17 281.05 0.342 0.2769 144.52 3.665 0.280 3.574 4957 5209 17,734 18,633
Additional fuel gas components

22 Acetylene C2H2 26.04 −119 — 0.3966 — 0.899 0.069 14.572 1424 1474 20,769 21,502
23 Methyl CH3OH 32.04 148.1 4.63 0.3231 473 1.106 0.084 11.841 766 867 9,066 10,258
alcohol
24 Ethyl alcohol C2H5OH 46.07 172.92 2.3 0.3323 367 1.590 0.121 8.236 1448 1599 11,918 13,161
25 Ammonia NH3 17.03 −28.2 212 0.5002 587.2 0.588 0.045 22.279 359 434 7,966 9,667
26 Hydrogen H2 2.02 −423.0 — 3.4080 193.9 0.070 0.005 188.217 274 324 51,625 61,095
27 Oxygen O2 32.00 −297.4 — 0.2186 91.6 1.105 0.084 11.858 — — — —
28 Nitrogen N2 29.16 −320.4 — 0.2482 87.8 0.972 0.074 13.472 — — — —
29 Carbon CO 28.01 −313.6 — 0.2484 92.7 0.967 0.074 13.546 321 321 4,347 4,347
monoxide
30 Carbon CO2 44.01 −109.3 — 0.1991 238.2 1.519 0.116 8.621 — — — —
dioxide
31 Hydrogen H2S 34.08 −76.6 394.0 0.2380 235.6 1.177 0.090 11.133 587 637 6,537 7,097
sulfide
32 Sulfur dioxide SO2 64.06 14.0 88 0.1450 166.7 2.212 0.169 5.923 — — — —
33 Water vapor H2O 18.02 212.0 0.9492 0.4446 970.3 0.622 0.047 21.061 — — — —
34 Air — 28.96 −317.6 — 0.2400 92 1.000 0.076 13.099 — — — —
Fuels 81
TABLE 4.8 (continued)

Physical Constants of Typical Gaseous Fuel Mixture Components
Unit Volume per Unit Volume of Combustible Unit Mass per Unit Mass of Combustible
Required for Required for Flammability

Stoichiometric Stoichiometric Stoichiometric Stoichiometric Limits (vol.%
Combustion Flue Gas Products Combustion Flue Gas Products in Air Mixture)
Fuel Gas
No. Component O2 N2 Air CO2 H2O N2 SO2 O2 N2 Air CO2 H2O N2 SO2 Lower Upper No.
Paraffin (alkane) series (CnH2n+2)

1 Methane 2.0 7.547 9.547 1.0 2.0 7.547 — 3.989 13.246 17.235 2.743 2.246 13.246 — 5.0 15.0 1
2 Ethane 3.5 13.206 16.706 2.0 3.0 13.206 — 3.724 12.367 16.092 2.927 1.797 12.367 — 2.9 13.0 2
3 Propane 5.0 18.866 23.866 3.0 4.0 18.866 — 3.628 12.047 15.676 2.994 1.624 12.047 — 2.0 9.5 3
4 n-Butane 6.5 24.526 31.026 4.0 5.0 24.526 — 3.578 11.882 15.460 3.029 1.550 11.882 — 1.5 9.0 4
5 Isobutane 6.5 24.526 31.026 4.0 5.0 24.526 — 3.578 11.882 15.460 3.029 1.550 11.882 — 1.8 8.5 5
6 n-Pentane 8.0 30.186 38.186 5.0 6.0 30.186 — 3.548 11.781 15.329 3.050 1.498 11.781 — 1.4 8.3 6
7 Isopentane 8.0 30.186 38.186 5.0 6.0 30.186 — 3.548 11.781 15.329 3.050 1.498 11.781 — 1.3 8.3 7
8 Neopentane 8.0 30.186 38.186 5.0 6.0 30.186 — 3.548 11.781 15.329 3.050 1.498 11.781 — 1.3 8.3 8
9 n-Hexane 9.5 35.846 45.346 6.0 7.0 35.846 — 3.527 11.713 15.240 3.064 1.463 11.713 — 1.1 7.7 9
Naphthene (cycloalkane) series (CnH2n)

10 Cyclopentane 7.5 27.939 35.81 5.0 5.0 28.939 — 3.850 11.155 14.793 3.146 1.283 11.155 — — — 10
11 Cyclohexane 9.0 33.528 42.970 6.0 6.0 33.528 — 4.620 13.386 17.750 3.146 1.283 11.155 — 1.2 8.4 11
Olefin series (CnH2n)

12 Ethene 3.0 11.320 14.320 2.0 2.0 11.320 — 3.422 11.362 14.784 3.138 1.284 11.362 — 2.7 36 12
(Ethylene)
13 Propene 4.5 16.980 21.480 3.0 3.0 16.980 — 3.422 11.362 14.784 3.138 1.284 11.362 — 2.0 11.7 13
(Propylene)
14 1-Butene 6.0 22.640 28.640 4.0 4.0 22.640 — 3.422 11.362 14.784 3.138 1.284 11.362 — 1.6 10 14
(Butylene)
15 Isobutene 6.0 22.640 28.640 4.0 4.0 22.640 — 3.422 11.362 14.784 3.138 1.284 11.362 — 1.6 10 15
16 1-Pentene 7.5 28.300 35.800 5.0 5.0 28.300 — 3.422 11.362 14.784 3.138 1.284 11.362 — 1.3 10 16
Aromatic series (C0H2n-6)

17 Benzene 7.5 28.300 35.800 6.0 3.0 28.300 — 3.072 10.201 13.274 3.380 0.692 10.201 — 1.2 8.0 17
18 Toluene 9.0 33.959 42.959 7.0 4.0 33.959 — 3.125 10.378 13.504 3.343 0.782 10.378 — 1.2 7.1 18
19 o-Xylene 10.5 39.619 50.129 8.0 5.0 39.619 — 3.164 10.508 13.673 3.316 0.848 10.508 — 1.0 7.6 19
20 m-Xylene 10.5 39.619 50.129 8.0 5.0 39.619 — 3.164 10.508 13.673 3.316 0.848 10.508 — 1.0 7.0 20
21 p-Xylene 10.5 39.619 50.129 8.0 5.0 39.619 — 3.164 10.508 13.673 3.316 0.848 10.508 — 1.0 7.0 21
Additional fuel gas components

22 Acetylene 2.5 9.433 11.933 2.0 1.0 9.433 — 3.072 10.201 13.274 3.380 0.692 10.201 — 2.5 80 22
23 Methyl 1.5 5.660 7.160 1.0 2.0 5.660 — 4.498 4.974 6.482 1.373 1.124 4.974 — 5.5 44 23
alcohol
24 Ethyl alcohol 3.0 11.320 14.320 2.0 3.0 11.320 — 2.084 6.919 9.003 1.911 1.173 6.919 — 3.3 19 24
25 Ammonia 0.75 2.830 3.582 — 1.5 3.330 — 1.409 4.679 6.008 1.587 5.502 — 15.5 27 25
26 Hydrogen 0.5 1.887 2.387 — 1.0 1.887 — 7.936 26.323 34.290 8.937 26.353 — 4.0 74.2 26
27 Oxygen — — — — — — — — — — — — — — — — 27
28 Nitrogen — — — — — — — — — — — — — — — — 28
29 Carbon 0.5 1.877 2.387 1.0 — 1.887 — 1.410 1.897 2.468 1.571 — 1.870 — 12.5 74.2 29
monoxide
30 Carbon — — — — — — — — — — — — — — — — 30
dioxide
31 Hydrogen 1.5 5.660 7.160 — 1.0 5.660 1.0 4.682 6.093 0.529 4.682 1.880 4.3 45.5 31
sulfide
32 Sulfur dioxide — — — — — — — — — — — — — — — — 32
33 Water vapor — — — — — — — — — — — — — — — — 33
34 Air — — — — — — — — — — — — — — — — 34
The table presented here gives the number of compound and the composition of the gases in terms of
moles/gram of each gas and their normalized vol- mole or mass fractions. Having assembled this informa-
ume, calculated from the mole fraction of each gas tion, the following formulae are used to calculate heat-
to the total moles of mix gas. ing values:
Gas Mass Percent (%) mol/g

Normalized
Volume Basis (%)
HVv = ∑ HV
v,i × yi (4.4)
H2O 10 0.00555 16.11

CO2
N2
25
65
0.00568
0.0232
16.49
67.36
HVm = ∑ HV
m ,i × xi (4.5)
where
Example 4.2 HVv is the heating value of mixture, volume basis
HVv,i is the heating value of component i, volume basis
Find the density of flue gas at 350°F (180°C) and HVm is the heating value of mixture, mass basis
2 atm. The composition of the flue gas by volume
HVm,i is the heating value of component i, mass basis
is 17% H2O, 6% CO2, and 77% N2.
To calculate the density of flue gas, one should
take the density of air at standard temperature Example 4.3
and pressure and molecular weights of flue gas
and dry air. The molecular weight of flue gas can Determine the HHV of a gas mixture comprised
be found by taking the volume percentages and of 60% natural gas and 40% H2 in Btu/ft3.
molecular weights of each gas. For the gas mixture, one should take 1 ft3 of
the gas mix coupled with the HHV of each indi-
vidual gas.
mole H 2 O g mole CO 2 g
MWFG = 0.17 × 18 + 0.06 × 44 The HHVs are presented as follows:
mole FG mol mole FG mol
CH4 = 1012 Btu/ft3
mole N 2 g g
+ 0.77 × 28 = 27.26 H2 = 325 Btu/ft3
mole FG mol mol
ft 3 CH 4 Btu ft 3 H 2 Btu Btu
For the molecular weight of air, it will be com- 0.6 × 1012 3 + 0.4 × 325 3 = 737.2 3
posed of 21% dry O2 and 79% N2, thus resulting in ft mix
3
ft ft3 ft ft
 g   g  g Example 4.4
MWAir = 0.21  32 + 0.79  28 = 28.84
 mol   mol  mol
Determine the HHV of a gas mixture comprised
of 30% natural gas, 50% H2, and 20% N2. Calculate
Using the temperature of air at STP, one can solve heating values based on the mass of gases.
for the density of flue gas at the given tempera- Similar to the earlier example, one should take
ture and pressure: 1 lbm of mix gas with the heating values to get the
mixture HHV. The HHV of each gas based on
mass will be
 2   27.26   288  kg
ρFG = 1.2    = 1.46 3
 1   28.84   273 + 176  m CH4: 23,875 Btu/lbm
H2: 61,095 Btu/lbm
4.2.2 Lower and Higher Heating Values lbm CH 4 Btu lb H Btu

0.3 × 23, 875 + 0.5 m 2 × 61, 095
lbm mix lbm lbm mix lbm
The LHV of a gas is the heat released by combustion of
a specific quantity of that gas with the products of com- Btu
= 37 , 710
bustion remaining as vapor. The higher heating value lbm
(HHV) adds to the LHV the latent heat of any steam pro-
duced as a combustion product. It represents the total
4.2.3 Specific Heat Capacity
heat obtained by first burning a fuel and then cooling
the products to standard temperature. Heating values The specific heat capacity of a gas is the energy that must
may be provided on a volume basis, typically Btu/scf be added to a specific amount of the gas to raise its tem-
(kJ/m3), or a mass basis such as Btu/lbm (kJ/kg). perature by 1°. If the gas is maintained at constant pres-
To determine the heating value of a mixture of sure during this heating process, the value is referred
gases, it is necessary to know the heating value of each to as cp. If the gas is maintained at constant volume, the
Fuels 83
value is referred to as cv. Specific heat is not a constant The specific heat of the gas mix coupled with the
for a given gas; it is a function of temperature. Specific molecular weight can be used to solve for the spe-
heat can be defined on a volume basis, typically Btu/ cific heat on a mass basis:
lbmol-°F (kJ/kgmol-°C); or on a mass basis such as Btu/ Btu 1 lbmol Btu
lbm-°F (kJ/kg-°C). 7.24 × = 0.265
lbmol °F 27.26 lbm lbm °F
To determine the specific heat of a mixture of gases, it
is necessary to know the specific heat of each compound
at the mixture temperature and the composition of the 4.2.4 Flammability Limits
gases in terms of mole or mass fractions. Having assem- Flammability limits define the range of fuel concentrations
bled this information, the following formulae are used to in air that will sustain a flame without additional air or
calculate specific heat (cp and cv formulae are analogous, fuel. The upper flammability limit (UFL) is the maximum
only cp formulae are shown): fuel concentration that can sustain a flame and the lower
flammability limit (LFL) is the minimum. These limits are
cp ( vol) = ∑c p,i ( vol) × yi (4.6) often tabulated for fuels at some standard temperature,
typically 60°F (16°C). Flammability limits are not constants
for a given gas; they are functions of the air/fuel mixture
cp (mass) = ∑c p,i (mass) × xi (4.7) temperature. An extensive discussion of this subject can be
found in the work of Coward and Jones.15
Wierzba and Karim16 present a method for estimating
where
the flammability limits as a function of mixture temper-
cp (vol) is the specific heat of mixture, volume basis
ature by calculating adiabatic flame temperature (AFT).
cp,i (vol) is the specific heat of component i, volume
First, the AFT for the standard temperature mixture is
basis
determined. Next, the mixture temperature is set to the
cp (mass) is the specific heat of mixture, mass basis
desired level and the fuel concentration is varied until
cp,i (mass) is the specific heat of component i, mass
the calculated AFT for the nonstandard temperature
basis
matches the AFT for the standard temperature. They
provide an approximating method for calculating AFT
Example 4.5 for sub-stoichiometric mixtures.
Both Coward and Jones and Wierzba and Karim indi-
Calculate the specific heat of a gas mixture on a
mole basis or mass fraction. The mixture consists
cate that a form of Le Chatelier’s rule can be used to
of 17% H2O, 6% CO2, and 77% N2. calculate LFL and UFL for many combinations of fuels
and inerts. Both references also mention that this rule
fails to accurately predict for a few important situa-
Btu/ Btu/ cal/ cal/ tions. One notable example is a mixture of ethylene and
Gas MW lbmol° F lbm · °F g · °C g · mole · °C carbon dioxide that differs substantially from normal
N2 28 6.94 0.248 0.248 6.94 calculated LFL and UFL. Another example is any mix-
H2O 18 8.02 0.445 0.445 8.02 ture of chemicals that is prone to react with another at
CO2 44 8.80 0.200 0.200 8.80 temperatures below the ignition point, such as ethylene
and hydrogen. Mixtures involving significant amounts
of inert compounds (e.g., H2O, N2, and CO2) require
The mix gas-specific heat, based on percentage special treatment either by the AFT method described
values, is calculated to be earlier or by grouping the inerts with fuel components
in known proportions matching conditions for which
moleH 2 O Btu mole CO 2 Btu LFL and UFL have been measured. This latter method
0.17 × 8.02 + 0.06 × 8.80
moleMix mole °F moleMix mole °F is described in detail by Coward and Jones.14
With these exceptions in mind, the following mixing
mole N 2 Btu Btu
+ 0.77 × 6.94 = 7.24 rules can be used to calculate LFL and UFL for most
moleMix mole °F lbmol °F
common gas mixtures:
100
Taking the molecular weight of the gas mixture LFL = (4.8)
will yield

∑ yi/LFLi
100
lbm UFL = (4.9)
∑ y /UFL
MWmix = 0.17(18) + 0.06 ( 44) + 0.77(28) = 27.26
lbmol i i
4.2.5 Burning Velocity 4.2.6 Viscosity

Flame speed and burning velocity are terms that are Viscosity is discussed in detail in Chapter 8. A useful
often used for the rate of propagation of a flame through mixing rule is also provided.
a flammable mixture. The burning velocity is depen-
dent not only on the fuel gas composition, gas/air mix-
ture ratio and temperature of the mixture, but also on 4.2.7 Derived Quantities
the physical dimensions of the combustion device and
In addition to the specific properties described earlier,
the interaction of the flame with heat absorbing sur-
there are a number of useful derived parameters that
faces. Burning velocity is thus not an absolute value, but
may be of interest when studying combustion systems.
is measurement device dependent. Quoted values for
flame speed or burning velocity should include infor-
mation regarding the method of measurement. Since
4.2.7.1 Partial Pressure
the measured burning velocity is not an absolute value,
comparisons of burning velocities for various fuels Partial pressure is the pressure exerted by a single
must be made on the basis of data collected on identical component of a mixture when that component alone
test instruments. occupies the entire volume at the mixture temperature.
Figure 4.3 shows burning velocity data for a variety Dalton’s law states that the total pressure of a mixture
of fuel gases. Note that the maximum velocity generally is the sum of the partial pressures of the components.
occurs in a fuel-rich mixture with less than 100% theo- While this law has been demonstrated to be somewhat
retical combustion air in the fuel/air mixture.8 in error, especially at high pressures, it is often useful
Blue water gas

3.6 9.0
Hydrogen
3.2 8.0
Burning velocity, ft/s (except hydrogen and acetylene)
Burning velocity, ft/s (hydrogen and acetylene only)

2.8 7.0
2.4 6.0
Ethylene
Acetylene
2.0 Carbureted 5.0

water gas
1.6 4.0
Coke oven gas
Carbon Illuminants
Butane
1.2 monoxide 3.0
Ethane
0.8 2.0
Producer
gas
0.4 Natural gas

Methane 1.0
Propane
0 0
0 20 40 60 80 100 120 140
Primary air, % of theoretical
FIGURE 4.3
Burning velocity for various gaseous fuels.
Fuels 85
for estimating purposes to determine whether a more mixtures that also require enrichment and supplemen-
detailed analysis is justified. The basic relationship is tal fuel firing to maintain ignition. Fertilizer plants and
other chemical plants produce ammonia, which may
pi = yi ↔ TP (4.10) be sent to a flare in an emergency. Waste gases that are
sent to flares in these facilities may be pure ammonia or
Partial pressures are of interest when estimating the diluted with nitrogen or water vapor.
probability of forming condensate in a gas mixture. The variety of gases and the hazards associated with
When the partial pressure of a component exceeds the each requires careful review of all aspects of system
vapor pressure of that component at the mixture tem- design to ensure that these fuels are safely handled,
perature, condensation is likely. whether in a flare, a furnace, or an incinerator.
4.2.7.2 Adiabatic Flame Temperature

The AFT is the temperature at which the enthalpy of the
products of combustion equals the sum of the enthalpy
4.3 Liquid Fuels
of the reactants plus the heat released by the combus-
tion process. Heat loss due to radiation, convection, or 4.3.1 Production, Refining, and Chemistry
conduction is not included; hence the reference to adia-
The primary concern for a typical refinery is to convert
batic. Accounting for dissociation of combustion prod-
a barrel of crude oil (42 U.S. gallons) into usable prod-
ucts is important. Customarily, the AFT is determined
ucts. A barrel of crude oil can typically be refined to
for a stoichiometric fuel/air mixture, although other
provide 11 gal of gasoline, 5.3 gal of kerosene, 20.4 gal of
mixtures such as LFL and UFL are sometimes studied
gas-oil and distillates, and 5.3 gal of heavier distillates.9
for special purposes, as discussed in Section 4.3.2.
The end products derived from crude oil number in the
thousands. Table 4.9 provides a listing of many of these
4.2.7.3 Other Special Cases products. The processes that produce these different
products are vast and complicated. Figure 4.4 provides a
Landfills and digester facilities produce an off-gas that
general refinery flow diagram.
must be disposed to prevent odor problems in the com-
munity. The gas is generally a mixture of CO2 and CH4. TABLE 4.9
Landfills are rarely above 30%–40% methane, while
Quantitative Listing of Products Made by the U.S.
digesters may be as high as 60%–70% methane. In some Petroleum Industry
landfills, perimeter wells are used to draw air into the
edges of the landfill, which prevents the spread of anaer- Product Classification Number of Individual Products
obic bacteria and methane. In these cases, the methane Lubricating oils 1156
content is even lower and some air is also sent to the flare. Chemicals, solvents, misc. 300
Marine and truck loading facilities burn the vapor Greases 271
displaced from the tankers or trucks during the loading Asphalts 209
operation. In many cases, the displaced vapor is mostly air Waxes 113
with some amount of evaporated gasoline or diesel fuel. White oils 100
Depending on the ambient temperature, the resulting mix- Rust preventatives 65
ture could be very rich in hydrocarbon vapor, or very lean. Diesel and light fuel oils 27
Medical equipment, such as bandages or hypodermic Motor gasolines 19
needles, is often sterilized by contact with ethylene oxide Residual fuel oil 16
(ETO) vapors. ETO sterilizer flares are designed to receive Liquified gases 13
the ETO vapor after the sterilization process is complete. Other gasolines 12
The composition coming to these flares generally con- Transformer and cable oils 12
sists of a mixture of ETO and either air or nitrogen. It Kerosenes 10
Aviation gasolines 9
should be noted that ETO has a flammability range from
Jet fuels 5
3% to 100% and a very low ignition temperature.
Carbon blacks 5
Flares are often used as backup equipment for inciner-
Cokes 4
ators during maintenance or malfunctions. In this type
Fuel gas 1
of service, the waste gas is usually enriched with a sub-
stantial amount of clean fuel gas to ensure reliable burn- Total 2347
ing. Steel mills produce off-gases that consist mainly of Source: Gary, J.H. and Handwerk, G.E., Petroleum Refining, 3rd edn.,
H2, H2O, CO, CO2, and air. These are generally low LHV Marcel Dekker, New York, 1994.
Crude
Typical refinery
products
LPG and gas Vapor Refinery fuel

Crude
recovery gas
unit
Stabilizer Straight run gasoline

section LPG
Regular
gasoline
Naphtha Hydro Reformate

Crude Reforming Premium
treating
unit gasoline
ATMOS
TWR Middle distillates Hydro Solvents
section treating
Aviation fuels
HVY atmos gas oil Catalytic Gasoline

cracking Diesels
VAC gas oil
Crude Heating oils

Fraction bottoms
unit
Treatingandblending
Lube oils
Vacuum Solvent
section extraction
Greases
Gasoline, naphtha,
Lube base stocks
Hydro and middle distillates
Asphalts
cracking
Lube oils Industrial

Solvent fuels
Propane dewaxing Waxes
deasphalter
Refinery
fuel oil
Gasoline, naphtha, and middle distillates
Visbreaker
Fuel oil
Asphalt
Delayed Gasoline, naphtha, and middle distillates

coker
Coke
FIGURE 4.4
Refinery flow diagram. (Adapted from Gary, J.H. and Handwerk, G.E., Petroleum Refining, 3rd edn., Marcel Dekker, New York, 1994.)
Fuels 87
The primary chemical components of crude oil are constituents of crude oils are alkanes (paraffins), cyclo-
carbon, hydrogen, sulfur, oxygen, and nitrogen. The alkanes (naphthenes), and aromatics.
percentages of these elements found in a crude oil are Alkanes (also called paraffins after the Latin parum
most frequently used to characterize the oil. Two terms affinis, “little affinity”) are those chemical structures
frequently used when referring to crude oil are “sweet” that are based on carbon atoms having only single bonds
crude and “sour” crude. Sweet crude is oil that contains and that are completely saturated with hydrogen atoms.
less than 0.5 wt% sulfur, while sour crude contains greater Some of the alkane hydrocarbons are listed in Table 4.8.
than 0.5 wt% sulfur. Sulfur content is of importance and The basic chemical formula for an alkane is CxH2x+2,
concern, due to the sulfur oxides that are produced during where “x” is the number of carbon atoms present. Crude
combustion. SO2, for example, is a gas that has been shown oils can contain structures with up to 70 carbon atoms.9
to contribute significantly to several different environ- However, the vast majority of the compounds contain
mental problems—namely, in acid rain formation and in 40 carbon atoms or less. When the number of different
its ready conversion to sulfuric acid, H2SO4. The nitrogen constitutional isomers (different chemical connectivity
content of crude oil is of special interest to the combustion and different physical properties, yet identical chemi-
industry due to the high levels of nitrogen oxides or NOx cal formulae) is considered (tetracontane [C40H82] has
(see Chapter 10) produced during combustion of these over 62 trillion possible isomers18), it is evident that the
fuels (e.g., approximately 0.2 lb/MMBtu NOx or 142 ppm compositional diversity between differing crude oils is
can be attributed to “Fuel NOx” for an oil that contains almost limitless.
0.47 wt% nitrogen). Like SOx, NOx is an environmentally Cycloalkanes (cylcoparaffins or naphthenes) are
damaging group of gases. Any time a fuel is burned in alkanes in which all or some of the carbon atoms are
air with a hot flame, NOx gases are produced. The greater arranged in a ring. When a cycloalkane contains only
the flame temperature of the combustion, the greater the one ring, the general formula is CxH2x. The most sta-
amount of NO that will be produced. NO is then oxidized ble cycloalkane is cyclohexane, while cyclobutane and
to form NO2 (over a period of minutes or hours), which is a cyclopropane are the least stable. The properties of
major contributor to photochemical smog. In general, the cycloalkanes are very similar to those of alkanes, as
fate of SO2 and NO are intertwined, as can be seen by the shown in Table 4.8.
following reaction sequence17: Aromatic compounds are those compounds that con-
tain at least one benzene-like ring. Benzene, discovered
SO2 + OH− → HSO 3− in 1825, has a chemical formula of C6H6, and is stable
and nonreactive relative to alkanes and cycloalkanes.
HSO 3− + O 2 → SO 3 + HOO − Aromatics, such as the heterocyclic compounds pyridine
and furan, are composed of rings that contain elements
SO 3 + H 2O → H 2SO 4 (g ) other than carbon. For example, the benzene ring contains
six carbon atoms, whereas the pyridine ring contains five
H 2SO 4 (g ) H
2O
→ H 2SO 4 (aq ) carbon atoms and one nitrogen atom. Properties of some
of the aromatic compounds are contained in Table 4.8.
The following is a parallel reaction that takes place It is worth mentioning the group of compounds called
between nitrogen oxide and the hydroperoxy radical, alkenes (olefins). Alkene compounds do not occur natu-
thus producing more of the hydroxyl radical to feed the rally in crude oil, but are produced by reaction during
aforementioned initial reaction: the refining process. Therefore, it should be expected
that a refined end product will have some percentage of
ethylene, propylene, or butylene, for example. Alkenes
HOO − + NO − → OH − + NO 2−
have the general formula of CxH2x and contain a carbon–
carbon double bond. Properties of some of the alkenes
The overall reaction is then are contained in Table 4.8.
When a crude oil is refined, the first step is, invariably,
distillation. The purpose of distillation is to separate
SO 2 + NO − + O 2 H
2O
→ NO −2 + H 2SO 4 (aq)
lighter components from heavier ones, based on their
respective volatility. The target of distillation is to sepa-
Crude oil compositions are relatively constant. rate the crude oil into different fractions. Each fraction
However, slight deviations in composition can result in consists of a boiling point range that will yield a mixture
vastly different refining methods. Crude oils also con- of hydrocarbons; see Table 4.10. Some of these mixtures
tain inorganic elements such as vanadium, nickel, and can then be used as product (fuels, solvents, etc.) or fur-
sodium, and usually contain some amount of water and ther refined into gasoline or other desirable mixtures.
ash (noncombustible material). The main hydrocarbon Catalytic cracking is a typical process used to break
TABLE 4.10 to gas flames, while heavier oils require a more compli-
General Fraction Boiling Points cated process and produce flames that are quite radiant
and more highly dependent on atomization techniques
Distillation Fraction Temperature Range
than the light oils. Oils are fired in burners by them-
Butanes and lighter <90°F selves, or in combination with fuel gas, waste gas, or
Gasoline 90°F–220°F both. Naphtha is frequently fired in combination with a
Naphtha 220°F–315°F PSA or other waste gas, and requires good vaporization
Kerosene 315°F–450°F to provide a quality flame.
Fuel oils 450°F–800°F
Residue >800°F 4.3.2 Oils
Source: Leffler, W.L., Petroleum Refining for the Non- According to the American Standard Testing Methods
Technical Person, Penn Well Publishing, Tulsa,
OK, 1985. With permission.
(ASTM) D-396, fuel oils are divided into grades, based
on the types of burners for which they are suitable.20 The
grades are determined by those values determined to be
down and rearrange alkane mixtures produced via dis-
most significant in figuring performance characteristics.
tillation into smaller, highly branched alkanes by heat-
The two classifications that separate these fuel oils are “dis-
ing the mixtures to high temperatures in the presence of
tillates” and “residuals,” where distillates indicate a distil-
a variety of catalysts. Figure 4.5 shows a fluid catalytic
lation overhead product (lighter oils) and residuals indicate
cracking process.19 Due to the reactions that take place
a distillation bottom product (heavier oils). Table 4.11 helps
during catalytic cracking, the product streams are gen-
in differentiating between these various classifications, and
erally heavier than the feed streams. Alkanes that are
Table 4.12 reveals typical analyses for these oils.
more highly branched are desirable because they have
a higher octane rating than their unbranched cousins.
4.3.2.1 Light Oils
Of particular interest are the liquid fuels produced
during the various refining processes that are used Grade 1 and 2 oils are light distillate (fuel) oils used
by the hydrocarbon and petrochemical industries. primarily in applications that do not require significant
Refineries frequently burn these liquid fuels in process atomization by air or steam in order to reduce droplet
heaters so that the heat liberated during combustion can size for proper burning. No. 1 oil will typically vapor-
be used to drive a more profitable process. Light fuel oils ize when it comes into contact with a hot surface or gas.
are relatively easy to burn and produce flames similar No. 2 oil is significantly easier to burn than residual oil
Gas-
concentration
unit
Fuel gas
To C3C4 to treating
flue gas
system
Debutanized gasoline
Reactor
Naphtha
Main column
Regenerator
Light cycle oil
Heavy cycle oil
Main-column bottoms
Heavy recycle
(optional)
Air Fresh feed
FIGURE 4.5
Flow diagram of UOP fluid catalytic cracking complex. (From Meyers, R.A., Handbook of Petroleum Refining Processes, 2nd edn., McGraw-Hill,
New York, Chapter 3.3, 1997.)
Fuels 89
TABLE 4.11
Requirements for Fuel Oils (per ASTM D 396)
No. 4 Distillate
Classification No. 1 Distillate No. 2 Distillate (Heavy) No. 6 Residual
Density (kg/m ) @ 60°F (15°C), max
3 850 876 — —
Viscosity @ 104°F (40°C) mm/s2
Min 1.3 1.9 >5.5 —
Max 2.1 3.4 24 —
Viscosity @ 212°F (100°C) mm/s2
Min — — — 15
Max — — — 50
Flash point °F (°C), min 100 (38) 100 (38) 131 (55) 140 (60)
Pour point °F (°C), max −0.4 (−18) 21 (−6) 21 (−6) —
Ash, % mass, max — — 0.1 —
Sulfur, % mass, max 0.5 0.5 — —
Water and sediment, % vol., max 0.05 0.05 0.5 2.0
Distillation temperature °F (°C)
10% volume recovered, max 419 (215) — — —
90% volume recovered, min — 540 (282) — —
90% volume recovered, max 550 (288) 640 (338) — —
TABLE 4.12
Typical Analysis of Different Fuel Oils
No. 1 No. 2 No. 4 No. 6 Fuel
Fuel Oil Fuel Oil Fuel Oil Oil (Sour)
Ash (%) <0.01 <0.01 0.02 0.05
Hydrogen (%) 13.6 13.6 11.7 11.2
Nitrogen (%) 0.003 0.007 0.24 0.37
Sulfur (%) 0.09 0.1 1.35 2.1
Carbon (%) 86.4 86.6 86.5 85.7
Heat of combustion 20,187 19,639 19,382 18,343
(HHV), Btu/lb
Specific gravity 0.825 0.84 0.898 0.97
60°F/60°F
Density (lb/U.S. 6.877 6.96 7.488 8.08
gal)
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
Press, Boca Raton, FL, 2001.
due to the reduced amount of atomization and preheat-

ing requirements. Atomization is the breaking apart of a FIGURE 4.6
liquid into tiny, more easily combustible, droplets using Burner firing heavy oil no. 1. (From Baukal, C.E. (ed.), The John Zink
steam, air, fuel gas, or mechanical means. These light Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
distillate oils will typically distill out between 450°F
and 800°F (230°C and 430°C).
4.3.2.3 Residual Oils
4.3.2.2 Heavy Oils
No. 6 oil is a heavy residual oil sometimes referred to as
No. 4 oil is a heavy distillate oil typically blended from, Bunker C oil. This oil requires significant atomization for
and thus having characteristics of, both light distillates proper combustion. Due to its high viscosity, No. 6 oil
and residual oils. These oils do not readily combust and requires heating during handling and further heating prior
therefore require some type of atomization, but still fall to combustion chamber injection. No. 6 oil is usually pre-
into a viscosity range that does not require preheating heated to 150°F–200°F (66°C–93°C), to decrease its viscos-
prior to burning (Figures 4.6 and 4.7). ity, before being atomized and injected into the burner.
TABLE 4.13
Naphtha Elemental Analysis
Component Vol%
n-Heptane 1.610
Methylcyclohexane 2.433
2-Methylheptane 5.618
1c,3-Dimethylcyclohexane 3.252
1t,4-Dimethylcyclohexane 1.040
1t,2-Dimethylcyclohexane 1.169
n-Octane 16.334
1c,2-Dimethylcyclohexane 1.674
1,1,4-Trimethylcyclohexane 3.500
2,6-Dimethylheptane 2.094
1c,3c,5-Trimethylcyclohexane 2.638
m-Xylene 2.426
p-Xylene 0.797
2,3-Dimethylheptane 1.475
4-Methyloctane 3.417
o-Xylene 1.137
n-Nonane 10.120
Other 23.534
Total 100.000
320
FIGURE 4.7
Burner firing heavy oil no. 2. (From Baukal, C.E. (ed.), The John Zink 310
300
Temperature,°F
4.3.3 Liquid Naphtha 290
Liquid naphtha is similar in its characteristics to kerosene 280

(Table 4.13). Figure 4.8 shows a typical naphtha distilla-
270
tion curve. In general, naphtha will boil out of a mixture
between 220°F and 320°F (100°C and 160°C). Naphtha is 260
categorized, based on its volatility, into light, intermedi-
250
ate, and heavy naphtha. Naphtha is a major constituent
of gasoline; however, it generally requires further refin- 240
ing to make suitable quality gasoline. Prior to firing naph- 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Cumulative percent volume
tha in a burner, care must be taken to vaporize it so that
the combustion will be more complete and uniform.
FIGURE 4.8
Naphtha distillation curve. (From Baukal, C.E. (ed.), The John Zink
4.3.4 Physical Properties of Liquid Fuels Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
When liquid fuels are encountered, there are certain
properties that determine into which category they are that will ignite when exposed to an ignition source.
divided, and for what processes they are suitable. The standard method for determining flash point is
ASTM D-93. Under certain conditions, ASTM D-56 can
be used for light distillate oils. Some flash point values
4.3.4.1 Flash Point
are provided in Table 4.11. The flash point is an impor-
The flash point of a liquid is the lowest temperature at tant property for indication of volatility and for storage
which enough vapors are given off to form a mixture requirements.
Fuels 91
4.3.4.2 Pour Point When the Reynolds number is less than 2100, the flow
is typically streamlined and smooth, and called lami-
The pour point of a liquid is determined by ASTM D-99
nar. However, when the Reynolds number increases
and indicates the lowest temperature at which an oil
above 2100, internal agitation takes place, and the flow
will flow at a controlled rate. If the fluid temperature
is considered turbulent. As seen in the Equation 4.11, as
goes below this point, flow will be inhibited.
the viscosity increases, the flow becomes more laminar,
assuming the other properties stay constant. Viscosity is
4.3.4.3 Distillation divided into two different categories: kinematic viscos-
ity and absolute viscosity.
The distillation of a liquid gives an indication of its vola-
Kinematic viscosity (ν) is dependent on fluid den-
tility, as well as the ease with which it can be vapor-
sity, and has units of length 2/time. Typical units for
ized. The test evaluates the vaporization range of a fuel
kinematic viscosity are stokes (0.001 m 2/s), centistokes
between its end point (the point at which 100% of the
(stoke/100), Seconds Saybolt Universal (SSU), and
volume has vaporized) and the initial boiling point (the
Seconds Saybolt Furol (SSF). Because the density of
point at which the liquid begins to vaporize). Figure 4.9
a fluid is dependent on temperature, the viscosity of a
shows a typical crude oil distillation curve.
fluid is likewise dependent on temperature. As the
temperature increases, the viscosity of a fluid will
4.3.4.4 Viscosity decrease (become more fluid or less viscous), and vice
versa.
In layman’s terms, the viscosity is a fluid’s resistance to
Absolute viscosity (μ) can be calculated by multi-
flow. Technically, the viscosity is the ratio of shear stress
plying the kinematic viscosity by the density of the
to shear rate of a fluid in motion. Most fluids under con-
fluid. The most common units for absolute viscosity
sideration in this chapter (gases, fuel oils) are Newtonian
are the poise (1 Pa s) and the centipoise (cp), which is
fluids because the ratio given earlier is constant with
poise/100.
respect to time, at a given temperature and pressure.
The viscosity of oil is a very important consideration
A very important factor in the determination of fluid
in proper burner design. As previously mentioned, the
flow is the dimensionless quantity called the Reynolds
more viscous the fluid, the more preheating required
number. The Reynolds number is calculated as
prior to burning. Several useful conversions are listed
DVρ DV as follows:
Re = or Re = (4.11)
µ ν
1 lbm/ft h = 0.00413 g/cm · s
where
D is the pipe diameter 0.000413 kg/m · s
V is the fluid velocity 1 cP = 0.01 P
ρ is the fluid density 0.01 g/cm · s
μ is the fluid absolute viscosity 0.001 kg/m · s
ν is the fluid kinematic viscosity
6.72 × 10−4 lbm/ft s
1000
900 Residue
1 St = 0.0001 m2/s = 100 cSt
800 centistokes = (0.266 × SSU) − (195/SSU) for SSU 32–100
Boiling temperature, °F
700 (0.220 × SSU) − (135/SSU) for SSU > 100

Gas-oil
600
See also Table 4.11 and Figure 4.10.
500
400 Kerosene
300 4.3.4.5 Density, Gravity, Specific Volume,
Naphtha
200 and Specific Weight
Gasoline
100 Density is a fluid’s mass per unit volume, and is impor-
Butanes and lighter
0 tant due to its effect on other properties, such as viscos-
0 10 20 30 40 50 60 70 80 90 100
ity. Additionally, the density is used to calculate the heat
Cumulative percent volume
capacity of an oil. The densities of liquids are frequently
FIGURE 4.9
given as the American Petroleum Institute (API) gravity
Crude oil distillation curve. (From Baukal, C.E. (ed.), The John Zink or the specific gravity (SG). Density and specific gravity
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) are related in that a liquid with a specific gravity of 1 has
–30 –20 –10 0 10 20 30 40 50 60 70 80 90 100 120 140 160 180 200

200,000
500,000
50,000
20,000 100,000
10,000 50,000
5,000 20,000
2,000 10,000
5,000
1,000
3,000
400 2,000
200 1,000
600
N
o.
100
6
75 400
300
Kinematic viscosity, centistokes
50
No
40 200
.5
he
No
30
av 100
.4
y
No
20
Viscosity, SSU
.5
15 80
lig
ht
10 60
8
50
No
6
.2
5
40
4
3
35
2
32
0.5
–20 0 20 40 60 80 100 120 140 160 180 200 240 280 320 360 400
Temperature, °F
FIGURE 4.10
Viscosity of fuel oils. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
a density of 1 kg/m3 (62.4 lb/ft3). The specific gravity of a Water is frequently used as a reference substance
liquid can be calculated by the formula and, at 60°F (16°C), has a specific gravity of 1.0 and a
ρ density of 1.94 slugs/ft3 (999 kg/m3), where 1 slug =
SG = (4.12) 1 lbf ft/s2. Specific gravity for gases requires an addi-
ρref
tional assumption relating to pressure and tempera-
where ture. Gas-specific gravity is defined relative to air as the
ρ is the density of the substance in question at specific reference substance and is generally determined at a
conditions standard temperature and pressure. Under those condi-
ρref is the density of a reference substance at a specific tions, gas-specific gravity can be calculated as the ratio
condition of molecular weights.
Fuels 93
Example 4.6 The specific volume (volume per unit mass) is the
In a typical combustion situation, determine reciprocal of the density, and is commonly used in ther-
the molecular weight of flue gas at 15% excess modynamic calculations.
air, the wet O2 percent, and dry O2 percent pres- The specific weight of a fluid (γ) is defined as its weight
ent in the flue gas. Assume natural gas as the per unit volume. The relationship that relates specific
working fluid. weight to the density is γ = ρ × g, where ρ is the density,
Following the general combustion equation, the and g is the local acceleration (32.174 ft/s2 or 9.81 m/s2).
balanced stoichiometric equation will be The specific weight of water at 60°F (16°C) is 62.4 lbm/ft3
(9.80 kN/m3).
CH 4 + 2φ [ O2 + 3.76 N2 ]
→ CO 2 + 2H 2 O + 2φ  O 2 + 3.76 N 2  − 2O 2 4.3.4.6 Heat Capacity (Specific Heat)

The heat capacity, or specific heat, of a fluid is defined as
where ϕ = 1 + EA%. the amount of heat that is required per unit mass to raise
The following table provides the calculated the temperature by 1°. Typical units of heat capacity are
byproducts per mole of CH4 Btu/(lbm · °R) or kJ/(kg · K) in SI units. Heat capacity is
temperature dependent, and is defined in terms of con-
stant volume or constant pressure, as can be calculated
Moles/ by the following equations:
Mole CH4
 δh 
Moles/Mole CH4 @ 15% EA Grams/Mole CH4
Cp =  (4.14)
CO2 1 1 1 × 44 = 44  δT  p
H2O 2 2 2 × 18 = 36
O2 2ϕ−2 0.3 0.3 × 32 = 9.6
 δh 
Cv =  (4.15)
 δT  v
N2 7.52ϕ 8.648 8.648 × 28 = 242.1
The molecular weight of flue gas will be the total where

mass of resultant products divided by the Cp is the heat capacity at constant pressure
total moles produced, as can be seen here: Cv is the heat capacity at constant volume
δh is the change in enthalpy
g FG δT is the change in temperature
MWFG = ( 44 + 36 + 9.6 + 242.1)
mole CH 4
To calculate the heat capacity of a petroleum liquid, to
1 mole CH 4 g within 2%–4% accuracy, the following equations can be
× = 27.66

(1 + 2 + 0 .3 + 8 .648 ) mol
l e FG mol employed:
0.388 + (0.00045 × °F )
The wet and dry O2 products formed will be C= for units of Btu/(lbm -°R )
the moles of O2 produced divided by the total SG
moles including and excluding moles of H2O, (4.16)
respectively:
1.685 + (0.039 × °C )
0.3 C= for units of kJ/(kg-K ) (4.17)
O 2 wet = × 100 = 2.51% SG
11.948
0.3 where
O 2dry = × 100 = 3.0%
11.948 − 2 C is the heat capacity
SG is the specific gravity (relative density), so long as
°API is inversely proportional to specific gravity; the liquid temperature is between 32°F and 400°F
therefore, as °API increases, the density decreases. (0°C and 205°C) and the specific gravity is between
When a fluid and water are compared at 60°F (16°C), the 0.75 and 0.96 at 60°F (16°C)8
°API can be calculated as
Further information about gaseous and liquid fuels
141.5 and their properties can be obtained from the references
°API = − 131.5 (4.13)
SG listed at the end of this chapter.21–24
12. Nelson, W.L., Petroleum Refining Engineering, 3rd edn.,

McGraw-Hill, New York, 1949.
References 13. Meyers, R.A., Handbook of Petroleum Refining Processes,
1. U.S. Energy Information Administration, Natural Gas 2nd edn., McGraw-Hill, New York, 1997, Chapter 6.2.
Data Report, Natural Gas Reserves Summary as of 14. Meyers, R.A., Handbook of Petroleum Refining Processes,
December 31, December 30, 2010. 2nd edn., McGraw-Hill, New York, 1997, Chapter 12.1.
2. U.S. Energy Information Administration, Total Energy 15. Coward, H.F. and Jones, G.W., Limits of Flammability of
Data Report, Annual Energy Review, October 19, 2011. Gases and Vapors, U.S. Bureau of Mines, Department of
3. Reed, R.J., North American Combustion Handbook, Vol. I, Interior, Bulletin 503, Pittsburgh, PA, 1952.
North American Mfg. Co., Cleveland, OH, 1986. 16. Wierzba, I. and Karim, G.A., Prediction of the flamma-
4. Gas Processors and Suppliers Association, GPSA bility limits of fuel mixtures, AFRC/JFRC International
Engineering Data Book, Vol. I, 10th edn., GPSA, Tulsa, OK, Symposium, Maui, HI, October 1998.
1987. 17. Baird, C., Environmental Chemistry, W.H. Freeman,
5. Austin, G.T., Shreve’s Chemical Process Industries, 5th edn., New York, 1995, Chapter 3.
McGraw-Hill, New York, 1984. 18. Solomons, T.W., Organic Chemistry, 5th edn., John Wiley
6. Gas Processors and Suppliers Association, GPSA & Sons, New York, 1992.
Engineering Data Book, Vol. II, 10th edn., GPSA, Tulsa, 19. Meyers, R.A., Handbook of Petroleum Refining Processes,
OK, 1987. 2nd edn., McGraw-Hill, New York, 1997, Chapter 3.3.
7. McCabe, W.L., Smith, J.C., and Harriot, P., Unit 20. American Standard Testing Methods, ASTM D-396:
Operations of Chemical Engineering, 5th edn., McGraw- Standard Specification for Fuel Oils, 1998.
Hill, New York, 1993. 21. Dean, J., Lange’s Handbook of Chemistry, 14th edn.,
8. Perry, R.H., Green, D.W., and Maloney, J.O., Eds., Perry’s McGraw-Hill, New York, 1992.
Chemical Engineers’ Handbook, 7th edn., McGraw-Hill, 22. Munson, B.R., Young, D.F., and Okiishi, T.H.,
New York, 1997, Chapter 27. Fundamentals of Fluid Mechanics, 2nd edn., John Wiley &
9. Leffler, W.L., Petroleum Refining for the Non-Technical Sons, New York, 1994.
Person, PennWell Publishing, Tulsa, OK, 1985. 23. Heald, C.C., Cameron Hydraulic Data, 18th edn., Ingersoll-
10. Gary, J.H. and Handwerk, G.E., Petroleum Refining, 3rd Dresser Pumps, Liberty Corner, NJ, 1994.
edn., Marcel Dekker, New York, 1994. 24. Van Wylen, G.J., Sonntag, R.E., and Borgnakke, C.,
11. Reed, R.D., Furnace Operations, 3rd edn., Gulf Publishing, Fundamentals of Classical Thermodynamics, 4th edn., John
Houston, TX, 1981. Wiley & Sons, New York, 1994.
5
Oil Atomization
I.-Ping Chung and Stephen B. Londerville
CONTENTS
5.1 Introduction...................................................................................................................................................................... 95
5.2 Liquid Fuel........................................................................................................................................................................ 95
5.3 Theoretical Basis of the Atomization Process.............................................................................................................. 97
5.4 Parameters Affecting Atomization............................................................................................................................. 100
5.5 Spray Characteristics..................................................................................................................................................... 101
5.5.1 Droplet Size........................................................................................................................................................ 101
5.5.2 Spray Angle........................................................................................................................................................ 101
5.5.3 Patternation......................................................................................................................................................... 102
5.6 Atomizer Type................................................................................................................................................................ 103
5.6.1 Single Fluid Atomizer....................................................................................................................................... 103
5.6.2 Twin Fluid Atomizer......................................................................................................................................... 104
5.6.2.1 Process Burner Oil Gun..................................................................................................................... 104
5.6.2.2 Boiler Burner Oil Gun........................................................................................................................ 105
5.7 Oil Gun Performance.................................................................................................................................................... 106
5.7.1 Process Oil Gun.................................................................................................................................................. 106
5.7.1.1 Atomization Quality........................................................................................................................... 106
5.7.1.2 Energy Consumption......................................................................................................................... 107
5.7.1.3 Turndown Ratio.................................................................................................................................. 108
5.7.1.4 Pollutant Emissions............................................................................................................................ 108
5.7.2 Boiler Oil Gun.....................................................................................................................................................110
5.8 Prediction Model.............................................................................................................................................................110
5.8.1 Nomenclature......................................................................................................................................................111
5.8.2 Droplet and Carbon Burnout........................................................................................................................... 112
References..................................................................................................................................................................................113
high-velocity mediums, such as air, gas, or steam. For

viscous liquid fuel, for example, No. 6 heavy oil, it is
5.1 Introduction
especially effective to use other high-velocity media to
Atomization is a process of breaking up bulk liquid into assist atomization. The air-assist or steam-assist atomi-
many small droplets. In order to have good combustion, zation is called twin-fluid atomization.
fuel and air must mix well. A bulk liquid has a limited
surface area to contact with the air. This is the reason
that liquid fuel, or, more specifically, fuel oil, requires
atomization before burning. Atomization increases liq-
uid surface area, which results in an increase of fuel
evaporation rate and fuel–air mixing rate. This improves
5.2 Liquid Fuel
the combustion process. By definition, a liquid or fluid is a substance that deforms
The atomization mechanism can be kinetic energy continuously when a shear stress is applied. The fluid
from the liquid itself, or external mechanical energy is called “Newtonian” when the shear stress is propor-
through a special design, such as swirl, vibration, tional to the rate of deformation. The proportionality
and rotation, or even through assistance from other constant is called viscosity.
95
∂u by density, is used. The resultant unit is (length)2/(time).

τ=µ (5.1)
∂y Kinematic viscosity is the most commonly reported and
used in the industry. Kinematic viscosity in SI units is
a Stoke, defined as cm2/s. The common SI unit is the
where
centiStoke (cSt) = Stoke/100. Other common units are
τ is the liquid shear stress
Saybolt Universal Seconds (SSU) and Saybolt Furol
μ is the dynamic viscosity
Seconds (SSF) (see Figure 5.1).
u ∂y is the rate of deformation or velocity gradient
∂
Most liquids are Newtonian. Exceptions are some
long chain hydrocarbons and various kinds of mix-
The units and nomenclature used to define dynamic tures of solids and liquids generally called slurries.
viscosity are almost endless, but essentially (force) Non-Newtonian fluids can be shear thickening or shear
(time)/(length)2. Most of the time another measure, kine- thinning. For atomizer considerations, theses fluids
matic viscosity defined by dynamic viscosity divided have a special designation called “apparent” viscosity,
100
80
70 1 14.5° API Residuum
60 2 23.0° API Cylinder stock
3 23.1° API Heavy motor oil
50 4 (L05 SPGR) Road oil
5 24.4° API Red oil
40 6 26.1° API Light motor oil
7 27.1° API Light paraffin oil
8 32.8° API White oil
30
9 28.8° API Light paraffin oil
10 30.0° API Pressed distillate
11 35.2° API Crude oil
20 12 35.6° API Mineral seal
13 40.4° API Kerosene
14 Gasoline
10 15 Gasoline
1
8
7
6
2
5
3
Centipoises
4
4
3
5
11
2 6
8 7
1 9
10
0.8
0.7
12
0.6
0.5
13
0.4
0.3
14
0.2
15
100 200 300 400 500 600 700

Temperature (°F)
FIGURE 5.1
Viscosity of mid-continent oils. (Adapted from Maxwell, J.B., Data Book on Hydrocarbons, D. Van Nostrand Company, Princeton, NJ, 1950.)
Oil Atomization 97
where the viscosity is rated at the atomizer high shear 141.5

rates. The liquid viscosity is the single most important API gravity = − 131.5 (5.2)
SG
parameter defining atomizer performance in droplet
generation. Fuel oil impurities such as ash, sulfur, and nitrogen are
Most common liquid fuels are hydrocarbon-based important in emission predictors. Actual components in
with a small amount of nonhydrocarbon components the ash such a vanadium, sulfur, and sodium are impor-
such as sulfur, nitrogen, oxygen, ash, and other minor tant parameters in the evaluation of atomizer corrosion
impurities. The American Society of Testing and and deposits. These deposits may be on or about the atom-
Materials (ASTM) has divided fuel oils into five catego- izer, and in the furnace as well. Some heavy oil products
ries, designated as 1, 2, 4, 5, and 6. Number 1 and 2 are may contain ash that is very erosive such as catalyst fines.
called distillate oils and have viscosity close to water. Many other fuels are used in the industries that are not
Number 4 and 5 are typically blends and are uncommon based on petroleum extraction. These are waste liquids
today. No. 6 is a high viscosity oil that requires heating that could be very corrosive or blends of fluids and solids
to reduce viscosity for transport via pipe or atomization. (slurries) that could be both corrosive and erosive. At times,
At room temperatures, the viscosity of No. 6 oil is so it is impossible to use internal mix atomizers or mechani-
high that it can appear to be a solid. Pitch is a fuel oil cal atomizers for corrosive or erosive fuels. Special exter-
that has even higher viscosity than No. 6 oil. nal mix atoizers are used for such applications.
Other important parameters for fuel oils are specific
gravity, API number, distillation temperature, impuri-
ties, and Conradson carbon or Ramsbottom carbon (see
Table 5.1). 5.3 Theoretical Basis of the
Specific gravity (SG) is the ratio of fuel density to
Atomization Process
that of water at 60°F (16°C). The National Bureau of
Standards and the American Petroleum Institute (API) The process of atomization is to disintegrate the liquid
devised another measure of gravity called the API into droplets by external forces. The literature on the
number. They are related by the following formula1: theory of liquid disintegration is fairly extensive,2–9 but
TABLE 5.1
Liquid Fuel Properties
No. 4 No. 5 No. 5
Grade Fuel No. 1 No. 2 (Light) No. 4 (Light) (Heavy) No. 6
Specific gravity, 60°F/60°F 0.8499 0.8762 0.8762 — — — —
(deg API), max 35 min 30 min 30 max — — — —
Flash point, °F min 100 100 100 130 130 130 140
Pour point, °F max 0 20 20 20 — — —
Kinematic viscosity mm2/s
At 100°F min 1.4 2 2 5.8 >26.4 >65 —
Max 2.2 3.6 5.8 26.4 65 194 —
At 104°F min 1.3 1.9 — 5.5 >24 >58 —
Max 2.1 3.4 — 24 58 168 —
Saybolt viscosity
Universal at 100°F
Min — 32.6 32.6 45 >125 >300 >900
Max — 37.9 45 125 300 900 9000
Furol at 122°F
Min — — — — — 23 >45
Max — — — — — 40 300
Distillation temperature, °F
10% point max 420 — — — — — —
90% point min — 540 — — — — —
Max 550 640 — — — — —
Sulfur content, mass, max 0.5 0.5 — — — — —
Corrosion copper strip, max 3 3 — — — — —
Carbon residue, 10% b; % m, max 0.15 0.35 — — — — —
Water and sediment, % vol, max 0.05 0.05 0.5 0.5 1 1 2
Source: Schmidt, P.F., Fuel Oil Manual, 4th edn., Industrial Press, New York, 1985, pp. 20–21. With permission.
generally addresses low-velocity liquid output from Obtaining a general solution for liquid jet disintegration
simple geometric openings. The analysis theory differs requires solving Navier–Stokes equations by using the
greatly from practical applications. Nevertheless, the perturbation method. The perturbation method is a widely
theoretical analysis provides useful information, such used technique in fluid mechanics using the superposition
as control parameters. This section presents the theoret- of an arbitrary small vibration onto main fluid motion and
ical results for a simple single fluid jet for the purpose of then solving the resulting equations. A simple example
understanding the effect of parameters on atomization. will be illustrated in the following derivation:
The common process of atomization is described as a For two-dimensional liquid jet disintegration, the
phenomenon where waves are developed on the liquid Navier–Stokes equations can be expressed in the fol-
surface with gradually increased amplitudes as insta- lowing forms:
bility is approached, until finally droplets are disinte-
grated from a continuous liquid body. Rayleigh was the ∂u ∂u ∂u 1 ∂P  ∂ 2u ∂ 2u 
first to solve this problem in 1878.10 He considered the +u +v =− + ν 2 + 2 
∂t ∂x ∂y ρ ∂x  ∂x ∂y 
simplified case of vibration and breakup of a cylindrical
jet for a nonviscous liquid in a vacuum environment.
∂v ∂v ∂v 1 ∂P  ∂ 2v ∂ 2v 
The waves developed on the liquid surface can be asym- +u +v =− + ν 2 + 2 
metric waves, i.e., sinuous waves or symmetric waves, ∂t ∂x ∂y ρ ∂x  ∂x ∂y 

i.e., dilational waves, as shown in Figure 5.2.
Later, Weber11 extended Rayleigh’s derivation to vis- ∂u ∂v
+ = 0 (5.6)
cous liquid as follows: ∂x ∂y
1/6 The small perturbation as indicated in Equation 5.7

D  Wel 0.5 
= 1.436  1 + 3 (5.3) is then substituted into the Navier–Stokes Equation 5.6,
d0  Re 
where u = V + ul
D is the droplet diameter v = vl
(5.7)
d0 is the jet diameter
Wel (Weber number) and Re (Reynolds number) are P = Po + Pl
defined as
where
ρlVl d0
2 V is the liquid bulk velocity
Wel = (5.4) Po is the liquid pressure before superposition of
σ
perturbation
ρlVl d0 ul, vl, and Pl are the small perturbations
Re = (5.5)
µl
After a series of substitution, derivation, transforma-
where tion, simplification, and by adding some assumptions,
ρl is the liquid density the Equation 5.6 can be written in a dimensionless form
Vl is the liquid exit velocity as follows:
σ is the liquid surface tension
2
μl is the liquid viscosity H ( iw ) = 2 ( K + iL ) iw + R + iS = 0

λ where
H = 1 + M ctghε

(a) 2 2
K= ε (1 + N ctghε )
λ LP
1
L= ε We
(b) 2
FIGURE 5.2  1 
Liquid disintegration of a cylindrical jet caused by wave formations R = ε 2  εctghε − We

 2 
on liquid surface either by sinuous wave (a) or dilational wave (b).
Oil Atomization 99
We ρd 3 Π 1 = Dl a1V b1ρ g c1 = M −3 c1L1+ a1+ b1+ c1T − b1 = M 0 L0T 0

S = 2ε 3 , ε = kd0 , iω l = iω l 0
Lp σ
Π 2 = σl a 2V b 2ρ g c 2 = M 1+ c 2La 2 + b 2 − 3 c 3T −2 − b 2 = M 0 L0T 0
and Π 3 = µ ll a 3V b 3ρ g c 3 = M 1+ c 3 La 3 + b 3 − c 3 −1T − b 3 − c 3 = M 0 L0T 0 (5.9)
Π 4 = µ g l a 4V b 4ρ g c 4 = M 1+ c 4 La 4 + b 4 − c 4 −1T − b 4 − c 4 = M 0 L0T 0
ρ V 2 2d0 1 µ 12
We = 1 , =
σ Lp ρ1σ 2d0 Π 5 = ρll a 5V b 5ρ g c 5 = M 1+ c 5La 5 + b 5 − 3 c 5 − 3T − b 5 = M 0 L0T 0
(5.8)
ρ µ cosh ε
M = 2 , N = 2 , ctgh ε =
ρ1 µ1 sinh ε Applying the power product method to zero expo-
nents in Equation 5.9, one obtains the following results:
For a detailed derivation of the aforementioned equa-

tions (see Refs. [12,13]). In the aforementioned equations, D
Π1 =
ω represents vibration frequency, d 0 is the liquid jet l
diameter, and k is the wave number (k = 2π/λ, λ is σ 1
wave length). The solution of Equation 5.8 is a com- Π2 = =
ρ gV 2l We
plex form. Its imaginary portion determines the
character of motion’s stability. The real part denotes µl
Π3 = (5.10)
the vibration amplitude with positive sign for decay ρ gVl
and negative sign for instability. It is obvious that the
solution will include We (Weber number), Lp (Laplace µg
Π4 =
number, or Ohnesorge number, Lp = 1/Oh), M (gas– ρ gVl
liquid density ratio), and N (gas–liquid viscosity
ratio). Therefore, these four dimensionless numbers ρl
Π5 = =M
are the control parameters for the droplet diameter, ρg

which represents the liquid disintegration capability.
A simpler alternative method can be used to investi-
After rearranging Π3 and Π4 as follows:
gate the control parameters by using the dimensional
analysis method, or called Buckingham PI theory.14
Proper selection of the physical parameters is criti- Π 32 µ2 1
cal when using this method. Assume the atomization Π′3 = = l =
Π 2Π 5 ρlσl Lp
process can be characterized by the droplet diameter D (5.11)
and the following physical properties affect the droplet Π µ
Π′4 = 3 = l = N
diameter: Π4 µ g

l—Characteristic length of atomizer, i.e., jet diam- The equations become

eter or liquid sheet thickness, etc
V—Liquid exit velocity or liquid relative velocity  ρ gV 2l ρlσl ρ g µ g 
D
to ambient environment = f , 2 , ,  = f (We, Lp, M, N ) (5.12)
l  σ µ l ρl µ l 
σ—Liquid surface tension
ρl, ρg —Liquid and ambient gas density, respectively
The results are the same as Equation 5.8.
μl, μg —Liquid and ambient gas viscosity,
In reality, the problem is much more complicated. For
respectively
example, the disintegration of a liquid sheet can also be
caused by the liquid perforation, especially for high vis-
Here, eight parameters (D, l, V, σ, ρl, ρg, μl, μg) are listed, cosity fluids as shown in Figure 5.3.15 The photo is a swirl
but only three basic primary dimensions, i.e., M (mass), hollow-cone spray with high viscosity liquid. The liquid
L (length), and T (time). Therefore, n = 8 − 3 = 5 PI addi- viscosity is about six times that of typical water. From
tional variables are needed to describe the relationship. the photo, perforation on the liquid surface can be seen
Choosing l, V, and ρg as primary variables and reduc- in addition to waves developed on the surface. Adding
ing the other parameters to dimensionless (Π) variables the perforation effect to the atomization process makes
yields the analysis even more complex. It is difficult to have
TABLE 5.2
Effect of Dimensionless Parameters on Liquid Breakup
Length
Parameter Definition Physical Meaning Effect
ρlVl Inertia
Reynolds Re = Negative
number µl Viscosity
ρ gV 2 l Inertia
Weber number We = Negative
σ Surface tension
ρl σl Surface tension
Ohnesorge Oh = Positive
number µl2 Viscosity
ρg Gas density
Density ratio M= Negative
ρl Liquid density
FIGURE 5.3 µg Gas viscosity

Viscosity ratio N= Negative
A hollow-cone swirl spray with high viscosity liquid (ν = 6 mm2/s). µl Liquid viscosity
The liquid surface shows perforation and waves developed before
disintegration.
process and utility/industrial burners, the liquid prop-

an analytic solution to describe this hydrodynamic and
erty (or oil) has wide variations in viscosity with small
aerodynamic process. For this reason, empirical correla-
variations in surface tension and density. Process burn-
tions are generally developed for real atomizations.12
ers are generally used for the hydrocarbon, chemical,
and petrochemical industries. Utility/industrial burn-
ers are mostly used for power generation or steam
generation in commercial/industrial or industrial appli-
5.4 Parameters Affecting Atomization cations. The liquid fuel viscosity can vary by factors of
10 or even a 100, but surface tension and density only
The analysis in Section 5.3 shows that a single fluid atomi-
vary a few percentages, making oil viscosity the domi-
zation is affected by the liquid exit velocity (V), liquid
nate parameter. As mentioned in Section 5.2, oil viscos-
characteristic length (l), liquid density (ρl), viscosity (μl),
ity varies inversely proportionally to temperature—the
surface tension (σl), and ambient gas density (ρg) and vis-
higher the temperature, the lower the oil viscosity is. In
cosity (μg). The exit velocity and characteristic length, (i.e.,
industry, oil viscosity can be maintained in a reason-
jet diameter or liquid sheet thickness) are determined by
able range by heating the oil to a certain temperature.
the nozzle design. The liquid density, viscosity, and sur-
The usual target is to maintain oil viscosity below 200
face tension are the liquid properties. The gas density and
SSU (Saybolt Seconds Universal) or 42 cSt. The Saybolt
viscosity are the ambient gas properties. Therefore, the
Universal Second is a kinematic viscosity used in classi-
atomization quality, which is represented by the droplet
cal mechanics and defined as the time that 60 cm3 (3.66
sizes, is determined by the nozzle design, liquid proper-
in.3) of oil takes to flow through a calibrated tube at a
ties, and the ambient gas properties.
controlled temperature. (As discussed in Section 5.2, the
Before advanced laser instruments were available,
cSt is a unit used for kinematic viscosity.)
measuring the droplet size was difficult. The atomi-
The conversion of SSU to cSt may be approximately
zation quality was investigated by the liquid breakup
made as follows:
length—the shorter the breakup length, the better the
atomization. Literature describing empirical relation-
ships for breakup lengths with those parameters is 32 < SSU < 100 s, cSt = 0.226(SSU) − 195/(SSU )
(5.13)
fairly intensive.8,11,13,16–21 A general trend of the breakup SSU > 100 s, cSt = 0.220(SSU ) − 135/(SSU )
lengths varied with those parameters is summarized
in Table 5.2. If the breakup length increases with an
increase of parameter, it is considered a positive effect. For viscous oils, sometimes SSF is used. Conversion of
If breakup length decreases with an increase of the SSF to cSt can be approximated as follows:
parameter, it is considered a negative effect.
Although the aforementioned discussions are mainly 25 < SSF < 40 s, cSt = 2.24(SSF) − 184/(SSF)
for a single jet fluid atomization, the results can be (5.14)
SSF > 40 s, cSt = 2.16(SSF) − 60/(SSF)
applied to the twin-fluid atomization process. In most
Oil Atomization 101
Since oil viscosity in industry can be maintained in an

appropriate range, the atomization quality then is mainly
controlled by the liquid characteristic length (l), exit veloc- PDPA
processor
ity (V), and density ratio (M). In other words, atomization
quality in industrial applications is mainly a function of Computer
atomizer design and operational conditions. Oil gun
Laser
Spray
5.5 Spray Characteristics chamber
As stated earlier, good combustion requires good air–fuel PDPA
mixing. The purpose of atomization is to create many probe
small droplets to increase the liquid–air contact surface
area. Smaller droplets produce larger contact surface areas.
The atomization quality, therefore, is usually indicated by FIGURE 5.4
the average droplet size. However, in addition to droplet John Zink Spray Laboratory equipped with a PDPA.
size, other spray characteristics, such as spray angle and
patternation, also affect the liquid combustion. Droplet
In order to obtain the SMD droplet size, individual
size, spray angle, and patternation are discussed next.
droplet sizes must be measured. The John Zink Spray
Laboratory22 is equipped with an advanced laser instru-
5.5.1 Droplet Size ment called a phase doppler particle anemometer
The drops generated from atomization are heteroge- (PDPA).23 This instrument can measure the individual
neous and the size distribution may vary widely. It is, droplet size within a spray and calculate the SMD value.
therefore, necessary to express the droplet diameter in The PDPA as shown in Figure 5.4 collects the droplet
a representative mean value. The most common mean diameter and velocity simultaneously for each droplet
values are defined as follows: passing through the laser sample volume. Section 5.7
includes a discussion of PDPA measurements.
Length mean = D10 =

∑ NDi
i i
∑N i
i
5.5.2 Spray Angle
An appropriate spray angle is one of the key elements
1/2
∑ ND
 2  to maintain a stable oil flame in most burners. If the
i i
Surface mean = D20 =  i  spray angle is too narrow, the recirculation vortex is

 ∑N i
i


too small to bring the high temperature flue gas back
to the oil gun tip, resulting in a lift-off flame as illus-
1/3 (5.15) trated in Figure 5.5a. The lift-off oil flame could be

Volume mean = D30 = 
∑ i
N iDi 3 


 ∑ N i 
i
Sauter mean diameter = SMD = D32 =

∑ ND
i
i i
3

∑ ND
i
i i
2
Droplet combustion involves liquid evaporation and

combustion reaction, which consist of heat and mass
transfer. Therefore, the mean value of droplet size should
contain quantities of area and volume. The Sauter mean (a) (b) (c)
diameter (SMD), defined as ratio of volume to area, is the
most representative diameter for atomization applications FIGURE 5.5
Spray angle relative to a stable oil flame: (a) narrow spray angle
involving combustion. The reason is that droplet burnout causes a lift-off oil flame. (b) Good spray angle brings hot flue gases
time is directly proportional to droplet mass or volume back to the spray root and maintains a stable oil flame. (c) Wide spray
and inversely proportional to the reaction area or surface. angle causes coking on the burner tile.
FIGURE 5.6
A gun with a 90° machine angle. Its spray angle actually is about 30°.
extinguished or cause pulsation. If the spray angle is

too wide, the liquid fuel spray can impinge on the tile
edge, resulting in coke build up as illustrated in Figure
Oil gun
5.5c. An appropriate spray angle recirculates high tem-
perature flue gases back to the spray tip and keeps
new droplets ignited, resulting in a stable oil flame as
shown in Figure 5.5b.
In reality, however, most oil guns’ sprays have
curved boundaries. The actual spray angle is not the
same as the mechanical angle projected from a drilled Patternator
hole. For example, the spray angle of a gun with a 90°
machined angle could be actually about 30° as shown
in Figure 5.6. Experimentation is required to measure
the real spray angle. One method of experimentation is
to use a patternator.
5.5.3 Patternation
FIGURE 5.7
Patternation is the liquid distribution within the spray. Patternator to collect water sprayed out of an oil gun.
Combustion applications require definable liquid fuel
distribution to maintain good combustion efficiency
0.3
and low pollutant emissions. The apparatus used to
measure the liquid distribution of an oil gun is called 0.25
Normalized water distribution
a patternator. An example patternator that is designed

by the John Zink Company is shown in Figure 5.7. The 0.2
oil gun is set at the center of a curved plate. Several
small cups are mounted at the curvature to collect 0.15
the water that is sprayed from the gun. The water col-
lected in each cup is then measured by the graduated 0.1
cylinder.
The typical patternation sprayed out of a gun is 0.05
shown in Figure 5.8. (The oil gun tested is indicated in
Figure 5.6.) This graph indicates that water is distrib- 0
uted between −15° and 15°, resulting in a spray angle of –30 –25 –20 –15 –10 –5 0 5 10 15 20 25 30
about 30°. This method demonstrates that patternation Angle (°)
indicates both the uniformity of the liquid spray and FIGURE 5.8
also the actual spray angle of an oil gun. Patternation measurements for a gun shown in Figure 5.6.
Oil Atomization 103
The discharge coefficient for simplex swirl atomizer is

approximated by
5.6 Atomizer Type
5.6.1 Single Fluid Atomizer 0.5 0.25
 Ap   Ds 
Cd = 0.035   D  (5.16)
The single fluid atomizers are generally called “mechan-  DsDo   o
ical” because they use only the liquid fuel pressure for
atomization. Many years ago, it was discovered that where
reasonably good atomization for combustions systems Cd is the discharge coefficient
could be obtained by using mechanical atomizers. For a Ap is the area of swirled ports
plain single orifice jet, fluids must be accelerated to about Ds is the diameter of swirl chamber
400 ft/s (122 m/s) for good atomization.24 The breakup Do is the exit diameter
mechanism is proportional to velocity and discussed in
Section 5.3. This velocity corresponds to a liquid pres- The cone angle is a direct function of the atomizer
sure of around 1000 psig (69 barg). Because of the high geometry ratio, Ap/DsDo.
pressure requirements, single orifice atomizers are not Typical required oil pressures are 500–1000 psig
generally used in atmospheric combustion applications. (35–70 barg) with a lower operational pressure of about
A common mechanical atomizer used in combustion 300 psig (21 barg). The range of turn down can be extended
applications is the “pressure swirl” atomizer, sometimes enormously to 10:1 or more with the use of a wide range
called “Simplex” (see Figure 5.9). Fluid is introduced return flow atomizer. This design is essentially a simplex
tangentially (swirled) creating an air core and then exits style with a return flow to keep the swirl high in the cham-
a small hole forming a hollow cone of liquid spray rang- ber, but limiting the discharge amount (see Figure 5.10).
ing from 50° to 80°. The operating parameters of this
style of atomizer are well known and documented.25,26 Return flow
FIGURE 5.9 FIGURE 5.10

Simplex swirl atomizer. Simplex swirl atomizer with return flow.
Steam
Steam injection
port
Oil
Steam-oil
mixture
FIGURE 5.11
John Zink EA oil gun.
A common problem with simplex atomizers is pulsation small ports before exiting to the tip. The MEA design is
in the air core. Care must be utilized in the design of the similar to the EA oil gun except for a small modification
swirl chamber to prevent pulsation. on the tip design as shown in Figure 5.12. Mixing the oil
with steam not only eliminates the clogging problem, but
also reduces the fuel density and viscosity. This results in
5.6.2 Twin Fluid Atomizer improved atomization quality.
John Zink also developed a High Efficiency Residual
Twin fluid atomizers are widely used in process burn-
Oil (HERO®) gun. Compared to the EA and MEA oil
ers and utility/industrial burners. Typically, the heat
guns, the HERO gun results in much better atomiza-
released using a process burner is much smaller than
tion quality and consumes less atomizing medium, i.e.,
industrial and utility burners.
steam.22 The HERO gun is a registered trade mark of John
Zink in the United States and patented worldwide.27,28
The HERO gun, in addition to reduced liquid den-
5.6.2.1 Process Burner Oil Gun
sity and viscosity as discussed for the EA and MEA oil
As discussed in Section 5.3, the control parameters for guns, also employs a special design (a modification of
atomization quality include liquid characteristic length, Y-jet) to decrease the characteristic length of liquid fluid.
exit velocity, density, and viscosity. To achieve good atom- The HERO gun design is shown in Figure 5.13.
ization, those parameters should be manipulated within
practical limits. For example, it is well known that a smaller
jet diameter produces better atomization. However, a
small exit port may be easily clogged (the heavy oil, espe-
cially No. 6 residual oil always contains some solid parti-
cles). For process burner applications, the problem is more
challenging. The heat release for each process burner is
usually much smaller than that of the boiler burners. In
order to keep the process burner oil port diameter large
enough to be free from clogs, the number of ports has to
be reduced. This may result in poor oil gun spray patter-
nation and atomization quality, resulting in a poor flame
pattern. Because of this, most oil guns in process burners
mix steam with the oil. In this way, the fuel port diam-
eters can be enlarged with a reasonable number of exit
ports to maintain appropriate spray conditions. The John
Zink EA and MEA oil guns are typical examples of this
design. For the EA oil gun as shown in Figure 5.11, the fuel FIGURE 5.12
oil is mixed with the steam by injecting steam through John Zink MEA gun.
Oil Atomization 105
Atomizing ports (8)

WDH tip
Liquid port
FIGURE 5.15
WDH waste aqueous gun design with one liquid exit port sur-
rounded with eight atomizing ports.
of diameter D. If D = 1.1 d, the thickness of the liquid

cone may become 0.32d, decreasing the liquid character-
istic length by almost 70%. Forming a liquid cone also
FIGURE 5.13 increases the contact surface area between liquid and
John Zink HERO gun. The tip is a modified Y-jet. gas. A large liquid–gas contact surface area improves
The Y-jet29 is well known and widely used in large oil- the disintegration process.
fired boiler burners. In the Y-jet, fuel is injected into an One limitation for Y-jet is that every Y-jet has to be
atomizing medium stream with a certain angle as illus- an individual port. Clogging can become a problem
trated in Figure 5.14. As shown in the figure, the stream for process oil guns with small heat releases. For this
of the atomizing medium pushes liquid fuel against the reason, the Y-jet design is not normally applied to pro-
wall, which makes the cylindrical liquid jet form an cess oil guns. The John Zink HERO gun eliminates this
annular cone. Forming an annular liquid cone reduces drawback by mixing fuel oil with steam before entering
liquid characteristic length. For example, a liquid jet into individual Y-jet. The liquid fuel port d, therefore,
with a diameter d is injected into a mixing chamber may be enlarged. Mixing with steam also decreases the
fuel oil density and viscosity resulting in even better
Atomizing atomization.
medium A good oil gun design results in a definable and
repeatable patternation. In the attempt to prevent clog-
Liquid fuel ging, some oil gun designs sacrificed the patternation.
Steam port
For example, the WDH gun, as shown in Figure 5.15,
has liquid flowing straight down through one exit port
surrounded by several atomizing ports. While this
d
design has no problem with clogging, the patterna-
tion has been compromised. The patternation for the
D WDH is shown in Figure 5.16. The figure shows that
Fuel port
Exit port liquid distribution for the HERO gun is much uniform
than that for the WDH gun. A uniform patternation
has the advantage of less soot generation. The HERO
gun also has lower particulate emissions compared to
other oil guns.
5.6.2.2 Boiler Burner Oil Gun

Boiler burner oil guns have capacities that range from
3 MW (10 × 106 Btu/h) to 130 MW (450 × 106 Btu/h).
Smaller boilers in the range of 50,000–1 million pounds
per hour (PPH) (23,000–454,000 kg/h) steam generally
use a variety of twin fluid internal mix atomizers or tip
mix such as the Y-jet already discussed in detail earlier.
FIGURE 5.14 Many years ago, utility size boilers almost always used
Y-jet atomization principle. mechanical atomizers such as the Simplex wide range
Super HERO, 80/35, ma/mw = 0.24, mw = 1.8 gpm

0.4
DH, 78/30, ma/mw = 0.81, mw = 1.66 gpm 5.7 Oil Gun Performance
Normalized water distribution
The oil gun performance can be evaluated by atomiza-

0.3 tion quality (or droplet size), energy consumption (or
atomizing medium consumption), turndown ratio, and
pollutant emissions including particulate, CO, and NOx
0.2 emissions. In the following, the process oil gun and
boiler oil gun are discussed separately.
0.1
5.7.1 Process Oil Gun
5.7.1.1 Atomization Quality
0
–40 –30 –20 –10 0 10 20 30 40 The atomization quality is determined by the mean
Angle (°)
droplet size. As mentioned in Section 5.5, the SMD is
the most representative mean diameter for atomization
FIGURE 5.16
Patternation comparisons for HERO and WDH guns. The HERO gun
community. Differing operation conditions can change
shows much more uniform liquid distribution than the WDH gun. the oil gun’s atomization quality. Some PDPA measure-
ments for different oil guns at different operation con-
ditions are presented here to demonstrate the control
return flow (WRRF) design due to the excessive cost parameters. (In the PDPA experiments, the liquid is
of atomizing steam and because the furnace residence water and air is the atomizing medium.)
time is high. Today, utilities’ units have been moving to The PDPA measurements of the MEA oil gun for a
twin fluid atomizers utilizing low steam consumption, given water pressure (=60 psig or 4 barg) at various air
typically 0.05–0.1 mass ratios. The driving decision to pressures are presented in Figure 5.18. In the figure, Pw
change is usually based on emission reduction. represents the water pressure and Pa is the air pressure.
It is common to design special cap drilling patterns The abscissa represents the position across the spray
for boiler applications. For package boilers, the tip drill- cone of one exit port in an oil gun. The droplet size is
ing may be elliptical to produce elliptical flames to fit represented by SMD with standard deviation. The results
the aspect ratio of the furnace. Similar tip features can indicate that the higher air pressure generates the smaller
be utilized on tangentially-fired utility units to prevent droplet. This is expected because high air pressure con-
flame impingement (see Figure 5.17). The drilling pat- tributes to higher momentum to break up the water jet.
terns can vary to produce staged combustion to reduce
NOx emissions, carbon loss, and opacity.30 Internal mix
atomizers are used rather than Y-jets due to the complex
drilling patterns. Pw = 60 psi, Pa = 60 psi, ma /mw = 0.128
Pw = 60 psi, Pa = 80 psi, ma /mw = 0.315
Pw = 60 psi, Pa = 100 psi, ma /mw = 0.777
Pw = 60 psi, Pa = 120 psi, ma /mw = 1.580
SMD (μm)
–5.0 –2.5 0.0 2.5 5.0

Radial position (cm)
FIGURE 5.18
FIGURE 5.17 Droplet size measurements of the MEA oil gun at different air
Coen elliptical cap slots for low-NOx. pressures.
Oil Atomization 107
Now, the question is which parameter plays the more MEA oil gun, ma /mw = 0.128
important role: water pressures or air–water differen- MEA oil gun, ma /mw = 0.315
tial pressure. The testing results for the oil gun at the MEA oil gun, ma /mw = 0.777
same differential pressure but different water pres-
MEA oil gun, ma /mw = 1.580
sures, are shown in Figures 5.19 and 5.20. With similar
HERO gun, ma /mw = 0.18
air–water differential pressures, the oil gun generates
SMD (μm)
similar droplet sizes in Figure 5.19, but different results
are shown in Figure 5.20. This indicates that the same
air–water differential pressures do not necessarily gen-
erate the same droplet sizes. Also note that the test case
for water pressure of 60 psig (4 barg) in Figure 5.20 does
–5.00 –2.50 0.00 2.50 5.00

FIGURE 5.21
Droplet size comparison measured by PDPA for different oil gun
SMD (μm)
designs.
not have smaller droplet size than that for 40 psig (2.75
barg). This infers that neither water pressure nor dif-
ferential pressure is a control parameter. By reviewing
Pw = 60 psi, Pa = 80 psi, ma/mw = 0.300
the calculation of the air-water mass ratio (ma/mw), the
Pw = 80 psi, Pa = 100 psi, ma/mw = 0.295
parameter that dominates the droplet size becomes
Pw = 100 psi, Pa = 120 psi, ma/mw = 0.297
prominent. In Figure 5.19, the air-water mass ratio is
similar and the oil gun generates the similar droplet
–5.0 –2.5 0.0 2.5 5.0
sizes. In Figure 5.20, even though the air–water differ-
ential pressures are the same, the mass ratios are dif-
FIGURE 5.19 ferent and consequently the atomization droplet sizes
Droplet size measurements of the MEA oil gun at similar air–water are different—the higher the mass ratio, the better the
differential pressures. atomization. The same trend can be seen in Figure 5.21.
In conclusion, mass ratio is the main control parameter
for oil gun atomization quality.
Pw = 40 psi, Pa = 80 psi, ma /mw = 1.341 For different oil guns with similar mass ratio, i.e.,
Pw = 60 psi, Pa = 100 psi, ma /mw = 0.777
0.128 for the MEA and 0.18 for the HERO, the droplet
size for the MEA oil gun is almost three times that of
the HERO gun. From another perspective, to achieve
the same droplet sizes, the MEA oil gun requires a
mass ratio about eight times that of the HERO gun.
SMD (μm)
This implies that the HERO gun is much more energy

efficient. This validates the previous discussion in
Section 5.3 that liquid characteristic length is a domi-
nate parameter for atomization quality. The HERO gun
has a much smaller characteristic length than the EA
or MEA oil guns. In other words, oil gun design is one
of the control parameters.
–5.0 –2.5 0.0 2.5 5.0 5.7.1.2 Energy Consumption

Radial position (cm) In industry, steam is the most popularly used atom-
FIGURE 5.20
izing medium for the oil guns. The steam consump-
Droplet size measurements of MEA oil gun at the same air–water dif- tion for a typical oil gun in process burners is about
ferential pressure but different mass ratios. 0.3–0.4 lbsteam/lboil. The John Zink MEA or EA guns fall
Oil gun steam consumption Oil gun capacity

16
MEA oil gun
14
HERO gun
12
Heat release (MMBtu/h)

10
8
Mass ratio
0
0 20 40 60 80 100 120
Oil pressure (Psig)
0 2.5 5 7.5 10 12.5 15 17.5 20 FIGURE 5.24

Heat release (MMBtu/h) A typical oil gun capacity curve. It shows oil gun turndown ratio.
FIGURE 5.22
Steam consumption for different oil gun designs. 5.7.1.3 Turndown Ratio
An oil gun with a high turndown ratio is easy to oper-
ate. The turndown ratio of an oil gun is defined as the
in this category. The HERO gun is a much more energy ratio of the maximum heat release to the minimum heat
efficient oil gun. Therefore, it consumes less steam as release with stable oil flames. A well-designed oil gun
shown in Figure 5.22. will result in a higher turndown ratio compared to a
In process oil burners, the oil gun is usually operated poorly designed oil gun. A well-designed oil gun gener-
at a constant steam–oil differential pressure. The steam ates smaller droplets for an equal gas/liquid ratio. The
pressure is 20 psig (1.4 barg) or 30 psig (2.0 barg) higher smaller droplets vaporize fast and are ignited easily.
than the oil pressure. Under this condition, the steam Consequently, the atomizer has the capability to main-
consumption for one specific oil gun is a function of tain a stable flame at low oil pressure. The John Zink
heat release (or oil pressure), as indicated in Figure 5.23. HERO gun typically achieves a turndown ratio of 4:1 at
The steam consumption is high at low heat release and oil pressures of 100 psig (6.9 barg), and the MEA or EA
low at high heat release. oil guns usually achieve a turndown ratio of 3:1.
The oil gun turndown ratio increases with higher
available oil pressure. As an example, an oil gun capac-
ity curve is shown in Figure 5.24. If the available oil pres-
sure is 100 psig (6.9 barg), the oil gun can easily achieve
a turndown ratio of more than 4:1. If the oil pressure is
only 80 psig (5.5 barg), the turndown ratio is reduced to
Steam consumption
around 3:1. Boiler burner oil atomizers can be designed

for up to a 12:1 turndown.
5.7.1.4 Pollutant Emissions

It is well understood that good atomization generates
less particulate emissions. As indicated in Figure 5.18,
an increase of mass ratio by increasing air-water differ-
ential pressure improves the atomization (i.e., smaller
0 2 4 6 8 10 12 14
droplet size). This infers that an increase of differential
Heat release (MMBtu/h)
pressure reduces the particulate emissions. A combus-
FIGURE 5.23
tion test of the HERO gun was conducted by varying the
Steam consumption curve for a constant steam–oil differential pres- steam–oil differential pressures. The particulate emis-
sure oil gun. sions were measured using the EPA Method 531 and are
Oil Atomization 109
TABLE 5.3
Combustion Performance of the HERO Gun
Steam–Oil Particulate
Heat Release Air Temp. Furnace Dry Differential Mass Ratio Emission
(MW) (°C) Temp. (°C) O2 (%) Pressure (Bar) (%) (%) NOx (%)
4.5 232 867 3.9 2 100 100 100
4.5 232 846 4.0 2.7 110 65 108
TABLE 5.4
Combustion Test Results for the MEA Oil Gun
Heat Release Differential Furnace Steam–Oil Mass
(MW) Pressure (Bar) O2 (%) Temperature (°C) Ratio (%) NOx (%)
4.6 1.1 3.08 1043 175 116
4.6 2.1 3.17 1001 275 136
listed in Table 5.3. The combustion results confirm the Oil gun NOx comparison
trend of the PDPA measurements (i.e., the higher the dif-
ferential pressure, the lower the particulate emissions).
MEA oil gun
The table also shows that an increase of mass ratio by
HERO gun
increasing the differential pressure creates an increase
in NOx emissions. The same trend is also observed by
testing the MEA oil gun as listed in Table 5.4. In this test,
the particulate emissions were not measured by a quan-
titative method, such as EPA Method 5, but by a qualita-
NOx
tive method, the Bacharach test method. The Bacharach

number is lower for the high differential pressure case.
The EPA Method 5 and the Bacharach test method are
described in detail in Chapter 10.
In the comparison, the first case in Table 5.3 is used as a
baseline, or 100%, and the other cases in Tables 5.3 and 5.4
are divided by its values. The comparison of NOx emis-
sions for both oil guns is illustrated in Figure 5.25.
In conclusion, for one specific oil gun, the differential 0 2.5 5 7.5 10 12.5 15 17.5 20
pressure may be increased to reduce the particulate emis- Heat release (MMBtu/h)
sions, but NOx emissions may increase. In order to reduce
FIGURE 5.25
both particulate and NOx emissions, a different oil gun Comparison of NOx emissions for the HERO and MEA oil guns.
design may be employed. A comparison of the HERO
case as shown in Table 5.3 to the MEA case in Table 5.4
demonstrates that the HERO has better performance on
both pollutant emissions.32 TABLE 5.5
CO emissions usually correlated with the particulate
emissions. The parameters that affect the particulate Effect of Operational Parameters on Pollutant Emissions
emissions also have a similar effect on the CO emissions. Operational NOx CO Particulate
Other operational conditions, such as O2 level, furnace Parameters Emissions Emissions Emissions
temperature, and air temperature, also affect the pollut- O2 level Positive Negative Negative
ant emissions.33 A general trend of pollutant emissions Furnace Positive Negative Negative
varied with operating condition parameters is summa- temperature
rized in Table 5.5. A positive effect is defined when emis- Combustion air Positive Negative Negative
sions increase with an increase of the parameter value. temperature
Conversely, a negative effect is defined when emissions Differential Positive Negative Negative
pressure
decrease with an increase of the parameter value.
In summary, a good oil gun design not only reduces the

particulate and CO emissions, but also reduces NOx emis-
5.8 Prediction Model
sions. A strong understanding of atomization principles is
essential in the development of a well-designed atomizer. As discussed, predictive models have been developed for
pressure swirl (Simplex) atomizers for more than 50 years.
Recently Coen Co. has developed new predictive codes for
5.7.2 Boiler Oil Gun Y-jet atomizers and internal mix atomizers. In prior years,
Residence times for boilers are generally much lower testing full scale atomizers was common. The following is
than those of process applications. Thus, the importance an outline of the Y-jet code used to measure flow and pres-
of droplet distribution becomes paramount in the predic- sure predictions (see Figure 5.26). The solution involves
tion of opacity and carbon loss. For mechanical Simplex compressible flow calculations, partial momentum
atomizers, a significant amount of work has been pub- exchange, orifice losses, and a multiple iterative solution.
lished to predict mean droplet size. An excellent sum- Compute the Mach number through gas orifice with
mary can be found in Atomization and Sprays.14 An easy to known mixer pressure, mass flows, and iterating on sup-
use approximation for Simplex atomizers SMD is34 ply gas supply pressure as follows until convergence:
−0.5
ptg  k − 1 2
( k +1)/( 2− 2 k )
SMD = 2.25σ0.25µ 0L.25mL0.25 ( ∆P )L ρ−A0.25 (5.17) mg = A gM kg  1 + M  (5.18)
RTt  2 
where Compute gas temperature at the orifice:

σ is the surface tension liquid, kg/s2
μ is the kinematic viscosity liquid, kg/m s Tt
Tgo = (5.19)
m is the mass flow rate liquid, kg/s 1 + ((k − 1)/2)M 2
ΔP is the pressure drop liquid, Pa
Compute gas pressure at the orifice:
ρ is the density of ambient, kg/m3
Ptg
Little work has been openly published on drop size Pgo = (5.20)
relations for twin fluid atomizers since the early work of (1 + (k − 1/(2))M ) 2 k /( k − 1)
Wigg in 1959.35 The industry usually relies on full scale

testing for drop size relations as discussed previously.
Atomizing gas
Subsequent correlations are empirically developed based
on atomizer geometry, liquid properties, and gas/liquid Tg Pg
ratio. Similarly, the pressure and flow relations are devel-
oped based on measured fluid losses, gas dynamics, and
momentum exchange correlated to actual test data.
Atomizers are selected for boilers primarily based on Liquid fuel Dg
emissions and turndown requirements. Utility boilers Tf Pf
rarely require high turndown due to multiple burners Df
and the need to generally fire at high load. For these
reasons, a WRRF simplex atomizer is the overwhelming
choice in these applications where it can be used. Dm
When a large multiburner boiler requires steam atom- Pm
izers, the Y-jet atomizer is a good solution due to the abil-
ity to generate small enough droplets at low gas/liquid
mass ratios of 0.05–0.1. The usual turndown required is
4 or 5:1. Large single burners are more difficult applica-
tions and may require 8–20:1 turndown.
Small boiler furnaces have small residence times and
require small droplets to limit particulate and opacity
emissions. Mechanical atomizers are almost never used
in small furnaces with the exception of cold light off. Exit
Y-jet atomizers are preferred, but may not be used if Pme
nonuniform flame patterns are needed. Generally, for
boiler applications, inside mixers are used and require FIGURE 5.26
0.1–0.15 gas/liquid ratios for normal applications. Diagram of Y-jet.
Oil Atomization 111
Compute density at the gas orifice: the mixer gas density and temperature will depend on
partial heat transfer between fluids:
Pgoi k3 is the fraction of conserved gas stagnation velocity
ρ goi = (5.21)
RTgio
k 3ρ gmVgm
2
Pme = Pm + ∆Pm + (5.27)
Compute velocity through the gas orifice completing the 2 gc
first part of the converged solution with the assumed
mixer pressure: Compute the exit Cd using conservation of mass:
mg mg
Vg = (5.22) Cde = (5.28)
ρ go Ag cdg Vgm Amρ g
Next, using the same fixed flows, compute the mixer Compute the exit Mach number from pressure ratio:
pressure by iterating on the mixer inlet pressure
assumed in the first iteration part one, above for the k /( k −1)
Pme  k −1 2 
atomizing gas flow using the partial conservation of = 1+ M  (5.29)
momentum and compressible flow at the exit. Pa  2 
With the known fuel flow, compute the fuel velocity
in the fuel orifice: Find the error in part two iteration by assuming the
mixer inlet pressure as follows:
mf
Vf = (5.23)
( ρ f A f ) cd f ( k + 1)/( 2 − 2 k )
pme  k − 1 2
mg = Ae (Cde )( M ) kg  1 + M 
With the known mixer pressure compute the fuel sup- RTm  2 
ply pressure. Include known formulas for viscosity cor- (5.30)
rections to find the actual fuel pressure.
Iterate part two until the mixer pressure is found.
Iterate part one with new mixer pressure until the new
( p f − pm ) = ( ρ f Vf2 ) 2 g 1cd2 (5.24) gas supply pressure is found. Continue back and forth
c f
until a solution for part one and part two is found.
Next, balance partial momentum exchange in the This iterative technique requires experimental coef-
mixer before the exit orifice. Correct the gas density ficients and heat transfer considerations to yield cor-
for the actual mixer temperature and pressure. From rect estimates. For simplicity, normal and oblique shock
the conservation of mass compute the gas velocity in waves were omitted. When the gas inlet is choked, set
the mixer. the Mach number to 1, and correct the pressure and
Define Vr = Vfm/Vgm as the ratio of oil to gas velocity at temperature. When the exit flow is choked, the resultant
the mixer outlet, a function of gas to liquid mass ratio. external shock wave can generally be ignored, however
From the conservation of mass excessive under expansion at the exit can cause droplet
deposition on the atomizer surface.
 m f mg   1 
Vgm =  +   (5.25)
V ρ
 r f ρ   Am 
gmi 
5.8.1 Nomenclature
From the partial conservation of momentum define A area

Cd discharge coefficient
k1 as the fraction of oil inlet momentum conserved g gravitational constant
k2 as the fraction of gas inlet momentum conserved gc proportionality constant between force, mass, and
Compute the pressure rise or loss due to inlet momen- acceleration
tum less exit momentum: K ratio of specific heats; constant pressure/volume
M Mach number
M mass flow rate

( ∆pm ) Am gc = k1Vf m f + k2Vg mg − Vgm ( mg + Vr m f ) (5.26)
P pressure
R universal gas constant
Compute the mixer exit pressure by including the par- T temperature
tial conservation of mixer gas stagnation velocity where ρ density
Subscripts: For HC oxidation, several sources are available such

as Barnes.38 In general,
a ambient
d discharge d (C a H b )
e exit dt
( )
= −5.52 × 108 P −0.815 (T ) e(12 , 200/T )
f fuel
0.5
g gas × (C a H b ) (O 2 ) moll/cm 2 s (5.32)
i inlet
For CO destruction, several kinetic data are available
m mixer
such as Williams and Hottel39:
o orifice
t total pressure
d [ CO ] 0.5
= −1.8 × 107 e( −25 , 000/RT ) ( CO )( O 2 )
The code is complicated and requires multiple dt
simultaneous to be solved. Experimental coefficients 2
are obtained from full scale testing. Other codes based 0.5  P 
× (H 2 O ) 
on similar principles are used for internal mix atomi-
 RT  (5.33)
zation with good success. The correlation of SMD test
patterns to gas/liquid ratios has shown good success Most published CO rates involve H2O because CO destruc-
also. tion requires the OH radical to produce the reaction.
After, and partially during gaseous oxidation, the
original droplet becomes void of hydrocarbons and
5.8.2 Droplet and Carbon Burnout becomes a char “particle.” The particle can be a solid
As discussed earlier, the process of droplet burnout is sphere or a hollow ceno-sphere depending on the orig-
heat up, evaporation, gaseous oxidation, and char burn- inal droplet size and fuel properties such as Conradson
out. Extensive analysis has been done on droplet heat up Carbon or Ashaltene content. In either case, the resul-
and evaporation by, for example, Chin and Lefebvre.36 tant char size and mass has shown to be related to fuel
Most of this work has been related to gas turbines where Conradson Carbon and Asphaltene content via the
the rate of combustion is limited by evaporation rates. In coke formation index or CFI (see Figure 5.27).40
furnaces, the evaporation rate is small compared to the
25
available residence time.
A much more simplistic approach is to use the effec-
tive evaporation constant that includes heat up and
evaporation, as discussed by Lefebvre.37 It is defined 20
as follows:
D02
λ eff = (5.31)
te 15
% Fuel property
where
D0 is the initial droplet diameter in μm
te is the time in s 10
λ eff is the evaporation constant
For injection of droplets into hot flame zones λ eff is about

5
0.8 (μm2/ms).
For usual liquid fuels used in furnaces, the total
heat up and evaporation time ranges from 10 to 20 ms
and is not significant compared to mixing and char 0
burnout. 0 5 10 15 20
For oxidation of the gasified droplet gases, it % CFI
is customary to use two-step kinetics, where the
Asph % CCR % RCR % Coen CCR % Princt CL
hydrocarbon fuel is assumed to oxidize to CO and
CO oxidizes to final products. In practice, the lim- Poly. (CCR %) Poly. (asph %) Linear (princt CL)
iting chemistry is the final CO oxidation. Further,
large scale eddy mixing is, in part, responsible for FIGURE 5.27
the mixing oxidizer and fuel. CFI versus CCR, RCR and asphaltenes.
Oil Atomization 113
For hollow ceno-spheres, the CFI computes the resid- and

ual char size from the original oil droplet as follows:
2D O2 N 2
Km = cm/s (5.39)
D2 ρ Dc
CFI = 6 c3 cs δ (5.34)
Do ρo
where42
And for a solid char the relationship is
 1 1 
3
ρ DO2 N2 ≅ 0.00026 Tk3  + †
CFI = D c c (5.35)  MWO2 MWN 2 
3
Do ρo
where diffusivity in cm 2/s (5.40)
Dc is the char diameter
Do is the droplet diameter and
ρcs is the density of ceno-sphere char surface
ρo is the density of oil droplet MW = Molecular weight of O2 or N2
ρc is the density of char particle
δ is the ceno-sphere shell thickness Combining Equations 5.36, 5.37, 5.39, and 5.40 yields a
final char loss formula:
Smaller droplets will form solid char and larger
droplets will form ceno-spheres. After determining the
char size and structure, a char burnout model can be dm −12C og A p
selected. The char will then oxidize by diffusion of oxy- = g C/s (5.41)
dt  Dc 32 
gen to the surface and surface reactivity. Two models +

can be used depending on the initial atomizer droplet  (0.000116 ) Tk
1.5
K p RTk 
size. One model is shrinking diameter constant den-
sity for solid chars, and the other is constant diameter Combining the aforementioned equations dm/dt can
shrinking density for ceno-spheres. In either case, the be integrated numerically or directly for the case of con-
oxidation is defined as follows41: stant diameter/shrinking density with constant O2 and
temperature. Since this is rarely the case, the normal
dm −12C og A p solution will involve droplet size distribution, reduc-
= (5.36)
dt (1/K m ) + (1/K r ) ing O2, changing temperature, and a simple “marching”
solution that is simple enough to do in a spreadsheet.
where Either the constant diameter or constant density model
dm/dt is the grams carbon/s consumed can be used after defining the char size distribution.
Cog is the molar density of O2 g mol/cm3 Further this procedure will yield not only the carbon
Ap is the area of initial char, cm2 particulate, but the size distribution as well. From this
Kr is the surface reaction coefficient in cm/s distribution and mass loading, opacity can be predicted
Km is the diffusion coefficient in cm/s using simple light scattering equations.
and
K p RT k
Kr = cm/s (5.37)
32
where References
1. Schmidt, P.F., Fuel Oil Manual, 4th edn., Industrial Press,
K p = Ae( − E/Tk R′ ) †reaction rate char data New York, 1985, pp. 34, 35.
2. Schweitzer, P.H., Mechanism of disintegration of liquid
in g cm 2 s atm O 2 (5.38) jets, J. Appl. Phys., 8, 513–521, 1937.
where 3. Giffen, E. and Muraszew, A., The Atomization of Liquid
E is the activation energy in cal/g-mole Fuels, Chapman & Hall, London, U.K., 1953.
4. Krzywoblocki, M.A., Jets—Review of literature, Jet
R′ = 1.98 cal/g-mole K
Propul., 26, 760–779, 1957.
A is the preexponential factor in g/cm2 s atm O2 5. Rutland, D.F. and Jameson, G.J., Theoretical prediction of
R = 82.057 cm3 atm/g mole K the sizes of drops forms in the breakup of capillary jets,
Tk is the temperature in K Chem. Eng. Sci., 25, 11, 1970.
6. Phinney, R.E., The breakup of a turbulent liquid jet in a 26. Lefebvre, H., Atomization and Sprays, Hemisphere Pub
gaseous atmosphere, J. Fluid Mech., 60, 689–701, 1973. Corp., New York, p. 171, 1989.
7. Kitamura, Y. and Takahashi, T., Influence of nozzle on 27. U.S. Patent No. 6,478,239, High efficiency fuel oil atom-
breakup of a liquid jet, Proceedings of ICLASS-78, Tokyo, izer, January 2002.
Japan, 1978. 28. U.S. Patent No. 6,691,928, High efficiency oil atomizer
8. Mahoney, T.J. and Sterling, A.M., The breakup length of method, January 2003.
laminar liquid jets in air, Proceedings of ICLASS-78, Tokyo, 29. Mullinger, P.J. and Chigier, N.A., The design and perfor-
Japan, 1978. mance of internal mixing multijet twin fluid atomizers,
9. Simmons, H.C., The atomization of liquids: Principles and J. Inst. Fuel, 47, 251–261, 1947.
methods, Parker Hannifin Report, No. 7901/2-0, 1979. 30. Drennan, S., Mandayam, V., and Rice, G., Low NOx
10. Rayleigh, L., On the instability of jets, Proc. London Math. experiences firing residual oil in industrial boilers,
Soc., 10, 4–13, 1878. AFRC International Symposium, Ottawa, ON, Canada,
11. Weber, C., Disintegration of liquid jets, Z. Angew. Math. 1997.
Mech., 11(2), 136–159, 1931. 31. Environmental Protection Agency Method 5,
12. Lyshevskii, A.S., Processes of Fuel Atomization in Diesel Determination of particulate emissions from station-
Nozzles, Mashgiz, Moscow, Russia, 1963 (in Russian). ary sources, 40 CFR (Code of Federal Regulation) Chapter
13. Bayvel, L. and Orzechowski, Z., Liquid atomization, 1 Part 60 Appendixes A—Standards of Performance for New
in Combustion: An International Series, N. Chigier, Ed., Stationary Sources Test Methods, 2000.
Taylor & Francis Group, Boca Raton, FL, 1993. 32. Chung, I.P., Strupp, C., and Karan, J., New fuel oil
14. Lefebvre, A.H., Atomization and Sprays, Hemisphere atomizer for improved combustion performance
Publishing, New York, 1989. and reduced emissions, presented at 6th European
15. Chung, I.P. and Presser, C., Fluid property effects on Conference on Industrial Furnaces and Boilers, Lisbon,
sheet disintegration of a simplex pressure-swirl atom- Portugal, 2002.
izer, J. Prop. Power, 17(1), 212–216, 2001. 33. Chung, I.P., Baukal, C., and Ruiz, R., The effect of
16. Buckingham, E., On physically similar systems: operational parameters on the burner NOx emis-
Illustrations of the use of dimensional equations, Phys. sions, presented at 2004 International Flame Research
Rev., 4(4), 345–376, 1914. Foundation Members Conference, Noordwijkerhout, the
17. Grant, R.P. and Middleman, S., Newtonian jet stability, Netherlands.
AIChE J., 12(4), 669–678, 1966. 34. Lefebvre, A., Gas Turbine Combustion, Hemisphere Pub.
18. Hiroyasu, H. Shimizu, M., and Arai, M., The breakup Co., Washington, DC, p. 400, 1983.
of high speed jet in a high pressure gaseous atmo- 35. Wigg, L.D., The Effect of Scale on Fine Sprays Produced
sphere, Proceedings of the 2nd ICLASS-82, Madison, WI, by Large Airblast Atomizers, Report 236, National Gas
pp. 69–74, 1982. Turbines Establishment, Pyestock, England, 1959.
19. Arai, M., Shimizu, M., and Hiroyasu, H., Breakup length 36. Chin, J.S. and Lefebvre, A.H., The role of the heat-up
and spray angle of high speed jet, Proceedings of the 3rd period in fuel drop evaporation, Int. J. Turbo Jet Engines,
ICLASS-82, London, U.K., pp. 1B/4/1-10, 1985. 2, 315–325, 1985.
20. Dombrowski, N. and Johns, W.R., The aerodynamic 37. Lefebvre, A.H., Gas Turbine Combustion, Taylor & Francis
instability and disintegration of viscous liquid sheets, Group, Boca Raton, FL, pp. 47–48, 1999.
Chem. Eng. Sci., 18, 203–214, 1963. 38. Barnes R.H., Saxton M.H., Barrett R.E., and Levy A.,
21. York, J.L., Stubbs, H.F., and Tek, M.R., The mechanism Chemical Aspects of Afterburner Systems, EPA report EPA-
of disintegration of liquid sheets, Trans. ASME, 75, 1279– 600/7-79-096, NTIS PB298465, p. 21, April 1979.
1286, 1953. 39. Williams, G.C., Hottel, H.C., and Morgan, A.C., The com-
22. Chung, I.P., Karan, J., and Struff, C., High efficiency bustion of methane in a jet-mixed reactor, 12th Symposium
residual oil (HERO®) atomizer for process heater (International) on Combustion, the Combustion Institute,
burners, 2002 American Flame Research Committee Pittsburgh, PA, p. 913, 1969.
(AFRC), Spring Meeting, Ottawa, Canada, 2002. 40. Urban, D.L., Huey, S.P.C., and Dryer, F.L., Evaluation
23. Bachalo, W.D., Method for measuring the size and veloc- of the coke formation potential of residual fuels, 24th
ity of spheres by dual-beam light scatter interferometry, Symposium (International), The Combustion Institute,
Appl. Opt., 19(3), 363–370, 1980. Pittsburgh, PA, pp. 1357–1364, 1992.
24. Castleman, R.A., Mechanism of the Atomization of Liquids, 41. Smoot, L.D. and Smith, P.J., Coal Combustion and
Bureau of Standards, Paper # 281 p. 376, March 1931. Gasification, Plenum Press, New York, pp. 81–90, 1985.
25. Tanasawa, Y. and Kobayasi, K., Study on swirl atom- 42. Bird, R.B., Stewart, W.E., and Lightfoot, E.N., Transport
izer, Technical report of the Tohoku University, Vol. 20, Phenomena, John Wiley, New York, 1960.
Tohoku, Japan, pp. 27–58, 1955.
6
Solid Fuel Combustion in Suspension
Stephen B. Londerville and Timothy Webster
CONTENTS
6.1 Introduction.....................................................................................................................................................................115
6.2 Fuel Properties and Characterization..........................................................................................................................115
6.2.1 Coal.......................................................................................................................................................................116
6.2.2 Wood, Biomass, Pet Coke...................................................................................................................................116
6.3 Oxidation of Solid Fuels.................................................................................................................................................116
6.3.1 Heat-Up, Devolitization, and Volatile Oxidation...........................................................................................116
6.3.2 Char Oxidation....................................................................................................................................................118
6.3.3 Flammability Characteristics............................................................................................................................119
6.4 Fuel Conveying.............................................................................................................................................................. 120
6.4.1 Pressure Drop Calculations in Solid/Gas Conveying.................................................................................. 120
6.4.2 Horizontal Transport......................................................................................................................................... 120
6.4.3 Vertical Transport and Minor Losses............................................................................................................. 121
6.4.4 Conveying Options............................................................................................................................................ 121
6.5 Burner Designs............................................................................................................................................................... 121
6.5.1 Utility and Multiburner Applications............................................................................................................ 121
6.5.2 Industrial Burners.............................................................................................................................................. 121
6.5.3 Support Fuel....................................................................................................................................................... 123
6.6 Furnace and Control Considerations.......................................................................................................................... 123
6.7 Combustion Controls.................................................................................................................................................... 123
6.8 Emission Formation and Prediction........................................................................................................................... 123
6.9 Conclusions..................................................................................................................................................................... 125
References................................................................................................................................................................................. 125
with more concentration dedicated to conveying, burn-

ers, and the actual oxidation of solids in suspension.
6.1 Introduction
The utilization of solid fuels dates back to early civili-
zation and the discovery of fire, used for cooking and
warmth long before liquid or gaseous fuels were in use.
Early utilization of solid fuels in the last few centuries
started with cut wood or chunk coal burned in stoves
6.2 Fuel Properties and Characterization
and fireplaces. Early industrial use of solid fuels uti- For all solid fuels, the primary properties are defined by
lized various forms of grates with chunk wood or coal. proximate and ultimate analysis. An example of each
Today, solid fuels are used in boilers (both industrial type of analysis is shown in Table 6.1 for Pittsburgh
and utility), kilns, and air heaters. Seam Coal, West Virginia. Generally, both proximate
This chapter will present the use of solid fuels pneu- and ultimate analysis are required to design a system.
matically conveyed in suspension. This means the solid This allows computation of heat input, combustion air
fuels need to be ground or “pulverized” into much calculations, and most emissions. Each solid fuel type
smaller sizes and pneumatically conveyed, usually with will have grind-size requirements for conveying and
air, to a burner system. The preparation of the fuel and unit performance with respect to heat transfer and
delivery to the burner system can be very complex. emissions. The following sections will detail some spe-
Only the basics of this preparation will be presented cifics for some common, but not all, solid fuels.
115
TABLE 6.1 post-combustion cleanup. Sulfur in coal contributes

Coal Analyses on As-Received Basis to acid rain, opacity, and particulates. The ash content
contributes to opacity and particulates. Even with zero
ASTM D3172 ASTM D3176
carbon loss, the stack gas of almost all coal installations
Proximate Analysis Ultimate Analysis will require sulfur and particulate cleanup that is quite
Component % Weight Component % Weight expensive for all but large capacity users. A great deal
of information is known about coal fuels as it has been a
Moisture 2.5 Moisture 2.5
utility fuel for 100 years or so.
Volatile matter 37.6 Carbon 75.0
Fixed carbon 52.9 Hydrogen 5.0
Ash 7.0 Sulfur 2.3
6.2.2 Wood, Biomass, Pet Coke
Total 100 Nitrogen 1.5
Heating value Oxygen 6.7 Less common solid fuels include wood, various bio-
13,000 Btu/lbm Total 100 masses, and pet coke. Many of these fuels are “renewable.”
30,238 KJ/kg All have specific and sometimes unique requirements
for grinding, conveying, ignition, burnout, erosion, and
Source: Adapted from Stultz, S.C. and Kitto, J.B., Steam: Its
Generation and Use, 40th edn., Babcock and Wilcox, emissions. All of these fuels can be pneumatically con-
Barberton, OH, p. 8.5, 1992. veyed and burned in suspension. The low volatility fuels
may require support fuel. The high sulfur and ash fuels
usually require some kind of post-combustion cleanup.
6.2.1 Coal
Rice hulls, coke, char fines, and coal are the most erosive
Coal is, by far, the most commonly used solid fuel, sup- fuels and will require special wear plates. The high nitro-
plying 32% of the world’s energy. Coal is ranked by gen fuels will need to address conversion to NOx emis-
class from oldest to youngest, geologically.1 Coal starts sions. Staging air and fuel has shown to be effective in the
initially forming from vegetation and wood under reduction of NOx emissions.
pressure and temperature over a long period of time. Some other fuels not listed that have been used are,
During this time period, the initial formation goes from planer shavings, sander dust, sunflower seed hulls, pea-
humates to anaerobic and then peat. Final formation to nut shells, and olive pits. Wood waste from manufactur-
coal then proceeds in order to yield lignite, subbitumi- ing may need to include the glues used in the product
nous, bituminous and finally anthractic coal over time. not specified by the parent wood for emission calcula-
Coal is defined by class as shown in Table 6.2. tions. Table 6.3 outlines some of these fuels’ properties.
In recent years, coal has been dubbed a “dirty fuel,”
primarily because of the increased use of natural gas,
which has lower pollution emissions as opposed to
actual coal emissions that are generally reduced with
6.3 Oxidation of Solid Fuels
TABLE 6.2 All solid fuels oxidize in a similar manner. The pro-
Class of Coals cesses are heat-up, devolatilization, volatile oxidation,
Class 1 Anthracitic High rank with high carbon, low and finally char burnout. During the process, the solid
moisture, but high heating value, particle can swell or shrink, break up or not, and is very
greater than 14,000 Btu/lbm. It is specific to the fuel.
low in sulfur and volatiles. Premium
fuel most commonly used in
domestic markets 6.3.1 Heat-Up, Devolitization, and
Class 2 Bituminous Commonly burned in utilities with Volatile Oxidation
medium heating values, volatiles,
and fixed carbon All solid fuels require initial heat-up as the first step in
Class 3 Subbituminous High in moisture and volatiles. Low oxidation to dry the material. Depending on the fuel
ash and sulfur. Heating values at type, swelling, shrinking, and breakup may occur par-
10,000 Btu/lbm
tially at this stage. In normal burners heat-up rates are
Class 4 Lignitic Lowest heating values, below
8000 Btu/lbm and high moisture
very high, on the order of 105 K/s.2 For normal burners
and volatiles. Lowest in fixed carbon where fuel is delivered pneumatically in suspension,
and ash these heat-up times are very low (<5 to 15 ms) and can
Source: Adapted from Stultz, S.C. and Kitto, J.B., Steam Its generation
generally be neglected. Coal devolatilization can be
and Use, 40th edn., Babcock and Wilcox, Barberton, OH, modeled by the method of Kobayashi.3 This method
p. 8.5, 1992. uses a parallel pair of first order irreversible reactions
TABLE 6.3
Solid Fuel Combustion in Suspension
Fuel Properties of Biomass and Solid Fuels

Chemical Composition (% by wt., Dry Basis)
Bark Wood Coal
Rice Coffee Apple Delayed Char
Fuel Characteristics Pine Oak Pine Fir Bagasse Hulls Grounds Pomace Peat Coke Fines Sub A Bit A
Proximate analysis
Volatile matter 72.9 76.0 79.4 75.1 83.8 64.3 86.0 71.3 67.3 12.1 2.0 34.3
Fixed carbon 24.2 18.7 20.1 24.5 12.7 20.7 13.1 23.0 22.7 87.6 58.4 56.4 59.2
Ash 2.9 5.3 0.5 0.4 3.5 15.0 0.9 5.7 10.0 0.3 39.6 4.7 6.5
Ultimate analysis
Hydrogen 5.6 5.4 6.3 6.3 5.8 5.7 7.7 5.9 5.3 3.9 0.4 5.0 5.2
Carbon 53.4 49.7 51.8 50.7 48.8 44.9 57.3 42.2 52.2 89.6 58.6 75.1 79.5
Sulfur 0.1 0.1 0.0 0.0 0.0 0.1 0.2 0.1 0.4 4.2 1.0 1.0 1.3
Nitrogen 0.1 0.2 0.1 2.4 0.2 0.5 1.4 0.7 1.8 1.2 0.4 1.4 1.4
Oxygen 37.9 39.3 41.3 40.2 41.7 33.8 32.5 45.4 30.3 0.8 0.0 12.8 6.1
Ash 2.6 5.3 0.5 0.4 3.5 15.0 0.9 5.7 10.0 0.3 39.6 4.7 6.5
Heating value
Dry basis (Btu/lb) 9,030 8,370 9,130 8,795 8,100 7,226 10,950 8,090 9,057 15,400 8,758 13,595 14,366
Source: CoenBCS 1983 Bulletin.
117
and seems to fit measured data very well. Combined, the At high temperatures a long list of kinetic data is
rate of devolatilization can be expressed in first order as available for computing volatile oxidation. Generally,
a two-step mechanism is used where the HCs oxi-
dv
= ( y1k1 + y 2k 2 )c (6.1) dize to CO and then CO oxidizes to CO2. References
dt for HC oxidation include several sources such as
where Barnes.7
dv/dt is the rate of devolatilization on a mass basis In general,
y1 is the fraction of volatiles close to the proximate
analysis
d(CaH b )
y2 is the fraction of volatiles at high temperatures, = −5.52 × 108 (P −0.815 )(T )e(12 , 200/T)
often near unity dt
k1 = B1exp(E1/RT) is the Arrhenius rate at low × (CaH b )0.5 (O 2 ) mol/cm 2s (6.4)
temperatures
k2 = B2exp(E2/RT) is the Arrhenius rate at high
For CO destruction, several kinetic data are available
temperatures
such as Ref. [8].
c is the coal particle mass
T is the absolute temperature in Kelvin
2
R is the universal gas constant in consistent units d[CO]  P 
= −1.8 × 107 e( −25 , 000/RT ) (CO)(O 2 ).5 (H 2O).5 
dt  RT 
At high temperatures as experienced in real flames, coal
(6.5)
devolatilization ranges from 40 to 10 ms from lignite to
bituminous coals. Almost all published CO rates involve H2O because CO
For wood and biomass, a first order reaction is also destruction requires the (OH)−1 radical to produce the
commonly used as described by Roberts,4 using a con- reaction.
stant diameter, shrinking density model. After, and partially during, gaseous oxidation, the
dρ original solid particle becomes void of hydrocarbons
= −ρK (6.2) and becomes a char “particle” that can be treated as
dt
mostly carbon. The resultant char is rarely spherical
Integrating except in the case of pet coke.
f = ε − K∆T (6.3)
6.3.2 Char Oxidation
where
ρ is the density in consistent units Despite the fact that the final char is not always spheri-
f is the fraction devolatilized cal, it is convenient to treat it as such. With this assump-
ΔT is the time of devolatilization in seconds tion for a solid char, the relationship is
K = Aexp(E/RT) an Arrhenius rate with units of
inverse seconds Dc3 ρc
(1 − V ) = (6.6)
Do3 ρo
Considerable variations in the rate data have been
recorded and summarized by Malte.5 Suggested values where
for A and E are 7 × 107 s−1 and −30 K cal/gmol. Dc is the char diameter, consistent units
However, for the devolatilization of biomass, not only Do is the initial particle diameter corrected for
does the density decrease, but the biomass particles breakup, swelling, or shrinking
undergo breakup according to the aspect ratio of the ρo is the density of initial particle, consistent units
particle and material type.6 Thus, the primary function ρc is the density of char particle assuming it is carbon
of devolatilization in biomass oxidation is the reduc- V is the proximate volatile fraction
tion in size and density for the next oxidation step, the
reduction of the resultant char called char burnout. After determining the char size and structure, a char
A final note on devolatilization for a variety of pet burnout model can be applied to determine the burnout
coke derived fuels—these fuels are usually void of sig- time. The char will oxidize by diffusion of oxygen to
nificant volatiles and tend to oxidize as constant density the surface and surface reactivity. Two different models
shrinking diameter char discussed later under the sec- can be used, depending on the initial char size and fuel
tion on char oxidation. They are treated as near solid type. First, the more common model is shrinking diam-
carbon spheres for practical burnout times. eter, constant density. The second model is generally
Solid Fuel Combustion in Suspension 119
used for large ceno-sphere constant diameter, shrink- Burnout for subbituminous coal char at 15% EA
ing density. In either case, the oxidation is defined as 100
follows2: 90
dm −12Cog Ap 80
= (6.7)
dt (1/K m ) + (1/K r ) 70
Percent burnout
where 60
dm/dt is the grams carbon/second consumed 50
Cog is the molar density of O2 g mol/cm3
Ap is the area of initial char, cm2 40
100 μm
Kr is the surface reaction coefficient in cm/s 30 150 μm
Km is the diffusion coefficient in cm/s 200 μm
20
and 10
K p RTk 0
Kr = cm/s (6.8) 0 0.2 0.4 0.6 0.8 1
32 Seconds
where FIGURE 6.1
Subbituminous char burnout Coen code A = 60 and E = 17,150.
K p = Ae( − E/Tk R′ ) †reaction rate char data
in g/cm 2 ⋅ s ⋅ atmO 2 (6.9) constant O2 and temperature. Since this is rarely the
case, the normal solution will involve char size distri-
where bution, reducing O2, changing the temperature and
E is the activation energy in cal/g mol a simple “marching” solution that is simple enough
R′ is the 1.98 cal/g mol K to do in a spread sheet. Either the constant diameter
A is the preexponential factor (g/cm2 · s · atmO2) or constant density model can be used after defin-
–
R = 82.057 cm3 atm/(g mol K°) ing the char size distribution. Further, this procedure
Tk is the temperature in K will yield not only the final particulate, but the size
distribution as well. From this distribution and mass
and loading, opacity can be predicted using simple light
scattering equations.
2DO2 N2 See Figures 6.1 and 6.2 for examples of computed char
Km = cm/s (6.10)
Dc burnout for subbituminous coal char and pet coke char.
Note that pet coke char requires longer to burnout due
where Bird, Stewart, and Lightfoot demonstrated9 to the low surface reactivity.
 1 1  6.3.3 Flammability Characteristics

DO2 N2 ≅ 0.00026Tk3  + † (6.11)
 MWO2 MWN 2  The oxidation mechanism for solid fuels has been dis-
cussed from heat-up to devolatilization and char burn-
and out. These mechanisms define the flame volume and
DO2 N2 = Diffusivity in cm2/s the emissions. Solid fuels exhibit unusual flammabil-
MW = Molecular weight O2 or N2 ity characteristics not found in other liquid or gaseous
fuels. The rich limit of flammability is almost infinite
Combining 6.7 and 6.11 yields the final char loss formula: for solid fuel/air mixtures. This is due to the fact that
all of the air is present, but much of the fuel still exists
as unreacted char or volatiles. The gaseous air does not
dm −12Cog Ap
= g C/s initially blend with fuel vapor as the fuel vapor is still
((D /(0.000116)T ) + (32/K RT ))
(6.12)
dt c k
1.5
p k in solid form and thus the ignitable gaseous mixture
remains lean over a wide range, even though the actual
Combining the aforementioned equations, dm/dt can mixture of air and solids can be very rich. Also the
be integrated numerically or directly integrated for range of local stoichiometric mixtures near and around
the case of constant diameter/shrinking density with the solid particle ranges widely from high excess air to
Pet coke char at 15% EA 6.4.1 Pressure Drop Calculations

100
in Solid/Gas Conveying
90
Pneumatic conveying of solids can be classified as
80 dense phase or dilute phase transport. In the literature
and industry, there are varying definitions for these
70
flows.11,12 Dilute transport generally means that parti-
Percent burnout
60 cles are pneumatically transported at many times their

50 100 μm settling velocity. For example, 10%–20% of stoichiomet-
150 μm ric air is a typical flow with a velocity at 60–100 ft/s
40 200 μm (18–30 m/s). Actual flow and velocity will depend on
30 particle sizes. Dense phase is used more for transport
than in combustion systems.
20 When dense phase flow is used to transport solid fuels
10 to a burner, the limit is usually about 5% stoichiometric
air flow at a minimum of 80 ft/s (24 m/s). Dense phase
0
0 0.2 0.4 0.6 0.8 1 flow is not recommended for solid fuel delivery due to
Seconds slugging and plugging issues. An excellent treatment of
dilute phase transport has been presented by Hinkle13
FIGURE 6.2 and is summarized here for convenience.
Pet coke char burnout Coen code A = 15 and E = 19,000.
Coal dust velocity versus equivalence ratio stoichiometric 6.4.2 Horizontal Transport
(F/A) mass
35 For horizontal transport, Hinkle13 proposed the follow-
cm/s ing relationships:
30
Flame velocity cm/s
25
Vg2 γ g WVp fLγ gVg2   4 f pVp  W 
∆Ph = + + 1+  (6.13)
20 2g g 2 gD   fVg  Vg γ g 
15
10 The first term is the initial acceleration of the gas, the

second term is the initial acceleration of the particles,
5
the last term is composed of fluid losses and losses
0 based on friction of solids to fluid and
0 2 4 6 8 10 12
Equivalence ratio where
ΔPh is the horizontal pressure drop in lbf/ft2
FIGURE 6.3
Coal dust flame velocity versus equivalence ratio. (Adapted from Vg is the gas velocity in ft/s
Smoot, L.D. et al., Propagation of laminar pulverized coal-air flames, γg is the specific weight of the gas in lbf/ft3
16th Symposium International on Combustion, the Combustion Institute, g is the 32.2 ft/s2
Pittsburgh, PA, p. 375, 1976.) W is the solids loading in lbm/s/ft2
Vp is the particle velocity in ft/s
near zero when reacting. The situation can be quantified f is the Darcy pipe friction factor
as shown by measurements of Smoot et al.10 plotted in L is the pipe length in ft
Figure 6.3. D is the pipe diameter in ft
f p is the fluid to solid friction loss, defined next
and
2
6.4 Fuel Conveying 3 γ gCd D  Vg − Vp 
fp = (6.14)
The pneumatic conveying of solid fuels can cause ero- 2γ p Dp  Vp 
sion, depending on the fuel properties, pipe materials,
and conveying velocity. Further, it is critical to be able to and
keep the solid material in suspension and be able to pre- Vp
dict piping pressure drop for proper fan selection and
Vg
( )
= 1 − 0.179Dp0.3 γ 0p.5 (6.15)
distribution at the burner.
where
Cd is the particle drag coefficient
6.5 Burner Designs
Dp is the diameter of the particle in inches
γp is the specific weight of the particle in lbf/ft3 Burners for solid fuel share many of the same features as
gas and oil burners. Many of them may also be equipped
It should further be noted that the following term must to fire at a maximum on gas and/or liquid fuel. The
always be less than unity or set to unity: key differences are the installation of a solid fuel injec-
tion pipe or annular scroll. Depending on whether the
solid fuel is conveyed in the dense or dilute phase to
4 f pVp the burner affects the burner internal area required to
≤1
fVg accommodate this.
Since the control of fuel and air mixing is also very
important in controlling the fuel NOx generated by
these burners, solid fuel burners often employ multi-
6.4.3 Vertical Transport and Minor Losses ple air zones to delay fuel air mixing. The burners will
For vertical transport, Hinkel13 proposed the following: likely have to accommodate a variety of changes in fuel
conditions, such as moisture content, grind size, vola-
tility, and heating value over their operational life. To
  4 f pVp  W  WL help optimize performance as these conditions change,
fLγ gVg2
∆Pv = 1+   + (6.16) the burner will often have the ability to adjust ratios
2 gD   fVg  Vg γ g  Vp between the different air zones. The burner will also

have features like adjustable register louvers to control
the amount, and sometimes even the direction of com-
With ΔPv is the vertical pressure drop in lbf/ft2. bustion air spin.
After reviewing minor losses from a variety of refer- Because solid fuel is abrasive, it will abrade away the
ences for many years, the substitution of equivalent burner components that it contacts over time. To accom-
length of pipe for minor losses and the use of Hinkel’s modate this, solid fuel burners are typically equipped
horizontal transport equation are recommended. For with replaceable components such as nozzles and wear
pneumatic conveying, always use large radius elbows. plates.
Pressure drop for short radius elbows can be modeled
using Solids Handling,14 but it is not recommended due 6.5.1 Utility and Multiburner Applications
to the large pressure drop. The author’s experience
with using Hinkel’s method compares very closely In utility and multiburner applications, it is common
with Ogawa.15 to use multiple mills for grinding the solid materials
to the right size and multiple conveying lines for dis-
tribution. The solids are pneumatically conveyed and
supplied to the burners in the form of a round pipe
6.4.4 Conveying Options
in the center of the burner. Various styles of “regis-
Figure 6.4 illustrates a standard arrangement for con- ter” burners are utilized to deliver the secondary air
veying solid fuels from the fuel bin to the burner in what around the center in one or more zones with or without
is known as indirect firing. In this conveying option, swirl. The furnaces are generally large for solid fuel
the solid fuel is preground and delivered/metered to burnout. At times, some units collect and reinject any
a fuel bin. The fuel can be introduced upstream of the unoxidized char.
fan under negative pressure and conveyed through the For NOx control, it is common to use over fired air
fan. Alternatively, the fuel can be introduced down- (OFA), staged levels, or burners out of service (BOOS).
stream from the fan via a rotary air lock feed and sup- Multiple variants to the simple round pipe have been
ply fuel under positive pressure. Both systems can be implemented to control NOx by limiting mixing with
used and both have obvious and subtle pros and cons. the secondary register air.
For example, rotary air locks can leak and running solid
fuel through a blower can cause fan problems. Note the
6.5.2 Industrial Burners
recommended minimum horizontal run and the five-
fold larger pipe radius elbows. In direct fired systems, Industrial burners designed for solid fuels generally
as are commonly used in utilities, the fuel is metered use an annular zone to inject pneumatically conveyed
and delivered directly to the mill and conveyed to the fuel into the furnace. These applications are usually a
burners. single burner and the furnace is small requiring more
122
Shaker table
Fuel bin
with screen
Flex boot
Flex boot
Butterfly valve
Infeed with actuator
hopper shaft of valve to be
Tap for parallel to burner
conveying Fan
air pressure (fuel conveying) Windbox
switch Venturi
Damper
with screen
inlet
Damper
Injection tee
control
actuator Tap for conveying
air pressure switch
Fuel bin
Flex boot
Damper 2˝ NPT for connection Shaker table
control from infeed hopper with screen
actuator
Flex boot
Infeed hopper To conveying Transition
Rotary fan inlet radius
air lock
Option 1: Conveying fan

after fuel inlet
Damper
with screen
inlet
Injection
Fan tee
(fuel conveying)
FIGURE 6.4
Fuel introduction for conveying options.
Once the ash has cooled, some portion of it will drop

out in the furnace, called “bottom ash,” and some por-
tion will be carried out of the boiler with the flue gas,
called “fly ash.” The boiler furnace must have some pro-
vision, such as dropout sections, to allow the bottom ash
to be collected and then removed from the furnace. The
fly ash is typically captured in the exhaust gas cleaning
systems with baghouses, cyclones, or precipitators.
One alternative to the system aforementioned, where
the ash is allowed to cool and be removed “dry” from
the system is to design the boiler furnace to operate in
a “slagging” mode. In this case, the temperatures in
the system are designed to keep any “bottom ash” at a
high enough temperature that it remains liquid. This
liquid ash, or “slag” is then allowed to drain to a low
FIGURE 6.5 point in the furnace and exit through a “slag tap” into
Front of Coen biomass burner.
a cooling pool.
rapid mixing of the fuel and air. The burner is designed
for higher turn down and may have features to reduce
emissions (see Figure 6.5).
6.5.3 Support Fuel 6.7 Combustion Controls

The use of support fuel varies according to specific job Control of solid fuel firing is complicated by the fact
applications. Fuels with high ash and or moisture as that it is more subject to changes in fuel heating value
well as low volatility fuels may require support fuel at than gas or liquid fuels. One of the major factors influ-
all times. Size distribution may be important, depend- encing the change in heating value is solid fuel’s abil-
ing on the fuel type. In many cases, the fuel delivery rate ity to absorb moisture during its transport and storage.
can vary requiring support fuel. In general, the amount Through seasonal changes in humidity and rainfall,
of support fuel can vary from 2% to 10%. For utility the moisture content of solid fuels can vary throughout
applications, support fuel is generally not required due the year. Since the fuel metering systems are based on
to preheated air and multiple burners. Many utility providing either a metered mass or volume of fuel to
applications employ warmup and stabilizing fuel injec- the burner, changes in heating value due to either com-
tors using either natural gas or oil. They are used for position changes, moisture content, or grind size need
startup and stabilization as the fuel properties vary. to be addressed. In most cases, the use of oxygen trim
systems that monitor the oxygen in the flue gas and
make adjustments to the combustion air flow are used
to compensate for these changes. For boiler systems that
utilize induced draft fans, the furnace is controlled at
a negative pressure, less than atmospheric. Therefore,
6.6 Furnace and Control Considerations
care must be taken to minimize the amount of infil-
The primary furnace considerations when firing solid trated air from leakages (tramp air) into the system as
fuels revolve around the high levels of ash that are gen- this will interfere with the ability of oxygen trim sys-
erated. Furnace designs must contain provisions to min- tems to effectively control the combustion air flow.
imize this ash from contacting cool boiler surfaces until
the ash has cooled enough that it is no longer liquid or
“sticky” and will generate slag formation. This typically
involves increasing the size of the combustion chamber
relative to what is used in a gas or oil fired application.
6.8 Emission Formation and Prediction
For example, package boilers firing gas or oil may have
a volumetric heat release in the furnace between 60,000 Coal is a domestically abundant hydrocarbon fuel
and 80,000 Btu/h-ft3 (620 and 830 kW/m3), while those with a lower cost per unit of energy than other com-
same designs, when employed in solid fuel applica- mon industrial or utility fuels, such as natural gas or
tions, may be derated to a range between 30,000 and fuel oil. Biomass combustion can make use of many
50,000 Btu/h-ft3 (310 and 520 kW/m3). waste products found in the timber, manufactured
wood products, and agricultural industries. It not only UBHC, VOCs, and NOx are all important emissions.
provides a source of essentially free fuel, but can also Kinetic rate equations such as 6.4 and 6.5 can be uti-
eliminate many of the disposal problems associated lized when the flow and temperature fields are known
with these byproducts. The environmental regulations to predict reduction in emissions. For the formation
faced by industry are one of the primary governing of emissions, rather than destruction or oxidation of
factors that must be addressed when designing a new hydrocarbons, the equations are similar as shown next.
system. The associated costs to comply with mandated
d(NO)
NOx emissions limits can be significant, especially if it = 2 Ae( − E/RT ) (O 2 )eq (N 2 ) (6.17)
requires the addition of costly flue gas treatment equip- dt
ment. In these cases, the ability to reduce NOx emissions
significantly through changes to the combustion equip- and
ment can make the difference as to whether a system
can meet required NOx levels in a cost effective manner. ko
(O 2 )eq = (O 2 )0eq.5
In order to make significant reductions in the NOx (RT )0.5
emissions from solid fuel combustion without the use
of flue gas treatment, a method must be found to reduce One generally accepted practice is to assume (O2) in
the conversion of fuel bound nitrogen (FBN) to NOx. equilibrium with (O) and (O2) concentration using the
Through the use of air staging, it has been possible to Westenburg16 results for ko for (O2) equilibrium and
reduce the total NOx emissions by more than 50% over Zeldovich constants, A, E as measured by Bowman.17
that of an unstaged system. The utilization of gas kinetic data for emission forma-
Coal, depending on type and origin, will contain tion or hydrocarbon oxidation can be computed from
between 1% and 3% FBN by weight. Raw wood typically detailed temperatures and species concentrations in the
contains 0%–0.25% nitrogen by weight. However, the flow field in many ways such as the Rayleigh flux the-
glues used in manufacturing the panels can increase orem. Simply stated, in all cases, one can post process
the nitrogen content of this waste to as high as 7%, thermal map data in some discrete volume form and/or
with the largest FBN being observed with plants that insert into a CFD code using the Rayleigh flux theorem
use new faster drying resins. The agricultural industry as follows:
also produces a significant amount of biomass waste,
such as rice husks or wheat straw, and increasing air Cv Cs
∂
quality regulations are greatly restricting the amount of
field burning allowed. Plowing this material back into
∂t ∫ nρdv =
∫ nρ(V ⋅ da) (6.18)
the soil can result in increased levels of crop disease and

lower crop yields. These raw biomass materials can also where
contain comparatively high nitrogen levels, with typical n is the chemical in mass units
values ranging from 0.5% to 2.5% by weight. t is the time
One major concern to utilizing these materials as an ρ is the density
industrial fuel is the effect that they may have on the v is the volume
amount of regulated emissions being generated. In a is the area
order to significantly reduce the amount of NOx from V is the velocity vector
the biomass fuels, it is necessary to target the “Fuel
NOx” conversion, since this results in the largest single where, described in words, the formation of (n) through
contribution to the NOx. One proven method of reduc- the volume surface is equal to the integrated rate of forma-
ing the amount of bound nitrogen that converts to NOx tion over the control volume. It is a simple extrapolation
is to introduce the fuel into a primary combustion zone to extend this concept for even coarse volumes as follows:
that is oxygen deficient, or substoichiometric. In this
dn
∑ dt ρ∆v = nρ(V ⋅ a)
way, the nitrogen that is liberated from the fuel has little
(6.19)
or no free oxygen with which to bond and therefore,
mostly recombines to the inert N2. The balance of the air
required to complete the combustion is introduced fur- Solid fuels generally contain a significant amount of
ther downstream in a secondary combustion zone. This elemental nitrogen in the fuel. A portion of the elemen-
guarantees complete burnout of the fuel and reduces tal fuel nitrogen is converted to NOx during the entire
the temperature of the combustion products. process of oxidation. The range of this conversion is very
Solid fuel applications require similar emission dependent on the burner and furnace type as well as
predictions as in gas fuel burners. Carbon monoxide, any NOx reduction techniques utilized. The range can
vary from 4% to 60% and is highly dependent on the Although this chapter only briefly touched on upfront
amount of FBN in the fuel. The FBN contribution can be fuel handling, such handling can be a major problem to
more or less than the thermal NOx contribution. get everything sized correctly and this should be only
From a practical perspective, carbon particulate can be left to truly experienced professionals for design.
calculated on every application as outlined with great
accuracy using Equation 6.12. For most solid fuel appli-
cations ash and sulfur compounds need to be added
to the total particulate. Then, opacity can be computed
based on total particulate loading and size distribution.
Very expensive or large applications will utilize com- References
putational fluid dynamics (CFD) to compute the total 1. S.C. Stultz and J.B. Kitto, Steam Its Generation and Use,
flow field, and post process emissions utilizing the equa- 40th edn., Babcock and Wilcox, Barberton, OH, p. 8.5,
tions in this chapter with the appropriate destruction 1992.
and formation kinetics. The fuel contribution to NOx is a 2. L.D. Smoot and P.J. Smith, Coal Combustion and Gasification,
function of many factors and each burner manufacture Plenum Press, New York, pp. 81–90, 1985.
will have a set of algorithms used for prediction. 3. H. Kobayashi, J.B. Howard, and A.F. Sarofim, Coal
devolatilization at high temperatures, 16th Symposium
International, p. 411, 1976.
4. A.F. Roberts, A review of kinetic data for the pyrolysis of
wood and related substances, Combustion and Flame, 14,
261–272, 1970.
6.9 Conclusions
5. P.C. Malte and B. Dorri, Western States Section of the
The solid fuel fired burner system is a complex process Combustion Institute, The Behavior of Fuel Particles in
involving heat-up, devolatilization, kinetics oxidation, Wood Waste Furnaces, Pullman, Washington, DC, April
char oxidation, diffusion, surface reactivity, and two 13–14, 1981.
phase particle/gas flow. Solid fuel firing offers the advan- 6. S. Londerville, Coen Company internal files and codes
on Wood oxidation, 1982–2010.
tage of being able to utilize lower cost hydrocarbon fuels
7. R.H. Barnes, M.H. Saxton, R.E. Barrett, and A. Levy,
such as coal or petroleum coke. Unique breakup/swell- Chemical Aspects of Afterburner Systems, EPA report EPA-
ing or shrinking particle models will need to be applied 600/7-79-096, NTIS PB298465, p. 21, April 1979.
based on experience with different fuel characteristics. 8. G.C. Williams, H.C. Hottel, and A.C. Morgan, The com-
Erosion needs to be considered for many solid fuels. bustion of methane in a jet-mixed reactor, 12th Symposium
Equally important is the upfront handling, supply, grind- (International) on Combustion, The Combustion Institute,
ing, and conveying of the solid material. Pittsburgh, PA, p. 913, 1969.
Biomass fuels have potential additional advantages, 9. R.B. Bird, W.E. Stewart, and E.N. Lightfoot, Transport
such as being carbon neutral or eliminating waste disposal Phenomena, John Wiley, New York, 1960.
costs. Overall, there are many unique challenges that must 10. L.D. Smoot, M.D. Horton, and G.A. Williams,
be considered in the design of the system, including: Propagation of laminar pulverized coal-air flames, 16th
Symposium International on Combustion, The Combustion
Institute, Pittsburgh, PA, p. 375, 1976.
1. Fuel storage, preparation, and transport complexity 11. D. Mills, M. Jones, and V. Agarwal, Handbook of Pneumatic
2. Higher NOx emissions due to fuel bound Conveying Engineering, Marcel Dekker, Inc., New York,
nitrogen 2004.
3. Higher particulate emissions due to ash content 12. F.A. Zenz and D.F. Othmer, Fluidization and Fluid-Particle
Systems, Reinhold Pub. Co, New York, 1960.
4. Higher SOx emissions due to sulfur content 13. B.L. Hinkle, PhD thesis, Institute of Technology, Atlanta,
5. Ash collection and disposal GA, June 1953.
14. K. McNaughton (Ed.), Solids handling, in Chemical
6. Potential for slag formation, corrosion, and ero-
Engineering, McGraw Hill, New York, p. 130, 1981.
sion in the boiler 15. A. Ogawa, Separation of Particles from Air and Gases, Vol. 1,
7. Larger furnace requirements than gas or oil CRC Press, Boca Raton, FL, pp. 124–129, 1985.
fired boilers 16. A.E. Westenberg, Turbulence modeling for CFD,
8. Increased fuel delivery system and burner wear Combustion Science and Technology, 4, 59, 1971.
17. C.T. Bowman, Kinetics of pollution formation and
due to abrasion
destruction in combustion, Progress in Energy Combustion
9. Increased monitoring of changes to fuel compo- Science, 1, 33–45, 1975.
sition, such as moisture content
7
Heat Transfer
Stephen B. Londerville, Charles E. Baukal, Jr., and Jay Karan
CONTENTS
7.1 Introduction.................................................................................................................................................................... 127
7.2 Thermodynamics........................................................................................................................................................... 127
7.2.1 Energy.................................................................................................................................................................. 128
7.3 Boiler Basics.................................................................................................................................................................... 130
7.3.1 Air Heaters and Duct Burners......................................................................................................................... 133
7.4 Modes of Heat Transfer................................................................................................................................................. 136
7.4.1 Conduction......................................................................................................................................................... 136
7.4.1.1 Thermal Conductivity........................................................................................................................ 136
7.4.1.2 One-Dimensional Steady State Conduction.................................................................................... 137
7.4.2 Convection.......................................................................................................................................................... 138
7.4.2.1 Newton’s Law of Cooling................................................................................................................... 139
7.4.2.2 Composite Wall................................................................................................................................... 139
7.4.2.3 Nondimensionalized Numbers........................................................................................................ 140
7.4.2.4 Flow over a Flat Plate...........................................................................................................................141
7.4.2.5 Turbulent Internal Flow..................................................................................................................... 142
7.4.3 Radiation............................................................................................................................................................. 143
7.4.3.1 Radiation from Surface to Ambient................................................................................................. 143
7.4.3.2 Radiation from Nonluminous, Nonreacting Gases to Surface.................................................... 143
7.4.3.3 Radiation from Nonluminous Reacting Flames to Surface.......................................................... 147
7.4.3.4 Radiation and Convection from Reacting Flames to Surfaces..................................................... 148
7.5 Conclusions..................................................................................................................................................................... 150
References................................................................................................................................................................................. 150
7.1 Introduction 7.2 Thermodynamics
Heat transfer is one of the fundamental purposes of The basics to understanding thermodynamics involve
combustion in all industrial applications. The objective understanding the interaction between energy and mat-
of many industrial combustion applications is to trans- ter. Energy is a driving force that acts as a conduit for
fer energy, in the form of heat, to some type of load for change, manifested in the form of heat and work.2 Like
thermal processing of that load.1 An understanding of many of the topics discussed in previous chapters, ther-
heat transfer and thermodynamics is essential to the modynamics follow a unique set of laws. The first law of
successful design and operation of combustion equip- thermodynamics is simply a subset for the law of conserva-
ment. The objective of this chapter is to review helpful tion of energy, wherein energy can be neither created nor
concepts of thermodynamics and heat transfer, focusing destroyed, merely transferred from one form to another.
on those topics as applied to combustion, particularly in The second law of thermodynamics illustrates that energy
industrial applications. is associated with both quality and quantity, where energy
127
is directed from higher qualities to lower. An example of hCO2 (t) = −1.4678 × 10 −7 T 3 + 1.7328 × 10 −3 T 2
the second law in action would be a hot pie left on a table
would eventually cool. Conversely, ice left in a cup would + 7.928T − 662.14

inevitably melt before reaching room temperature. In both
cases, energy is transferred from the hot medium toward hO2 (t) = −4.0674 × 10 −8 T 3 + 5.6679 × 10 −4 T 2
the cool. Only when the two mediums have reached the
same temperature does the transfer of heat stop occurring. + 6.594T + 13.943

At this point, the two bodies are said to be in thermal equi-
librium. Lastly, the zeroth law of thermodynamics states hH2 O (t) = −5.437 × 10 −8 T 3 + 1.0623 × 10 −3 T 2
that if two bodies are in thermal equilibrium with a third
+ 6.6416T + 371.14
body, then that body is also in equilibrium with each other.
The expressed enthalpies are in units of Btu/lbmole and
7.2.1 Energy are useful for atmospheric combustion calculations. The
error with a third order curve fit is about 0.6% and can
As discussed in earlier chapters, enthalpy is the energy
be improved using a fourth or fifth order to less than
content of a substance and is defined as
0.2%. One important aspect to note is that these equa-
h = u(t) + Pv (7.1) tions can only be used for low pressure conditions. In
the case for high pressure conditions, tables or correc-
where
tions may be necessary. Having all equations in a third
u(t) is the internal energy as a function of temperature
order polynomial makes it very convenient to derive a
only
single third order expression for a mix of gases using
P is the pressure
the mole fractions of components
v is specific volume of fluid
4
Enthalpy of steam and gases at low pressures is tabu-
lated in tables by temperature or by heat capacity, cp for
hmix (t) = ∑ f h (t);
1
i i

gases. Enthalpy, heat capacity, and heat of combustion
are covered in Chapter 3. where one simply sums the polynomial coefficients
Enthalpy can be very useful in finding the temperature multiplied by the mole fraction of the species.
of a gas mixture in which constituents initially had differ- Table 7.2 is provided for N2 gas over a range of
ent temperature. The specific heat of gases relative to tem- 60°F–200°F (16°C–93°C) for use in the following example.
perature (assuming constant atmospheric pressure) can be Figure 7.3 shows an example of a nitrogen gas mix
very closely approximated by curve fitting, as presented chamber.
in Table 7.1.3 Figures 7.1 and 7.2 show the specific heats of
air and flue gases in volume and mass units, respectively. Example 7.1
Equations for enthalpy can be curve fitted from tables
What is the mix temperature of gaseous N2 if the
in third order to be more useful in simple calculations gas comprises 5 lbm(2.26 kg) of N2 at 200°F(93°C)
from Keenan and Kaye as follows as a function of °R4: and 15 lbm(6.8 kg) of N2 at 70°F(20°C)?
hN2 (t) = −3.7432 × 10 −8 T 3 + 5.2874 × 10 −4 T 2 First, calculate the enthalpy values of N2 at their
respective temperatures using the gas table pro-
+ 6.269T + 194.5 vided or the curve fitted enthalpy polynomials.

TABLE 7.1
Specific Heat Curve Fit Equation
Gas Curve Fit Equations (Btu/lbmole°F) Range (°R) Max Error
N2 Cp = −196.08ϴ + 256.38ϴ − 122.56ϴ + 9.3355
−3 −2 −1.5 540–6300 0.43%
H2O Cp = −0.88407ϴ + 19.778 ϴ 0.5 − 43.868ϴ 0.25 + 34.190 540–6300 0.43%
CO2 Cp = 5.7835 * 10−3ϴ2 + 7.2967ϴ 0.5 − 0.98074ϴ − 0.89286 540–6300 0.19%
O2 Cp = 4.8044 * 10−3ϴ1.5 − 42.679ϴ−1.5 + 56.615ϴ−2 + 8.9465 540–6300 0.30%
Temperature (°R ) Temperature(°K )

θ= =
180 100
Source: Adapted from Wylen, G.J.V. and Sonntag, R.E., Fundamentals of Classical
Thermodynamics: Series in Thermal and Transport Sciences, 2nd edn., John Wiley & Sons,
Inc., New York, 1973.
Heat Transfer 129
11
10
Cp (Btu/lbmole°F)
9
7
Air
Flue gas vol.% = 6% CO2
17% H2O 77% N2
6
0 1000 2000 3000 4000
Temperature (°F)
FIGURE 7.1
Specific heat of air and flue gas in Btu/lbmole°F versus temperature. Data plotted from Table 7.1.
0.38
0.36
0.34
0.32
Cp (Btu/lbm°F)
0.3
0.28
0.26
0.24 Air
Flue gas vol.% = 6% CO2
0.22
17% H2O 77% N2
0.2
0 1000 2000 3000 4000
Temperature (°F)
FIGURE 7.2
Specific heat of air and flue gas in Btu/lbm°F versus temperature. Data plotted from Table 7.1.
Btu Now, find the total energy content of the final

N 2 at 200°F → h = 4585.8 mixture
lbmole
Btu  Btu   Btu 
N 2 at 70°F → h = 3680.8  5 lbm × 163.77 †lb  +  15 lbm × 131.45†lb 
lbmole m m
Convert lbmole to lbmass to obtain = 2790.60 †Btu

Btu 1 lbmole Btu
h(70) = 3680.8 × = 131.45 Find the enthalpy of the mixture as follows:
lbmole 28 lbm lbm
1 28 lbm Btu
Btu 1 lbmole Btu 2790.60 Btu = × = 3906.8 †
h(200) = 4585.8 × = 163.77 20 lbm 1 lbmole lbmole
lbmole 28 lbm lbm
TABLE 7.2 Example 7.2

Enthalpy of N2 Gas as Function of Temperature Oftentimes the specific heat of gases does not vary
N2 Temperature (°F) Enthalpy (Btu/lbmole) greatly with temperature; in this case, another
method of solving for mixing temperatures of
60.3 3611.3 gases is through the use of specific heat. Solve the
70.3 3680.8 previous example using this method.
80.3 3750.3 Using Table 7.1, calculate the specific heat of
85.3 3785.1 nitrogen at the specified temperatures.
90.3 3819.9
95.3 3854.7
Btu 
100.3 3889.5 70°F → 6.954
lbmole ⋅ ∆°F  Btu
105.3 3924.3  Average = 6.691
Btu  lb ⋅ ∆°F
110.3 3959.1 200°F → 6.969 mole
lbmole ⋅ ∆°F 
115.3 3993.9
120.3 4028.7
125.3 4063.5
Btu 1 lbmole Btu
130.3 4098.3 6.961 × = 0.2486
lbmole ⋅ ∆°F 28 lbm lbm °F
135.3 4133.1
140.3 4167.9
145.3 4202.7 Through the first law of thermodynamics using
150.3 4237.5 the average specific heat of the two gases to obtain
155.3 4272.3
160.3 4307.1
Heat†loss†by†5lbm = Heat†gain†by 15 lbm
165.3 4342.0
170.3 4376.8
175.3 4411.6
Btu
180.3 4446.4 5 lbm × .2486 × (200 − T )°F
lbm ⋅ ∆°F
185.3 4481.3
190.3 4516.1 Btu
= 15 lbm × .2486 × (T − 70)°F
195.3 4551.0 lbm ⋅ ∆°F
200.3 4585.8
Source: Adapted from Keenan, J.H. and Kaye, J., Gas Tables: Solving the equation for temperature,
Thermodynamic Properties of Air Products of Combustion
and Component Gases, John Wiley & Sons, Inc., New York.
T = 102°F (39°C)

According to Table 7.2, for enthalpy of N2, interpo-
late for temperature over a range of 100°F–110.3°F
Temp (°F) Enthalpy (Btu/lbmole)

100.3 3889.50
T 3906.80 7.3 Boiler Basics
110.3 3959.10
The main task that boilers accomplish is to heat water
and provide steam for various applications. Although
T = 102.79°F (39.3°C)
there are various types of boilers available, the primary
5 lbm @ 200°F components for a boiler are combustion chamber, burner,
heat exchanger, controls, and enclosure.5 Calculation for
the total steam heat absorbed by a boiler is as follows:
Mixing chamber Q = m
 ( hg − hliq ) (7.2)
15 lbm @ 70°F
where
.
Mix temp Q. is the total heat output
m is the steam flow rate
FIGURE 7.3 hliq is the inlet water enthalpy
Nitrogen gas mix chamber. hg is the outlet steam enthalpy
Heat Transfer 131
TABLE 7.3 TABLE 7.4

Enthalpy Values for Saturated Steam and Water Based Enthalpy Values for Saturated Steam and Water
on Temperature Based on Pressure
H2O H2O Pressure
Temperature (°F) hf (Btu/lbm) hfg (Btu/lbm) hg (Btu/lbm) (psia) hf (Btu/lbm) hfg (Btu/lbm) hg (Btu/lbm)
50 18.07 1065.03 1083.10 90 290.67 894.89 1185.60
55 23.07 1062.23 1085.30 95 294.67 891.89 1186.60
60 28.08 1059.32 1087.40 100 298.51 888.99 1187.50
65 33.08 1056.52 1089.60 110 305.78 883.44 1189.20
70 37.08 1054.72 1091.80 120 312.55 878.20 1190.80
75 43.07 1050.83 1093.90 130 318.92 873.21 1192.10
80 48.07 1048.03 1096.10 140 324.92 868.45 1193.40
85 53.06 1045.24 1098.30 150 330.61 863.88 1194.50
90 58.05 1042.35 1100.40 160 336.02 859.49 1195.50
95 63.04 1039.56 1102.60 170 341.19 855.25 1196.40
100 68.03 1036.67 1104.70 180 346.14 851.16 1197.30
110 78.02 1030.98 1109.00 190 350.89 847.19 1198.10
120 88.00 1025.20 1113.20 200 355.46 846.33 1198.80
130 97.99 1019.41 1117.40 250 376.09 825.47 1201.60
140 107.99 1013.61 1121.60
Source: Adapted from Cengel, Y.A., Heat and Mass Transfer: A
150 117.99 1007.71 1125.70
Practical Approach, 3rd edn., McGraw Hill, New York, 2007.
160 128.00 1001.98 1129.98
170 138.02 995.88 1133.90
Using the steam tables, the input and output
180 148.04 993.76 1141.80
enthalpies are
190 158.08 987.62 1145.70
200 168.13 981.37 1149.50 Btu Btu
hg = 1195.5 and hliq = 168.13
210 178.20 971.30 1149.50
lbm lbm
220 180.21 970.09 1150.30
230 188.28 965.02 1153.30 The total heat output supplied by boiler is
240 198.37 958.63 1157.00
lbm Btu
250 208.49 952.01 1160.50 Output = 100, 000 × (1195.5 − 168.13)
h lbm
Source: Adapted from Cengel, Y.A., Heat and Mass Transfer: A
Practical Approach, 3rd edn., McGraw Hill, New York, 2007. Btu
= 102.7 × 10 6
h
Several enthalpy values for steam are found in Tables
7.3 and 7.4 or in the appendix. Refer to the ASME steam In the case of a perfect boiler, the heating output
tables for data on water and steam.6 By finding the heat- would be the same as the input. However, the
ing output, one can find the efficiency or input using the boiler has an efficiency of 80%; therefore, the heat-
ing input is
following relation:
102.7 × 10 6 Btu
Heating†output = 128.4 × 10 6
η= (7.3) 0.8 h
Heating†input
where, η is boiler efficiency. Stack

losses
Tables 7.3 and 7.4 are provided for saturated steam
over a temperature range of 50°F–250°F (10°C–120°C)
and a pressure range of 90–250 psia (6–17 bara).
Fuel in
Example 7.3 Feed Boiler

Find the heating input for the boiler with satu- water in
Steam out
rated feed water input at 175 psia at 200°F with
steam output at 160 psia. The flow rate for steam
is 100,000 pph with boiler efficiency of 80%. See FIGURE 7.4
Figure 7.4. Boiler example.
Example 7.4 Summing the total heat loss will yield

24,741 Btu/mole CH4. Heat is also lost through the
Estimate the boiler efficiency with a nominal
formation of H2O. The total energy lost through
capacity of 100,000 pph. The flue gas exits at
H2O formation is
15% excess air with a stack temperature of 350°F
(176°C). Use CH4 as the fuel for a HHV and LHV
basis. To compute the boiler efficiency, consider a lbmole H 2 O lbm
2 × 18 × 970 (Latentheat
100% efficient boiler and solve for losses, which lbmole CH 4 moleH
2O offormation)
will lower the overall efficiency. The losses con-
Btu Btu
sidered will be dry gas loss at the stack, formation = 34, 920
of H2O, skin loss, and H2O in air. The balanced lbmH 2 O lbmole CH 4
stoichiometric equation for combustion will be
CH 4 + 2φ[O 2 + 3.76N 2 ] Therefore, the total stack losses will be
→ CO 2 + 2H 2 O + 2φ[O 2 + 3.76N 2 ] − 2O 2

Btu Btu Btu
where, 24, 741 + 34, 920 = 59, 661
lbmole CH 4 lbmole CH 4 lbmole CH 4
EA
φ = Excess†air†factor = 1 +
100%
Assuming complete combustion of CH4, one
From this reaction, determine the moles of each
can take the atomic weight and higher heat-
molecule produced per mole of CH4.
ing value to determine the total possible heat of
combustion.
Moles Moles
Produced/ Produced Mole
Molecules Mole CH4 at 15% EA Fraction lbm Btu Btu
16 × 23, 875 (HHV) = 382, 000
CO2 1 1 0.0824 lbmole lbm lbmole
H2O 2 2 0.1848
O2 2ϕ-2 0.3 0.0277
N2 8.648ϕ 8.648 0.7992 Therefore, the stack and steam formation loss will
11.948 result in a drop in efficiency of
Determine enthalpies of gas from Ref. [4] or 59, 661

Loss% = × 100% = 15.62%
Section 7.2.1 equations or Appendix C. Calculate
382, 000
the dry heat loss per lbmole for each molecule at
350°F (177°C). Assume outside air temperature of
70°F (21°C). The enthalpies of each gas at the speci- For the sake of simplicity, the skin loss from the
fied temperatures are given in the following table. boiler casing for a boiler sized at 100,000 lb/hr
(45,000 kg/h) will have a typical heating loss of
Enthalpy at 70°F Enthalpy at 350°F 0.46%.7 Another method is shown in Section
Reactants (Btu/lbmole) (Btu/lbmole) 7.4.3.1 where casing losses are derived.
Lastly, the loss through H2O presence in the
CO2 3968.3 6655.4
combustion air is assumed to be at 80°F (27°C)
H2O 4202.2 6479.7
with 100% relative humidity. The mass percent of
O2 3676.1 5674.9 water in the air and air to fuel ratio of the fuel
N2 3680.8 5634.5 burned must be determined.
Based on a steam table, the saturated pressure
Solving for the total dry gas loss per lbmole of for 80°F (27°C) steam is 0.5074 psia (3.5 kPa). The
CH4, to get following saturation pressure corresponds to a
volumetric percentage of air of
Moles/ h(350)-h(70) Heat Lost (Btu/
Gases Molecule (Btu/Mole Gas) Mole CH4)
0.5074 psia
CO2 1 6655.4–3968.3 2687.10 = 2.07%
14.7 psia
H2O 2 6479.7–4202.2 4555.00
O2 0.3 5674.9–3676.1 599.64
Meaning, at standard pressure, only 2% of
N2 7.52 5634.5–3680.8 16895.60
the volume occupied by air is water. A 2.07% by
Heat Transfer 133
volume translates to 1.2% mass of water. Next, the Next, repeat the same calculation, but for lower
air to fuel ratio for CH4 can be solved to determine heating value conditions. Since the fuel burned
the amount of air released. The equation for air to is still CH4, the dry gas losses remain the same
fuel ratio by weight is expressed from previous without considering the H2O formed from com-
chapters. bustion. However, the heat input changes to
 A  A 1 Btu lbm Btu

  =   × 21, 495 × 16 = 343, 920
F wt F Vol SG fuel lbm lbmole lbmole

 A  B
  = 4.76 × EA  A +  resulting in total losses of
F Vol 4

24, 741(Dry†gas†loss) + 482 (H 2 O †from†air†loss)
where, 343, 920 †
A, B are subscripts for fuel CAHB
SGfuel is the specific gravity of fuel = 7.3%†drop†in†efficiency

Therefore, the air to fuel ratio for CH4 is Therefore, the overall boiler efficiency for a lower
heating value will be
 A  15%   4
  = 4.76 ×  1 +   1 +  = 10.948
F Vol 100%   4 η = 100 − 7.3 (Dry†gas + H 2 O†losses)

 A 1 lbmair − 0.46 (Skin†losses)
= 10.948 × = 19.76
 F  wt 0.554 lbm CH 4
= 92.24%
Using the air to fuel ratio, one can calculate for Note that the boiler efficiency is dependent on
the amount of water in the air as a ratio to fuel the use of higher or lower heating values, but the
burned: mass flow of air and fuel remain unchanged. So
when reporting boiler efficiencies, they need to be
lbmair lb H O lbm CH 4 referenced to higher or lower heating values.
19.76 × 0.012 m 2 × 16
lbm CH 4 lbmair lbmole CH 4
1 lbmoleH 2 O lb H O 7.3.1 Air Heaters and Duct Burners
× † = 0.211 mole 2
18 lbmH 2 O lbmole CH 4
When approaching problems regarding air heaters and
duct burners, they can be solved in a similar fashion to
From the calculation of enthalpy per/molecule/
mole CH4, the value from H2O can be used to boiler problems from the chemistry and thermodynamic
determine the heat loss per lbmole of CH4 burned. view point. For air heaters, find the excess air required to
achieve an outlet temperature using the same procedures to
lbmoleH 2 O Btu compute adiabatic flame. Where the flow is highly diluted,
0.211 × (6479.7 − 4202.2)
lbmole CH 4 lbmole†H 2 O make some simplifying assumptions and use constant spe-
cific heat if the outlet temperatures do not exceed approxi-
Btu
= 482.83 mately 400°F (200°C). This is best shown by Example 7.5.
lbmole CH 4
Resulting is a loss due to H2O in the air of Example 7.5

An air heater is to be sized to heat air from 70°F
Btu
482 (21°C) to 250°F (121°C) using CH4 fuel. The air
lbmole CH 4
× 100% = 0.126% flow is 200,000 lbm/h (90,718 kg/h). Find the gas
Btu firing rate in lbm/h and heat input based on LHV.
382, 000
lbmole CH 4 Repeat for air at 70°F (21°C) to 1800°F (980°C).
Start with the standard oxidation equation and
Taking the summation of the losses, the boiler apply thermodynamics for an exact solution:
efficiency for a higher heating value will be
Given:
CH4 at an unknown excess air
η = 100 − 15.62(Stack†losses) − 0.46(Skin†losses)
Find:
− 0.126(H 2 O †from†air†losses) = 83.79% Adiabatic flame temperature

From the general form, For more accuracy, one can also apply an aver-
age specific heat between the two tempera-
 B tures. This average is about 0.245 Btu/lbm · °F.
C A H B + ε  A +  (O 2 + 3.76N 2 )
 4 The resulting firing rate is then 8.82 × 106 Btu/h
(2.58 MW), almost exactly the same as the exact
B  B  B solution.
→ ACO 2 + H 2 O + (ε − 1)  A +  O 2 + ε  A +  3.76N 2
2  4  4 Using the LHV of CH4, the rate of heat transfer

is equivalent to the product of gas flow rate and
By definition, lower heating value. One uses the LHV because
they cannot claim any sensible heat from the for-
Lower heating value = ∑moles†of†product mation or condensation of water vapor during
combustion
× (∑h(Tad) − ∑h(70))
products
Q = m gasLHV = m air Cp ∆T

Gas H(70°F) (Btu/lbmole) H(250°F) (Btu/lbmole) Btu Btu
8.82 × 10 6 † †(LHV) = m
 gas × 21, 495† →m
 gas
h lbm
N2 3680.8 4934.6
O2 3676.1 4951 lb
= 410 † m
CO2 3968.3 5650 h
H2O 4202.2 5657
Average specific heats can be used for up to
LHV = CO 2 (1681.7 ) + 2H 2 O(1454.8) around 1000°F (540°C) with fair accuracy for air
heater applications.
+ (ε − 1)(2)O 2 (1274.9) + ε(2)3.76N 2 (1253.8) In the case of air at 70°F (21°C) to 1800°F (982°C),

it becomes more difficult since the specific heat
With, LHV = 21,495 Btu/lb m × 16 lb m/lbmole = will vary greatly at the inlet and outlet in a non-
344 kBtu/lbmole, solving for ε = 28.54 and EA = 2755%. linear way. The exact solution is shown next, and
If the H2O in the products is in the gas phase as then shortcuts are examined.
in this case, use LHV. One generally needs to obtain
the enthalpy for the products and reactants. In this
Gas h(70°F) (Btu/lbmole) h(1800°F) (Btu/lbmole)
case, the reactants are at ambient temperature and
can be ignored for flame calculations. Generate a N2 3680.8 16,626
chart of enthalpy at 70°F and 250°F and write the O2 3676.1 17,375.8
corresponding thermodynamic equation as follows: CO2 3968.3 24,512.10
H2O 4202.2 20,144
200, 000 lbm †air †lbm †mole†air lbm †mole†fuel
× ×
h 28.84 lbm †air 28.54 × 2 × 4.76 lbm †mole†air
LHV = CO 2 (20, 543.8) + 2H 2 O(15, 942.6)
16 †lbm †fuel† 408.4 †lbm fuel 408.2 lbm fuel
× = + (ε − 1)(2)O 2 (13, 699.7 ) + ε(2)3.76N 2 (12, 945)
†lbm †mole†fuel h h
21, 495 Btu
× = 8.78 million Btu/h LHV Btu lbm kBtu
lbm fuel With LHV = 21, 495 × 16 = 344
lbm lbmole lbmole
This is also a case of highly diluted flows where
the heat input may be approximated just using Solving for ε = 2.557 and EA = 155.7%
constant specific heat and ignoring the mass of
the fuel input. 200, 000 lbm †air lbm †moleair lbm †molefuel
For relatively low temperature changes and × ×
h 28.84 lbm †air 2.557 × 2 × 4.76 lbm moleair
highly diluted flows, one can assume the specific
heat of air will not vary greatly from operation at 16 lbm fuel 4, 558 lbm fuel
ambient temperature. So one can use an average × =
lbm molefuel h
specific heat value for air, approximately 0.24. The
rate of heat transfer will be
4, 558 lbm †fuel 21, 495Btu
lbm Btu × = 98 millionBtu/h
Q=m
 air C p ∆T = 200 × 10 3 × 0.24 × (250°F − 70°F) h lbm †fuel
h lbm ⋅ °F
Btu
= 8.64 × 10 6 (LHV) Applying the dilution assumption and ignoring
h the fuel mass flow can be done in several ways,
Heat Transfer 135
first by just using the gas tables with air at 70°F The enthalpies for each gas at 1000 and 1440°F are
and 1800°F (21°C and 980°C): in the table given herewith:
lbm Btu Gas h(1000°F) (Btu/lbmole) h(1440°F) (Btu/lbmole)

200 × 10 3 × (577.51 − 126.66)
h lbm
N2 10,347 13,741
Btu O2 10,693 14,322
= 90.2 × 10 6 †(LHV)
h CO2 14,081 19,697
H2O 12,178 16,427
Another approximation can be found by using
the average specific heats between 1800°F Inserting enthalpies and solving for ϕ
and 70°F (980°C and 21°C). (0.29 + 0.24) ×
0.5 = 0.265 Btu/lbm · °F. These approximations
2(φ − 1)h(O 2 )1440 + 2φ6.25h(N 2 )1440 + [ 2φ(0.341) + 1] h(CO 2 )1440
will always be lower than the actual, in these
cases by 7%–8%, but provide a good simple + [ 2φ(0.741) + 2] h(H 2 O)1440
reality check.
 2φh(O 2 )1000 + 2φ6.25 ( h(N 2 )1000 ) + 2φ(0.341)h(CO 2 )1000 
lbm Btu − 
200 × 10 3
× 0.265 × (1800°F − 70°F)  + 2φ(0.741)h(H 2 O)1000



h lbm ⋅ °F
Btu Btu lbm
= 91.7 × 10 6 †(LHV) = 21, 495 × 16 LHV
h lbm lbmole

φ = 5.35
Unlike air heaters, duct burners firing TEG have
added H2O and CO2 at high temperatures and the Taking the ratio of TEG to fuel on a volume basis,
long exact solution is almost always required. and later on a mass basis will give the fuel firing rate.
Example 7.6  TEG 

  : (2)(5.35)(1 + 6.25 + 0.341 + 0.741)
TEG in a duct burner is to be heated from F  Vol
1000°F (540°C) to 1440°F (782°C) at 2 × 106
lb/hr (90,718 kg/h). Find the firing rate of the fuel. 89.15†mole†TEG
=
Again, use CH4 on a LHV basis. The volume per- mole†CH 4
centage of TEG is 75% N2, 12% O2, 4.1% CO2, and
8.9% H2O. MWTEG = 28.25
The exact solution will require the use of exact
chemistry to solve for the fuel firing rate. TEG lbmole †TEG 1 mole†CH 4
 TEG 
will vary in composition, so create a general solu-   : 89.15 ×
F  Wt lbmole †CH 4 16 lbm †CH 4
tion to solve for all cases. Given that the composi-
tion of TEG by volume is 12% O2, 75% N2, 4.1% lbm †TEG lb †TEG
CO2, and 8.9% H2O, normalize the TEG to 1 mole × 28.25 = 157 m
mole†TEG lbm CH 4
on an O2 basis.
The fuel firing rate will then be
12 O2 1 O2
75 N2 6.52 N2 lbm †TEG 1†lbm CH 4
m
 Fuel = 2.0 × 10 6 ×
4.1 CO2 0.341 CO2 h 157 lbm †TEG
8.9 H2O 0.741 H2O
lbm CH 4
= 12.7 × 10 3
h
CH 4 + 2φ[O 2 + 6.25N 2 + 0.341CO 2 + 0.741H 2 O]
It is usually difficult to find shortcuts for TEG
→ CO 2 + 2H 2 O + 2(φ − 1)O 2 + 2φ[6.25N 2
applications unless highly diluted. It is instruc-
+ 0.341CO 2 + 0.741H 2 O] tive to look at a shortcut using the high inlet and
outlet temperatures and enthalpy. The average
specific heat cannot be taken and one must rely
Taking the net enthalpy change of the products on gas tables or the enthalpy equations given in
and reactants, will yield the LHV on a lbm basis. the previous section. Using the enthalpies and
highly diluted flow, ignoring fuel input solve for
∑hp − ∑hr
g g = ∆hc(LHV) heat input ignoring changes in composition to
determine error in this method.
 lbm   lbm  heating. Major heat transfer processes in b

oilers or fur-
MWTEG = 0.75  28 + 0.12  32
 lbmole   lbmole  naces include the following:
 lb m   lbm  1. Conduction through the furnace refractory and

+ 0.041  44 + 0.089  18
 lbmole   lbmole  convection from the wall of the furnace to the
surrounding air
lbm
= 28.25 2. Radiation exchange between the flame, the sur-
lbmole
rounding walls, and water tubes
h1 (1000°F) 3. Convection from the hot furnace gases to the
Btu Btu water tubes and from water tube walls to the
= 0.75(10, 347 ) + 0.12(10, 693) fluid flowing through the tubes
lbmole N 2 lbmole O 2
Btu Btu The consequences of the performance of these heat
+ 0.041(14, 081) + 0.089(12, 178)
lbmole CO 2 lbmoleH 2 O transfer mechanisms may significantly impact product
Btu 1 lbmole Btu throughput and quality, furnace efficiency, equipment
= 10, 704 × = 379 lifetime, and safety. Other critical phenomena for con-
lbmole 28.25 lbm lbm
sideration may include the effect of heat transfer mech-
h2 (1440°F)
anisms on the fired equipment itself (e.g., heat transfer
effects on burner fuel tips), or the effect of heat transfer
Btu Btu on the performance of the fired equipment with respect
= 0.75(13, 741)† + 0.12(14, 322)
lbmole †N 2 lbmole O 2 to NOx emissions, flame stability, and flame shape.
Btu Btu
+ 0.041(19, 697 ) + 0.089(16, 427 )
lbmole CO 2 lbmole H 2 O
7.4.1 Conduction
Btu 1 lbmole Btu
= 14, 293 × = 506 Heat transfer via conduction refers to the transfer of
lbmole 28.25 lbm lbm
energy from a more energetic to less energetic state due to
The heat transfer and gas rate will then be
particle interactions and can occur in solids, liquids, and
gases that are either stationary or moving. In gases and
Q=m
 Fuel × LHV = m
 TEG [506 − 379] liquids, energy is transferred through random collisions
Btu
by molecules—also called diffusion of energy. In liquids,
= 254 × 10 6 there tend to be stronger and more frequent molecular
h
interactions from the more closely packed structure. For
lbm conduction in solids, energy is transferred due to a com-
m
 = 11.8 × 10 3
h bination of wave induced lattice motion and translational
and error of about 7%, but suitable as a cross movement of free electrons. In a solid conductor, energy
check. transfer is governed by motion of free electrons and vice-
versa for nonconductor solids. As indicated with the sec-
ond law of thermodynamics, heat flows in the direction
of decreasing temperature, from hot to cold bodies. The
total heat conduction is calculated through Equation 7.4:
7.4 Modes of Heat Transfer
dT
Heat is a form of energy upon which the majority of all Q cond = − kA (7.4)
dx
combustion processes are based. Heat transfer is a sci-
ence that seeks to understand and predict the energy where
transfer between masses, resulting from differences in A is the normal area
temperature.8,9 Heat transfer is commonly divided into dT is the temperature gradient
three mechanisms or modes for classification: conduc- dx is the thickness
tion, convection, and radiation. These mechanisms all k is the thermal conductivity
have importance as applied to combustion processes.
Consider a typical boiler that consists of tubes with flow-
7.4.1.1 Thermal Conductivity
ing fluid to be heated, and a burner, or group of burners,
designed to provide the required energy for the desired Thermal conductivity is a measure of the ability of
process, a radiant section and a convection section for a material to conduct heat. It is defined as the rate of
Heat Transfer 137
heat transfer per unit area per temperature difference, insulating materials with low conductivity and heat con-
expressed as Btu/(h · ft · °F) or W/(m · K). Higher values of duction. The opposite proves to be true for metals, which
thermal conductivity in a material represent good heat tend to have high conductivity and heat conduction.
conduction while lower values are representative of poor Thermal conductivity of pure metals varies as a func-
heat conductors or insulators. Properties and ranges for tion of temperature, usually decreasing as temperature
typical materials can be seen in Tables 7.5 and 7.6. increases. Alloys are similar to metals, but may increase
Generally speaking, solids have generally higher or decrease with an increase in temperature. Nonmetallic
thermal conductivity relative to liquids and liquids liquids behave similarly to metals. Lastly, gases see an
higher than gases. Among solids, nonmetals tend to be increase in conductivity with temperature, but decrease
as molecular weight increases. For calculation purposes,
TABLE 7.5
thermal conductivity is taken to be constant at the aver-
Thermal Conductivity of Common Materials age material temperature. More detailed information on
Material Btu/h ft °F W/m °C thermal conduction heat transfer is available in the books
specifically written on that subject.10–15
Gases at atmospheric pressure 0.004–0.70 0.007–1.2
Insulating materials 0.01–0.12 0.02–0.21
Nonmetallic liquids 0.05–0.40 0.09–0.70 Example 7.7
Nonmetallic solids (brick, stone, concrete) 0.02–1.5 0.04–2.6
Determine the heat flux/ft2 of a 2 in. Firebrick
Liquid metals 5.0–45 8.6–78
wall (0.58 Btu/(h · ft · °F)) with a temperature dif-
Alloys 8.0–70 14–121 ference from outer to inner wall of 500°F.
Pure metals 30–240 52–415 Using Fourier’s law of conduction,
Btu 500°F 12″ Btu
Press, Boca Raton, FL, 2001. Q = 0.58 × × = 1740 2
h ft ∞F 2″ 1†ft ft

TABLE 7.6
Properties of Various Substances at above 32°F (0°C)
(Except for Steam as Noted in the Table Herewith Provided)
7.4.1.2 One-Dimensional Steady State Conduction
ρ cp k In the case of one-dimensional steady state conduction,

heat is transferred through only one direction. Heat
lb/ft3 Btu/lb °F Btu/h ft °F
transfer in all other directions is deemed negligible or
Metals zero. Steady state is reached when the temperature at
Copper 559 0.09 223 every point of the system stays constant over time and
Aluminum 169 0.21 132 is valid when analyzing a furnace wall under steady
Nickel 556 0.12 52 state operating conditions. This does not apply during
Iron 493 0.11 42 the startup and shutdown phases, where heat input and
Carbon steel 487 0.11 25 temperature gradients change with respect to time.
Alloy steel 18Cr 8Ni 488 0.11 9.4 For any isotropic material, the general heat conduc-
Nonmetal solids tion equation in three dimensions is written as
Limestone 105 ∼0.2 0.87
Glass pyrex 170 ∼0.2 0.58 ∂  ∂T  ∂  ∂T  ∂  ∂T   ∂T
Brick K-28 27 ∼0.2 0.14 k + k  + k  + Q = ρcp (7.5)
∂x  ∂x  ∂y  ∂Y  ∂z  ∂Z  ∂t
Plaster 140 ∼0.2 0.075
Kaowool 8 ∼0.2 0.016
where
Gases
k. is the thermal conductivity
Hydrogen 0.006 3.3 0.099
Q is the rate of heat generation in system
Oxygen 0.09 0.22 0.014
ρcp∂T/∂t is the time rate of change of sensible energy
Air 0.08 0.24 0.014
Nitrogen 0.08 0.25 0.014
Steama 0.04 0.45 0.015
7.4.1.2.1 Plane Wall
Liquids Using the heat conduction equation in the case of a
Water 62.4 1.0 0.32 plane wall, steady state conditions and constant thermal
Sulfur dioxide (liquid) 89.8 0.33 0.12 conductivity simplifies Equation 7.5 to
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook,
CRC Press, Boca Raton, FL, 2001. ∂ 2T ∂ 2T ∂ 2T
+ + = 0 (7.6)
∂x 2 ∂y 2 ∂z 2
a Reference temperature for steam is 212°F (100°C). All other
temperatures are 32°F (0°C).

T∞,1 qx̋
Ts,1
Ts,2
T∞,2 q˝contact
qx
TA
Hot fluid
T∞,1, h1 Cold fluid
ΔT
x x=L T∞,2, h2
FIGURE 7.5 TB
Plane wall conduction. (From Baukal, C.E. (ed.), The John Zink
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) q˝gap
Furthermore, one-dimensional analysis of Equation

7.6 will yield
A B
d 2T
= 0 (7.7) FIGURE 7.6
dx 2 Temperature drop due to thermal contact resistance. (From Baukal,
Using the plane wall in Figure 7.5, integration of C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca
Equation 7.7 provides a general solution in the form of Raton, FL, 2001.)
dT 7.4.1.2.2 Contact Resistance

= C1 (7.8)
dx In composite systems, the temperature drop across the
and interface between the walls might be appreciable. This
temperature drop caused by the contact resistance, Rt,c,
T ( x) = C1x + Ts1 (7.9) between two solid materials A and B, is illustrated in
Figure 7.6. If the heat flux for a unit area of interface is
C1 is a constant of integration which equals qx in. then the contact resistance can be defined as
Ts 2 − Ts1 TA − TB
C1 = (7.10) Rt", c = (7.14)
L q"x
Substituting Equation 7.10 into 7.9 yields The thermal contact resistance for different combina-
tions of solids is available in standard texts.16
T −T 
T =  s1 s 2  x + Ts1 (7.11)
 L 
7.4.2 Convection
Equation 7.11 will yield the temperature of the plane Convection heat transfer occurs due to fluid movement.
wall for one-dimensional steady state conduction as a A combination of molecular conduction and macro-
function of thickness. To obtain the rate of heat conduc- scopic fluid motion contribute to convective heat trans-
tion from Equation 7.4 substitute dT/dx with Equations fer. Convection takes place adjacent to heated surfaces as
7.8 and 7.9 to get a result of fluid motion past the surface. All convection
processes fall into three categories—natural convection,
(T − Ts1 ) forced convection, and mixed convection.
Q = − kA s 2 (7.12)
L Natural convection occurs when fluid motion is cre-
ated as a result of local density differences alone, with-
This can be simplified to out any mechanical driving force such as a fan or pump.
(T − T ) Theoretical analyses of natural convection require
Q = s1 s 2 (7.13) the simultaneous solution of the coupled equations of
L
motion and energy.
kA
Forced convection can occur when mechanical
where L/kA is the thermal resistance of the plane wall forces from devices such as fans and pumps give
with units of K/W. motion to the fluid. Forced convection is the most
Heat Transfer 139
commonly employed mechanism in the process TABLE 7.7

industries. Schematically, hot and cold fluids, sepa- Typical Convective Heat Transfer Coefficients
rated by a solid boundary, are pumped through the
Condition Btu/h ft2 °F W/m2 °C
heat-transfer equipment; the rate of heat transfer
Air, free convection 1–5 6–30
being a function of the physical properties of the flu-
Air, forced convection 5–50 30–300
ids, the flow rates, and the geometry of the system.
Steam, forced convection 300–800 1,800–4,800
Flow is generally turbulent and the flow duct varies in
Oil, forced convection 5–300 30–1,800
complexity from circular tubes to baffled and finned
Water, forced convection 50–2,000 300–1,200
tubes. Theoretical analyses of forced convection heat
Water, boiling 00–20,000 3,000–120,000
transfer have been limited to relatively simple geom-
etries and laminar flow. Usually, for complicated Source: Baukal, C.E. (ed.), The John Zink Combustion
geometries, only empirical relationships are avail- Handbook, CRC Press, Boca Raton, FL, 2001.
able. However, computational fluid dynamics (CFD); If Ts > T∞ where the surface is hotter than the fluid, then
see Chapter 9, the science of computer modeling flows heat is convected away from the surface by the fluid
and heat transfer, has advanced enough to provide and the surface is cooled. If Ts < T∞ where the surface is
good information based on the available semiem- cooler than the fluid, then heat is convected to the sur-
pirical CFD models.17 Forced convection, heat transfer face by the fluid. This expression is called Newton’s law
coefficients are strongly influenced by the mechan- of cooling. Table 7.7 illustrates the typical range of heat
ics of flow occurring during forced convection heat transfer coefficients for specific fluids.
transfer. Intensity of turbulence, entrance conditions,
and wall conditions are some of the factors that must 7.4.2.2 Composite Wall
be considered for greater accuracy.
Mixed convection, refers to those situations where In industry, different furnace designs are used for dif-
both natural and forced convection are at work. A good ferent heat transfer operation. Design economics require
example would be the convective heat transfer process that these furnaces often have several walls in series to
taking place on the outside surface of a furnace wall increase strength, provide better insulation, or even to
when there is some wind blowing. In the absence of enhance appearance. The one-dimensional steady state
wind, the wall would be cooled purely by natural con- heat transfer analysis can be applied to these cases also.
vection, but with wind, both mechanisms are present Multiple walls in series may be considered to be a com-
simultaneously. A number of books are dedicated spe- posite wall as shown in Figure 7.7.
cifically to convection heat transfer.18–29 These should be The total thermal resistance is evaluated as
consulted for a complete development of the theory and 1 LA LB Lc 1
equations discussed in this section. ∑R = h A + k A + k A + k A + h A (7.16)
t
1 A B C 2
7.4.2.1 Newton’s Law of Cooling T∞,1
Ts,1
As noted before, when a fan or a blower causes the cool- T2
ing, the heat transfer is by forced convection. When the
cooling air is simply moving due to buoyancy forces T3
caused by density difference, the transfer of heat is
Ts,4
taking place by free, or natural, convection. Advection
is energy transfer by bulk motion alone. In convec- LA LB LC
tion, heat transfer takes place due to random molecu- kA kB kC T∞,4
Hot fluid
lar motion and bulk motion combined. Any convective T∞,1, h1 Cold fluid
transfer of heat can be represented by a general heat bal- A B C T∞,2, h2
x
ance equation:
T∞,1 Ts,1 T2 T3 Ts,4 T∞,4

q′′ = h (Ts − T∞ ) (7.15)
1 LA LB LC 1
where h1A kAA kBA kCA h4 A
q″ is the heat flux
FIGURE 7.7
h is the convective heat transfer coefficient Equivalent thermal circuit for a series composite wall. (From
(Ts − T∞) is the temperature difference between the hot Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
fluid and the surrounding cold fluid Boca Raton, FL, 2001.)
An overall heat transfer coefficient U can then be surface temperature is ≤140°F (60°C). The com-
defined posite wall has a ¼ in. thick steel wall (k = 25 Btu/
[h · ft · °F]) with inside air preheat temperature of
−1 600°F (320°C) and ambient temperature of 100°F
1  1   L   L   L   1   (38°C). The outside wall is exposed to free or
U= =   +  A  +  B  +  c  +    (7.17)
Rtot A  h1   k A   kb   kc   h2   forced convection and radiation that can be com-
bined as h = 1.7 Btu/ft2 · °F for this example.
Given the coefficient of heat transfer at the out-
Example 7.8
side wall, take the heat flux equation (7.13) applied
A composite furnace wall has an inside temper- to composite walls as follows:
ature of 3000°F (1600°C) with 9 in. (23 cm) fire-
600 − 100 600 − 140
brick (k = 0.58 Btu/h-ft-°F), 2 in. (5.08 cm) inst. Q= =
brick steel wall (k = 0.05 Btu/h-ft-°F) and 0.5 in. ∆Lins ∆Lsteel 1 ∆L1 L2
∆L
+ + +
(1.27 cm) steel wall (k = 26.5 Btu/h-ft-°F), with A1kins A2 k steel hA3 A1k1 A2 k 2
outside temperature of 120°F (49°C). Determine
the heat flux and temperature drop through For this case, A1 = A2 = A3. Likewise, from
each medium. Example 7.8, the temperature drop through steel
Using the heat flux Equation 7.12, substitute for can be neglected, eliminating the ΔLsteel/A2ksteel term.
the temperature, thickness, and thermal conduc- Isolating for the thickness of insulation, to obtain
tivity, neglecting the frontal area.
∆Lins 1
+
∆T 600 − 100 k h k
q = = ins ⇒ 1 + ins
∑R t
600 − 140 ∆Lins
kins
∆L1h

(3000 − 120)°F Btu
q = 0.029
 9   2   .5  h ⋅ ft ⋅ °F †× †12 in
 (12)1(.58)  +  (12)1(.05)  +  12 (26.5)  1.087 = 1 +
     ( )  Btu ft
∆L1 × 1.7
h ⋅ ft 2 ⋅ °F
3000 − 120
q = ∆L1 = 2.35 "
1.29 + 3.33 + .0015
One would choose to use a 3” thick piece of insu-
obtaining
lation in order to accomplish this.
Btu
q = 623
h ⋅ ft 2 7.4.2.3 Nondimensionalized Numbers
Using this, find the temperature drop through A common practice utilized throughout convection and
each medium. fluid dynamic studies is to limit the number of variables
For firebrick, Kaolin insulation, and steel using a number which lacks dimensions. The most com-
mon numbers used throughout the study of convection
∆T1 are the Nusselt, Reynolds, and Prandtl numbers. Due to
= 623 †⇒ ∆T1 = 805°F
9 complexities relating flow fields, these numbers are used
12(.58) to reduce the complexities of solving these problems.
∆T2 7.4.2.3.1 Nusselt Number
= 623 ⇒ ∆T2 = 2076°F
2
In order to solve problems involving fluid flow, there
12(.05)
must be a signification of the convective forces taking
∆T3 place across the boundary layer. One important mea-
= 623 ⇒ ∆T3 = 0.97°F
.5 sure used to determine the convective effectiveness of
12(26.5) any fluid is the dimensionless number known as the
Nusselt number. The Nusselt number signifies the ratio
The temperature drop through the steel turns of heat transfer through convection to that of conduc-
out to be less than 1°F. Therefore, any temperature tion, which is given by
drop through steel can be said to be negligible.
hLc
Nu = (7.18)
Example 7.9 k
Determine the necessary thickness of windbox A higher Nusselt number would result in a greater
insulation (k = 0.029 Btu/(h · ft · °F) ) such that the amount of heat transfer via convection, while a lower
Heat Transfer 141
Nusselt number would mean a greater amount of heat For this case, laminar flow applies to all Reynolds
is transferred via conduction. numbers less than 500,000 and turbulent for all num-
bers greater than 500,000 but less than 50,000,000.
7.4.2.3.2 Reynolds Number
One of the most important nondimensional numbers Example 7.10
when regarding fluid flow is the Reynolds number. The
Reynolds number is used to determine the flow state of Air flows over a plate at 20 ft/s (6.1 m/s) with an
the fluid—laminar or turbulent—defined as the ratio of ambient temperature of 100°F (38°C). Determine
the coefficient of heat transfer and rate of heat flux
inertial forces to viscous forces and given as
for a 20 ft (6.1 m) flat plate with a surface tempera-
ture of 800°F (450°C).
VLc ρVLc
Re = = (7.19) Because the temperature across the boundary
υ µ layer of the plate varies with distance, one must
take the average temperature in order to deter-
where mine the fluid properties.
V is the average fluid velocity
Lc is the characteristic length Tsurf + Tamb (100 + 800)°F
Tf = ⇒ = 450°F
υ is the kinematic viscosity 2 2
μ is the dynamic viscosity
The properties for air at 450°F under 1 atm pres-
The range for laminar and turbulent flow is governed sure are
heavily on whether the flow is external or internal.
Btu
7.4.2.3.3 Prandtl Number k = 0.02305
h ⋅ ft ⋅ °F
Lastly, the Prandtl number relates the relative thickness
of the boundary layer, expressed as the ratio of diffusiv- ft 2
υ = 4.154 × 10 −4
ity of kinematic viscosity to the diffusivity of heat. The s
Prandtl number is defined as Pr = 0.6953

υ µcp Using the kinematic viscosity of air, one can
Pr = = (7.20)
α k evaluate to find the Reynolds number in order to
determine the flow state.
where
α is the thermal diffusivity ft
20 × 20 ft
cp is the specific heat of fluid VLc s
Re = ⇒ 2 = 9.62 × 10
5
υ −4 ft
4.154 × 10
The Prandtl number can vary between 0.01 and 100,000 s
depending on the fluid properties. Further text will elab-
orate on the many factors affecting these numbers. Since the Reynolds number is greater than
5.0 × 105, the flow is turbulent. The Nusselt num-
ber relation for turbulent flow is
7.4.2.4 Flow over a Flat Plate
Solving problems involving fluid flow over a flat Nu = 0.037 Re 0.8 Pr 1/3

plate starts with identification of fluid properties as a
function of temperature. An important detail to note
Solving for the Nusselt number will yield
would be the difference in temperature differs from
the surface to the outer edge of the boundary layer.
For these cases, the average temperature of the surface Nu = 0.037(9.62 × 10 5 )0.8 (.6953)1/3 ⇒ 2005
and ambient conditions is called the film temperature.
Under this flow, the Nusselt number correlations are Taking the relation between Nusselt number and
two: heat transfer coefficient, to obtain
Nu = 0.664Re 0.5Pr 1/3 (Laminar) (7.21) Btu

2005 × 0.02305
hL h ⋅ ft ⋅ °F = 2.31 Btu
Nu = c ⇒ h =
Nu = 0.037 Re 0.8 Pr 1/3 (Turbulent ) (7.22)
k 20 ft h ⋅ ft 2 ⋅ °F
With the coefficient of heat transfer, one can solve Sieder–Tate equations are applicable for both the cases
for the rate of heat flux using Newton’s law of of uniform surface temperature and heat flux condi-
cooling. tions. Petukhov32 has given a correlation that gives more
accurate results than the Dittus–Boelter or Sieder–Tate
Btu
q = h(Ts − Tamb ) ⇒ 2.31 (800 − 100)°F equations, but is more complex to use. The correlation is
h ⋅ ft 2 ⋅ °F
= 1620
Btu ( f /8)ReD Pr
Nu D = (7.27)
h ⋅ ft 2 1.07 + 12.7( f /8)1/2 (Pr 2/3 − 1)
7.4.2.5 Turbulent Internal Flow where the friction factor, f, is obtained from the Moody
diagram.33
Internal or conduit flow is a flow field in which the fluid
completely fills a closed stationary duct. On the other
Example 7.11
hand, external or immersed flow is where the fluid flows
past a stationary immersed solid. With internal flow, the Determine the exit temperature and rate of heat
heat transfer coefficient is theoretically infinite at the transfer for air flow in a pipe at 15 ft/s (4.5 m/s)
location where heat transfer begins. The local heat trans- with an inlet temperature of Ti = 150°F (65°C).
fer coefficient rapidly decreases and becomes constant, The pipe has a diameter of 1.75 in. (4.45 cm) and
constant surface temperature Ts = 80°F (26°C).
so that after a certain length the average coefficient in the
Assume fully developed flow.
conduit is independent of the length. The local coefficient Similar to the example provided in convective
may follow an irregular pattern, however, if obstructions flow over a flat plate, find the bulk temperature of
or turbulence promoters are present in the duct. the fluid. However, since the exit temperature is
Since the analysis of turbulent flow heat transfer is unknown, guess for Te = 110°F (43°C).
quite complex, calculations must rely upon empirical
correlations. The Chilton–Colburn analogy provides an Ti + Te 150 + 110
important correlation for the Nusselt number in turbu- Tb = ⇒ = 130°F
2 2
lent flow heat transfer:
The fluid properties of air at 1 atm and 130°F are
Nu D f
StPr 2/3
= Pr 2/3 = (7.23)
ReD Pr 8 ρ = .06727 lbm/ft3
Cp = 0.2405 Btu/lbm · °R
where St is Stanton number and f is the friction factor, k = .01599 Btu/h · ft · °F
given by υ = 1.981 × 10−4 ft2/s
Pr = 0.7216
f = 0.184(ReD )−1/5 (7.24)
With the kinematic viscosity, evaluate for the
Thus, Reynolds number.
Nu D = 0.023 (ReD )4/5 (Pr)1/3 (7.25) ft  1 ft 

15 × 1.75  
VD s  12 in.
Re = ⇒ = 11, 042
Dittus and Boelter30 suggested a modification in υ ft 2
1.981 × 10 −4 †
Equation 7.25 by replacing the exponent of Prandtl num- s
ber by n, where n is 0.4 for heating and 0.3 for cooling.
It is to be noted that Equation 7.25, or its modification, Since the Reynolds number is greater than 10,000,
is good for the cases where the temperature difference the flow is considered turbulent. The Nusselt cor-
relation for turbulent flow is
(Ts − Tm) is moderate. When the temperature difference
is large, the equation suggested by Sieder and Tate31 is 4 1
 4
 1
recommended as follows: Nu = 0.023 Re 5 (Pr) 3 ⇒ 0.027  11, 042 5  (0.7216) 3 = 42

 

4 1 0.14
 µ With the Nusselt number, the coefficient of heat
Nu D = 0.027(ReD ) (Pr)  
5 2 (7.26)
transfer is
 µs 
where μs is the viscosity of the fluid determined at sur- Nu × k 42(0.01599) Btu

h= = = 4.55
face temperature and all the other properties are mea- D 1.75 h ⋅ ft 2 ⋅ °F
sured at the mean temperature. The Dittus–Boelter and 12
Heat Transfer 143
Since the exit temperature is unknown, solve for 7.4.3.1 Radiation from Surface to Ambient
it using the given relation:
The radiative heat transfer from a flat plate to the ambi-
ent environment is a practical application for hot sur-
Q = hA∆TLM = mC
 p ∆T
faces of boilers and other heater casing surfaces. From
first principles34,35
where
Ti − Te
∆TLM =
T − Te
ln s
( )
Q = Asσε s Ts4 − Ta4 (7.28)
Ts − Ti
where
∆T = Te − Ti Q is the radiant heat flow in Btu/h
As is the area of surface
Using an iterative approach, the exit temperature εs is the average emissivity surface
will be approximately 90°F. Hence, the rate of heat Ts is the average temperature of hot surface in°R
transfer will be Ta is the average temperature of ambient in°R
σ = 1.714 †× †10 −9 Btu/h ⋅ ft 2 †⋅ R 4
 
  1.75   150 − 90 
Q = hAs ∆TLM = 4.55 ×  × π × 1 ft  Example 7.12
 12  80 − 90 
 ln 
80 − 150  Find the radiative heat flux given surface emissiv-
ity = 0.8, ambient temperature = 100°F (37°C), and
Btu surface temperature = 140°F (60°C).
= −64
h Substituting into Equation 7.28
Q
7.4.3 Radiation As
( )
= σε s Ts4 − Ta4 = 1.714 × 10 −9 (0.8)
Thermal radiation heat transfer is the movement of Btu

energy by electromagnetic waves. Quantum the- (
× (140 + 460)4 − (100 + 460)4 ) h-ft 2
ory describes electromagnetic energy as photons or

= 42.8Btu/h-ft 2
quanta. Unlike conduction or convection, radiation
does not require any intervening medium for transfer. or the equivalent h = 42.8/40 = 1.07 Btu/h-ft2 that
Electromagnetic radiation, in the wavelength range of is sometimes easier to use when adding forced or
0.1–100 μm, is produced solely by the temperature of convective terms to casing losses. Using both radia-
a body. Energy at the body’s surface is converted into tion and forced convection, create a casing heat flux
loss chart shown in Figure 7.8.
electromagnetic waves that emanate from the surface
and strike another body. Some of the thermal radiation
is absorbed by the receiving body and reconverted into 7.4.3.2 Radiation from Nonluminous,
internal energy, while the remaining energy is reflected Nonreacting Gases to Surface
from or transmitted through the body. The fractions of This section on radiation will concentrate on the most dif-
radiation reflected, transmitted and absorbed by a sur- ficult subject, the heat transfer from hot gases and react-
face are known as reflectivity ρ, transmissivity τ, and ing flames to absorbing surface. The subject is immensely
absorptivity α, expressed as ρ, τ, and α respectively. The complex involving radiation from a reacting flame that
sum of these fractions equals one. Thermal radiation is can have significant luminosity, partial absorption, reflec-
emitted by all surfaces whose temperatures are above tion, transmission from the reflecting gas, wavelength
absolute zero. dependence, etc. To remain focused with something use-
Thermal radiation can pass through some gases like ful for an engineer to use, the process has been simpli-
air without absorption taking place. So these gases do fied. If the surface emissivity is high, the recommended
not affect radiative transfer. On the other hand, gases method is described by Hottel and Sarofim:34
like CO2, H2O, and CO affect radiation somewhat and
are known as participating gases. These participating (ε s + 1)
gases are, of course, common constituents of boiler flue Q = Acpσ
2
( )
ε gTg4 − α gTw4 (7.29)
gases and as such, play a significant part in the trans-
fer and distribution of heat to the heater tubes. Practical where
applications will be presented in this section from Q is the radiant heat flow in Btu/h
simple wall to atmosphere and radiation from reacting Acp is the area of equivalent cold plane absorbing
flames in simple furnaces. surface
Casing heat flux versus ambient temperature

4000
3500
3000 5 fps
10 fps
20 fps
2500 50 fps
Heat flux (btu/h ft2)
2000
1500
1000
500
0
80 130 180 230 280 330 380
Casing temperature (°F)
FIGURE 7.8
Calculated casing heat flux losses using 0.8 emissivity for various exterior velocities.
σ = 1.714 †× †10 −9 Btu/h-ft 2†R 4 Figures 7.9 and 7.10 show the emissivity of H2O and
εs is the emissivity of cold plane CO2, respectively, as a function of gas temperature. The
εg is the average emissivity of gas flame figures are based on experimental data taken in hemi-
Tg is the average temperature of gas flame °R spherical shape of the gas at 1 atm total pressure, in a
σg is the absorptivity of gases at cold plane temperature mixture with nonradiating gases. For pressure other
Tw is the average temperature of the cold plane wall than 1 atm, corrections must be incorporated. When
CO2 and H2O both appear together with nonradiating
gases, the total emissivity of the gas is obtained by
To compute the heat transfer, determine the emis-
sivity of the hot gases. In most combustion calcula- ε g = ε w + ε c − ∆ε (7.30)
tions, the dominant nonluminous radiation is from
H2O and CO2. The total emissivity is a function of the where the correction factor Δε can be obtained from
partial pressure, temperature, and mean beam length. Figure 7.11. Normal emissivity of various surfaces is tabu-
Gases emit and absorb radiation in discrete energy lated in Table 7.8. The mean beam length, Le, can be defined
bands dictated by the allowed energy states within as the radius of a hemispherical gas mass whose emissivity
the molecule. While the energy emitted by a solid is equivalent to that for the geometry of interest. Table 7.9
shows a continuous spectrum, the radiation emitted gives the mean beam length of numerous gas geometries
and absorbed by a gas is restricted to bands. The emis- and shapes from Hottel.34 For geometries not covered in
sion of radiation for these gases occurs in the infrared Table 7.9, the mean beam length can be approximated as
region of the spectrum. The inert gases and diatomic
gases of symmetrical composition such as O2, N2, and 3.4( Volume)
Le = (7.31)
H2 are transparent to thermal radiation. Important (Surface†area)
gases that absorb and emit radiation are polyatomic
gases such as CO2 and H2O and asymmetric molecules Using mean beam length L e instead of L (the radius of
such as CO. hemisphere), gas emissivity is obtained, which in turn
Heat Transfer 145
0.7
0.6
0.5
0.4 pw L = 20 ft-atm
10
0.3
5
3
0.2 2
1.5
1.0
Gas emissivity, εc of water vapor
0.6
0.1 0.4
0.08 0.3
0.2
0.06
0.15
0.05
0.10
0.04
0.08
0.03 0.06
0.05
0.04
0.02 0.03
0.02
0.015
0.01
0.01 0.007
0.08 0.005
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Temperature (R)
FIGURE 7.9
Total emissivity of water vapor at reference state. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
gives radiant heat transfer to a surface due to emission 0.65

 Tg 
from an adjoining gas: α c = Cc  
 Ts 
( )
ε c Ts , pc LeTs /Tg (7.35)
q = ε g AsσTg4 (7.32) When both CO2 and water vapors are present, the total
gas absorptivity is obtained by
where As is the surface area. The net radiation exchange α g = α w + α c − ∆α (7.36)
rate between the surface at temperature Ts and the gas at
Tg can be expressed as where
∆α = ∆ε

qnet = Asσ ε gTg4 − α gTs4 ( ) (7.33)
Example 7.13
The relationship for determining absorptivity of CO2 For a typical package boiler of 120,000 lb/h steam,
and H2O from their respective emissivity data was find the radiative heat flux in a boiler volume of
given by Hottel34 as follows: 7 × 9 × 24 ft that is all water wall equivalent cold
plane, given the cold plane absorbing surface
0.45 emissivity = 0.85, hot gas temperature of 2400°F,
 Tg 
α w = Cw  
 Ts 
(
ε w Ts , pw LeTs /Tg ) (7.34) and wall temperature of 500°F. The products of
combustion are from methane at 9% CO2, 18%
0.3
0.2
pw L = 5.0 ft-atm
3
0.1
2
1
0.08
0.6
Gas emissivity, εc, of carbon dioxide
0.06 0.4
0.05 0.2
0.04 0.10
0.06
0.03
0.04
0.02 0.02
0.010
0.008
0.006
0.01
0.005
0.008 0.004
0.003
0.006
0.002
0.005
0.004 0.001
0.003
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Temperature (R)
FIGURE 7.10
Total emissivity of carbon dioxide at reference state. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL,
2001.)
H2O. Determine heat absorbed by radiation and Compute absorption using (7.34) and (7.35):
net radiative heat transfer. 0.45
To find the mean beam length, one can use the  2860 
αw =  × 0.3 = †.
0 49
volume and surface area along with Equation 7.31.  960 

0.65
Surface area = 894 ft2, volume = 1572 ft3, and  2860 
αc =  × †0.12 = †.
0 24
mean beam length = 6 ft  960 

Using Figures 7.9 and 7.10 at 2860°R Net absorption at 500°F = 0.49 + 0.24 − 0.04 = 0.69
Use Equation 7.28 to compute heat flux:
For H2O: PwL = 0.18 × 6 = 1.08, εw ≈ 0.18
Q (0.85 + 1)
For CO2: PcL = 0.09 × 6 = 0.54, εc ≅ 0.095 = 1.714 †× †10 −9 †(0.235 × 2, 860 4 − 0.69 × 960 4 )
Sum = 0.275 next compute correction, Acp 2
L(Pw + Pc) = 1.62 from Figure 7.11, Δε = 0.04 Btu
= 23, 880
ft 2
Net εg = 0.235 for nonluminous hot gas emissiv-
Total radiative heat transfer = 894 × 23,880 = 21
ity at 2400°F. Repeat to find CO2 and H2O emissiv-
million Btu/h.
ity at 500°F.
For H2O: PwL = 0.18 × 6 = 1.08, εw ≈ 0.3 Example problem 7.13 is accurate for illustrative pur-
For CO2: PcL = 0.09 × 6 = 0.54, εc ≅ 0.12 poses. In most real applications the flame reacts in a fur-
Net εg = 0.42, wall gas emissivity at 500°F nace and the temperature varies along the flame path.
Heat Transfer 147
0.07
Tg ≈ 540°C Tg ≈ 930°C
Tg ≈ 125°C
0.06
L (pw + pc) = 5.0 ft-atm
0.05 3.0
Mixture correction Δε
2.0
1.5
0.04
1.0
0.03 3.0 3.0 0.75
2.0
2.0 1.5
0.02
1.5 1.0 0.5
0.75
0.01 1.0
0.75 0.5
0.5 0.3
0.3
0.3 0.2 0.2 0.2
0.00
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
FIGURE 7.11
Correction for water vapor and carbon dioxide. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Secondly, some convection contributes to the overall heat

transfer. Lastly, most flames have additional emissivity
due to luminosity caused by chemiluminescences and soot
within the flame. Although this problem seems immensely
TABLE 7.8
complicated, one can add some simple improvements to
Normal Emissivities, ε, for Various Surfaces arrive at answers closer to reality. The best method is to
Material Emissivity, ε Temp. (°F) Description use computational methods, although even CFD solutions
Aluminum 0.09 212 Commercial sheet
are currently limited by poor luminosity models.
Aluminum 0.63–0.42 530–930
oxide 7.4.3.3 Radiation from Nonluminous
Aluminum 0.27–0.67 212 Varying age and Reacting Flames to Surface
paint Al content
Historically, methods to resolve the issue of react-
Brass 0.22 120–660 Dull plate
ing flames have centered on the relationship between
Copper 0.16–0.13 1970–2330 Molten
Copper 0.023 242 Polished
TABLE 7.9
Cuprous 0.66–0.54 1470–2012
oxide Mean Beam Lengths Le for Various Gas Geometries
Iron 0.21 392 Polished, cast
Characteristic
Iron 0.55–0.60 1650–1900 Smooth sheet Geometry Length Le
Iron 0.24 68 Fresh emeried
Sphere (radiation to surface) Diameter (D) 0.65D
Iron oxide 0.85–0.89 930–2190
Infinite circular cylinder Diameter (D) 0.95D
Steel 0.79 390–1110 Oxidized at
(radiation to curved surface)
1100°F
Semi-infinite circular cylinder Diameter (D) 0.65D
Steel 0.66 70 Rolled sheet
(radiation to base)
Steel 0.28 2910–3270 Molten
Circular cylinder of equal height Diameter (D) 0.60D
Steel (Cr-Ni) 0.44–0.36 420–914 18–8 rough, after and diameter (radiation to
heating entire surface)
Steel (Cr-Ni) 0.90–0.97 420–980 25–20 oxidized in Infinite parallel planes Spacing between 1.80L
service (radiation to planes) planes (L)
Brick, red 0.93 70 Rough Cube (radiation to any surface) Side (L) 0.66L
Brick, fireclay 0.75 1832 Arbitrary shape of volume V Volume to area 3.6(V/A)
Carbon, 0.945 100–700 0.003 in. or thicker (radiation to surface of area A) ratio (V/A)
lampblack
Source: Hoogendoorn, C.J. et al., J. Inst. Fuel, 43, 511, 1970.
Water 0.95–0.963 32–212
With permission.
furnace exit temperature and equivalent flame tempera- nonluminous emissivity and absorptivity. Using the
ture based on the fact that this equivalent temperature methods outlined in Chapter 3 and in this chapter,
must be between the adiabatic flame temperature and one can determine the adiabatic flame temperature
the furnace exit gas temperature.36–38 to be 3352°F (1844°C). Also, the combustion products
are found to be 9% CO2, 18% H2O, 3% O2, and 70% N2
Eight correlations were reviewed36 and the correlation
at 126,750 lb/h. An estimate of velocity yields a con-
proposed by Long 39 seemed to be best suited and it is vection coefficient of 4 Btu/°F ft2 using Equation 7.39.
an average of the adiabatic flame and furnace exit tem- Next, h(t) is required. Referring to the curve fitted
peratures as follows: enthalpy equations in Table 7.1 one can c ompute
∑
4
the enthalpy function as hmix (t) = f i hi (t) by
DT = 0.25 × (Tad − Tfeg ) (7.37) 1
multiplying the polynomial coefficients times the

where volumetric fractions and sum. Finally, one needs
DT is the equivalent flame temperature—furnace exit to compute the MW of the products from the meth-
gas temperature ods of Chapter 3 and divide the resultant polyno-
Tad is the adiabatic flame temperature mial. The resultant enthalpy equation in Btu/lbm is
Tfeg is the furnace exit gas temperature
h(t) = −1.8162 × 10 −9 T 3 + 2.645 × 10 −5 T 2
Equation 7.29 can be rewritten to include convection,
DT, and equated to heat lost by combustion gases in + 0.23397T + 5.1794

terms of furnace heat flux. Further, there is the added
feature of estimating furnace exit gas temperature.
Equation 7.38 is now in terms of Tfeg only. It can
( )
Q (ε + 1) be easily solved with any iterative solution. The
( )
4
= σ s ε g 0.25x(Tad − T feg ) + T feg − α g Tw4 heat flux is 46,836 Btu/h-ft2 = 41.87 × 106 Btu/h =
Acp 2
@ 28.6% of the firing rate absorbed. The FEGT is
(( )
+ H 0.25x(Tad − T feg ) + T feg − Tw ) 2456°F (1347°C) and the equivalent hot gas tem-
perature is 2690°F (1477°C). For complete accu-
racy, one would need to redo the problem based
( )
= h(Tad ) − h(T feg ) M fg (7.38)
on the new equivalent hot gas temperature, but as
will be seen, the accuracy of the emissivity is still
where not corrected for luminosity.
H = (1 + 2.7γ fg Vfg ) (7.39)
7.4.3.4 Radiation and Convection from
Furnace wall convection coefficient40 Reacting Flames to Surfaces
h(t) is the enthalpy of gases as a function of tem- Up until now, the classical approach of flame to surface
perature Btu/lbm radiation excluding the effect of flame luminosity has
Mfg is the mass flow of gases lbm/h been covered. The history of computational methods
has produced a mixed success and is very dependent
The application of Equation 7.38 is best illustrated by an on burner type, fuel air fixing, soot particles, and fuel
example. properties. Hottel and Sarofim34 discuss a review start-
ing with a radiation model for clouds of soot/particles.
Example 7.14 Most of these models are based on the gray gas expres-
sion of the form
Repeat Example 7.13. The products of combus-
tion are from methane. The firing rate is 146 × 106
Btu/h firing methane at 15% EA. Determine heat ε s = 1 − e − kL (7.40)
absorbed by radiation and net radiative heat trans-
fer. The furnace exit gas temperature is unknown where
and the average flame temperature is unknown. εs is the emissivity due to soot
The boiler efficiency is 83% HHV. Determine the kL is the path length × effective soot-emission parameter
furnace exit gas temperature, heat flux to the wall,
and total heat absorbed in the radiant section of
the furnace. Hottel and Sarofim concluded that soot models
Example 7.13 was selected as a starting point i nvariably underpredict actual measured flame emissivi-
because the equivalent flame temperature does not ties. One example predicts the addition of 0.22 for soot
differ much from reality. So, select 2400°F (1300°C) emissivity for oil flames. Further references site natural
for the hot gas first guess for simplicity and retain the gas flame emissivity’s that can range from 0.2 to 0.6 for
Heat Transfer 149
highly luminous natural gas flames. They recommended For special applications such as substoichiometric fir-
adding 0.1 to natural gas flames to account for luminosity. ing, the soot contribution will increase significantly and
Godridge and Hammond,41 made detailed measure- additional factors will need to be added that are beyond
ments of a large residual oil flame and correlated to a the scope of this chapter.
soot model similar to Hottel and Sarofim. They found Repeat Example 7.14 and correct for luminosity. The
emissivities in large residual oil flames to vary from products of combustion are from methane. The firing rate
about 0.9 at the base of the flame to 0.45 at the flame is 146 × 106 Btu/h firing methane at 15% EA. Determine
ends. Arscott and Godridge42 found similar emissivities heat absorbed by radiation and net radiative heat trans-
in residual oil flames. fer. The furnace exit gas temperature is unknown and
A correction to nonluminous gas emissivity was mea- the average flame temperature is unknown. The boiler
sured for a large tangentially fired utility furnace.43 The efficiency is 83% HHV. Determine the furnace exit gas
correction factors were determined to be about 1.2 for temperature, heat flux to the wall and total heat absorbed
natural gas and 1.4–1.55 for oil firing. in the radiant section of the furnace.
A significant improvement over the simple gray gas
soot models was developed by Yuen and Tien.44 They Solution
( )
developed an expression for total emissivity for both Q (ε + 1)
( )
4
solid and gaseous fuels using gray gas and gray soot = σ s (ε g + LF ) 0.25x(Tad − Tfeg ) + T feg − α gTw4
Acp 2
models. Unfortunately, the practical application of this
model is limited due to the basic flame parameters that
are required such as, size distribution of soot, volume
((
+ H 0.25x(Tad − T feg ) + T feg − Tw ) )
fraction, optical properties of soot, partial pressure of
gases, and temperature distribution. The model can be
( )
= h(Tad ) − h(T feg ) M fg
used for CFD (see Chapter 9) solutions.
Some practical peak emissivity measurements are Example 7.15
listed by Trinks40 as a function of carbon/hydrogen mass As in the last Example 7.14, everything is the same
ratio and shown in Figure 7.12. Based on all the aforemen- except the flame emissivity has been corrected by
tioned details, the writer recommends an addition of 0.15 adding 0.15 for luminosity:
to normal natural nonluminous emissivity predictions.
For oil flames, the emissivity for # 2 oil will be about ε s = 0 .85
0.45–0.55 over the flame zone only. For # 6 oil flames,
ε g = 0.235
the emissivity will be about 0.75 over the flame zone
only. For coal and other highly luminous solid fuels, the
H =4 Btu/°F ft 2
emissivity varies between 0.85 and 0.95.
0.9
0.8
0.7
0.6
Emissivity
0.5
0.4
0.3
0.2
0.1
0
0 2 4 6 8 10 12 14 16 18 20
C/H mass ratio
FIGURE 7.12
Maximum emissivity versus C/H mass ratio. (Adapted from Grigull, U. and Sandner, H., Heat Conduction, Hemisphere, Washington, DC, 1984.)
Tad = 3812°R 8. F. Kreith, R.M. Manglik, and M.S. Bohn, Principles of Heat
Transfer, 7th edn., CL Engineering, Mason, OH, 2010.
Tw = 960°R 9. J.P. Holman, Heat Transfer, 10th edn., McGraw-Hill, New
York, 2009.
LF = 0.15 10. S. Arpaci, Conduction Heat Transfer, Addison-Wesley,
Reading, MA, 1966.
α g = 0.69 11. M.N. Özisik, Boundary Value Problems of Heat Conduction,

Dover, New York, 1968.
M fg = 126, 750 lb/h 12. U. Grigull and H. Sandner, Heat Conduction, Hemisphere,

Washington, DC, 1984.
The equation is, again, in terms of Tfeg only and 13. G.E. Myers, Analytical Methods in Conduction Heat
through substitution and solving. The heat flux is Transfer, Genium Publishing, Schenectady, NY, 1987.
53,617 Btu/h · ft2 = 51.51 × 106 Btu/h = @ 35.2% 14. B. Gebhart, Heat Transfer and Mass Diffusion, McGraw-
of the firing rate absorbed. The FEGT is 2232°F Hill, New York, 1993.
(1222°C) and the equivalent hot gas temperature 15. D. Poulikakos, Conduction Heat Transfer, Prentice Hall,
is 2523°F (1384°C). Based on past observations and Englewood Cliffs, NJ, 1994.
data collection, these results fit reality for the con- 16. F.P. Incropera and D.P. DeWitt, Fundamentals of Heat and
ditions. The luminosity correction increased the Mass Transfer, 4th edn., John Wiley & Sons, New York, 1996.
heat transfer in the radiant section by 23% and 17. C.E. Baukal, V.Y. Gershtein, and X. Li, Computational
decreased the exit temperature by 224°F (124°C). Fluid Dynamics in Industrial Combustion, CRC Press, Boca
Raton, FL, 2001.
18. V.S. Arpaci, Convection Heat Transfer, Prentice Hall,
Englewood Cliffs, NJ, 1984.
19. C.S. Fang, Convective Heat Transfer, Gulf Publishing,
Houston, TX, 1985.
7.5 Conclusions 20. S. Kakac, R.K. Shah, and W. Aung (eds.), Handbook of
Single-Phase Convective Heat Transfer, Wiley, New York,
As per the objectives stated in the Introduction, this 1987.
chapter provided a brief summary of the overarching 21. L.C. Burmeister, Convective Heat Transfer, 2nd edn., Wiley,
topic concerning the fundamentals of thermodynamics New York, 1993.
and heat transfer. The presented material was designed 22. S. Kakac and Y. Yener, Convective Heat Transfer, 2nd edn.,
to reinforce basic principles and equations, which are CRC Press, Boca Raton, FL, 1995.
applicable to combustion. Proper understanding of these 23. P.H. Oosthuizen and D. Naylor, Introduction to Convective
basic concepts will be invaluable for the understanding Heat Transfer Analysis, McGraw-Hill, New York, 1999.
of more advanced topics to be discussed in later chapters. 24. T. Cebeci, Convective Heat Transfer, 2nd Rev. edn.,
Horizons, Long Beach, CA, 2002.
25. A. Bejan, Convection Heat Transfer, 3rd edn., Wiley, New
York, 2004.
26. W.M. Kays, M.E. Crawford, and B. Weigand, Convective
Heat and Mass Transfer, 4th edn., McGraw-Hill, New
References York, 2004.
27. M. Favre-Marinet and S. Tardu, Convective Heat Transfer:
1. B. Gebhart, Heat Transfer, 2nd edn., McGraw-Hill, New Solved Problems, Wiley, Hoboken, NJ, 2009.
York, 1971. 28. M. Ghiaasiaan, Convective Heat and Mass Transfer,
2. Y.A. Cengel, Heat and Mass Transfer: A Practical Approach, Cambridge University Press, Cambridge, U.K., 2011.
3rd edn., McGraw Hill, New York, 2007. 29. M. Kaviany, Principles of Convective Heat Transfer, 2nd
3. G.J.V. Wylen and R.E. Sonntag, Fundamentals of Classical edn., Springer-Verlag, New York, 2011.
Thermodynamics: Series in Thermal and Transport Sciences, 30. F.W. Dittus and L.M.K. Boelter, Publications on
2nd edn., John Wiley & Sons, Inc., New York, 1973. Engineering, University of California, Berkeley, CA,
4. J.H. Keenan and J. Kaye, Gas Tables: Thermodynamic Vol. 2, p. 443, 1930.
Properties of Air products of Combustion and Component 31. E.N. Sieder and G.E. Tate, Heat transfer and pressure
Gases, John Wiley & Sons, Inc., New York, 1948. drop of liquids in tubes, Ind. Eng. Chem., 28(12), 1429–
5. E. Pita, Air Conditioning Principles and Systems: An Energy 1435, 1936.
Approach, Prentice Hall, Upper Saddle River, NJ, 2002. 32. B.S. Petukhov, T.F. Irvine, and J.P. Hartnett (eds.),
6. ASME International Steam Tables for Industrial Use, Advances in Heat Transfer, Academic Press, New York,
2nd edn., American Society of Mechanical Engineers, Vol. 6, 1970.
New York, 2009. 33. L.F. Moody, Friction factors for pipe flow, Trans. ASME,
7. S.C. Stultz and J.B. Kitto, Steam Its Generation and Use, 66, 671–684, 1944.
40th edn., Babcock and Wilcox, Barberton, OH, Fig 11, 34. H. Hottel and A. Sarofim, Radiative Transfer, McGraw-
p. 22.7, 1992. Hill, New York, 1967.
Heat Transfer 151
35. F. Kreith, Principles of Heat Transfer, 3rd edn., IEP- 40. W. Trinks and M.H. Mashinney, Industrial Furnaces, 5th
Dun-Donnelley, New York, 1976. edn., John Wiley & Sons, Inc., New York, 1961.
36. Heat transfer in the furnace chamber of pulverized-fuel- 41. A.M. Godridge and E.G. Hammond, Emissivity of a very
fired water tube boilers, Central Electricity generating large residual fuel oil flame, 12th Symposium International
Board, Marchwood Engineering, Journal of the Institute of on Combustion, p. 1219, 1968.
Fuel, July 1967. 42. J.A. Arscott and A.M. Godridge, Flame data from a large
37. D. Anson, Modeling experience with large boilers, residual fuel oil fired burner, 17th Symposium International
Third Symposium on Flames and Industry, Institute of Fuel, on Combustion, p. 737, 1978.
October 1966. 43. K.A. Bueters, J.G. Cogoli, and W.W. Habelt, Performance
38. M. Fishenden and O. Saunders, An Introduction to Heat prediction of tangentially fired utility furnaces by com-
Transfer, 3rd edn., Oxford University Press, London, puter model, 15th Symposium International on Combustion,
U.K., 1957. p. 1245, 1974.
39. V.D. Long, Estimation of the mean radiating temperature 44. W.W. Yuen and C.L. Tien, A simple calculation scheme
of a cylinder of combustion gases, J. Inst. Fuel, 35, 431, for the luminous flame emissivity, 16th Symposium
1962. International on Combustion, p. 1481, 1976.
8
Fundamentals of Fluid Dynamics
Stephen B. Londerville and Wes Bussman
CONTENTS
8.1 Introduction.................................................................................................................................................................... 153
8.2 Properties of Fluids....................................................................................................................................................... 154
8.2.1 Density................................................................................................................................................................ 154
8.2.2 Specific Heat....................................................................................................................................................... 154
8.2.3 Specific Weight................................................................................................................................................... 154
8.2.4 Viscosity.............................................................................................................................................................. 155
8.2.4.1 Absolute Viscosity............................................................................................................................... 155
8.2.4.2 Kinematic Viscosity............................................................................................................................ 155
8.3 Hydrostatics.................................................................................................................................................................... 156
8.3.1 Pressure............................................................................................................................................................... 156
8.3.2 Absolute and Gage Pressure............................................................................................................................ 157
8.3.3 Pressure Measurement...................................................................................................................................... 157
8.3.3.1 Manometers......................................................................................................................................... 157
8.3.3.2 Bourdon-Tube Gage............................................................................................................................ 158
8.4 Fluid Flow....................................................................................................................................................................... 158
8.5 Euler’s Equation............................................................................................................................................................. 160
8.6 Bernoulli Equation..........................................................................................................................................................161
8.6.1 Static, Velocity, and Total Pressure...................................................................................................................162
8.7 Continuity....................................................................................................................................................................... 163
8.8 Fan and Pump Sizing.................................................................................................................................................... 164
8.9 Pressure Losses through Pipes.................................................................................................................................... 165
8.9.1 Major Losses....................................................................................................................................................... 165
8.9.2 Minor Losses.......................................................................................................................................................167
8.9.3 Series and Parallel Flow.................................................................................................................................... 171
8.9.3.1 Parallel Flow........................................................................................................................................ 172
8.9.3.2 General Parallel Flow..........................................................................................................................174
8.9.4 Discharge Coefficient........................................................................................................................................ 175
8.9.5 Draft......................................................................................................................................................................176
8.10 Compressible Flow........................................................................................................................................................ 178
8.10.1 Mach Number.................................................................................................................................................... 178
8.10.2 Quasi–One-Dimensional Isentropic Flow...................................................................................................... 179
8.10.3 Mach Number Relationships........................................................................................................................... 179
8.11 Free Jet............................................................................................................................................................................. 180
8.12 Conclusions..................................................................................................................................................................... 182
References................................................................................................................................................................................. 182
These advances were brought about without the aid

of any mathematical analysis of fluid mechanics, but
8.1 Introduction
through trial and error and everyday life experiences.
Ancient civilizations were able to solve fluid mechanic Today, fluid dynamics is one of the most important
problems that enabled them to build boats, ships, canals, areas of engineering science. Over the last 100 years,
irrigation systems, bridges, and water lifting machines.1 research into the science of fluid dynamics has grown
153
at an exponential rate. Fluid dynamics is a broad sub- TABLE 8.1

ject because it is an important tool used in many engi- Fluid Properties of Various Gases
neering fields, for example, turbulence, acoustics, and
Gas Molecular Density Ratio of
aerodynamics. Composition Weight (lbm/ft3) Specific Heat
Fluid dynamics plays an important role in the design
Methane (CH4) 16.043 0.04238 1.31
and operation of combustion equipment. For example,
Ethane (C2H6) 30.070 0.07943 1.19
industrial boilers require analysis and control of the flow
Propane (C3H8) 44.097 0.11648 1.13
of air, steam, fuel, and hot combustion products through
Ethylene (C2H4) 28.054 0.07410 1.25
complex networks of pipes, ducts, valves, dampers, reg-
Carbon dioxide 44.010 0.11625 1.29
ulators, pumps, fans, etc. Engineers rely extensively on (CO2)
empirical fluid dynamic models, computational fluid Water (H2O) 18.015 0.04758 1.33
dynamic models, cold flow physical models, and experi- Oxygen (O2) 31.999 0.08452 1.40
ence gained from day-to-day observation of flow dynam- Nitrogen (N2) 28.013 0.07399 1.40
ics to help facilitate and guide in the design of combustion Carbon 28.010 0.07398 1.40
equipment. In general, the more familiar operators are monoxide (CO)
with the flow dynamics associated with their combustion Hydrogen (H2) 2.016 0.00532 1.41
equipment, the better they are equipped to operate and Air 28.970 0.00237 1.40
troubleshoot their systems in a safe and efficient manner. Note: Density is at 59°F (standard temperature) and 14.696 psi.
8.2.2 Specific Heat

8.2 Properties of Fluids By definition, the specific heat of a fluid denotes the
In fluid mechanics, fluid properties are used to charac- heat input required to raise the bulk temperature by
terize the behavior of the given fluid. Many of the topics 1 degree. Similar to the density of fluids, specific heat
covered in previous chapters will be briefly covered and of a liquid is a temperature dependent variable, which
explained in the context of fluid dynamics. tends to decrease as temperature increases. For gases,
the specific heat increases with temperature and the
value is dependent on the processes that the gas is
8.2.1 Density undergoing. Depending on the process, specific heat has
Density is defined as the ratio of mass to volume. For two distinct values Cp and Cv, where Cp is the specific
a solid, density is constant at every point of the mate- heat at a constant pressure and Cv is the specific heat at
rial. However, for a fluid, the density will vary, depend- constant volume. The latter of the two applies only for
ing on selected control volume. If the control volume is gases. Given the two specific heats, we can find the ratio
infinitesimally small, the density of the given fluid will of specific heat, defined as
fluctuate wildly as the number of molecules varies over
time. Likewise, a control volume which is too large will Cp
also affect the density due to macroscale variations. k= (8.2)
Cv
Therefore, density is more properly defined as
This ratio can be related to the overall compressibility
m
ρ = lim (8.1) of a gas as a result of applied pressure. Table 8.1 gives
v →0 v
the molecular weight, density, and ratio of specific
heat for given gases at standard temperature and
where2 pressure.
ρ is the density in kg/m3 or lbm/ft3
m is the mass of fluid in kg or lbm
v is the volume of fluid in m3 or ft3 8.2.3 Specific Weight
The specific weight, as implied by the name, specifies
Many properties of gases and liquids are dependent the weight per unit volume of a given substance. Specific
on the temperature of the fluid. At standard tempera- weight is related to density by the given relation:
ture and pressure (T = 59°F [15°C] and P = 14.7 psia
[101.35 kPa]) the density of air is approximately 0.075 g
lbm/ft3 (1.2 kg/m3). Table 8.1 lists the densities of com- γ=ρ (8.3)
gc
monly used fuels.
Fundamentals of Fluid Dynamics 155
where where
γ is the specific†weight†in†N/m 3 †or lbf /ft 3 τ is the shear†stress†in†N/m †or†lbf /ft
2 2
g is the acceleration†due†to†gravity†in†m/s 2 †or†ft/s 2 µ is the absolute (dynamic) viscosity†in†N ⋅ s/m 2 †or

gc is the proportionality constant in consistent units lbf ⋅ s/ft 2 †

dV/dy is the velocity†gradient
8.2.4 Viscosity
A more common unit used to express viscosity is the
In fluid dynamics, one of the most important prop- poise or centipoises (10−2 poise). A poise is equivalent to
erties is viscosity. Viscosity is a measure of a fluid’s
resistance to shear stress. A lower value for viscosity
Dyne ⋅ s g
will result in less resistance encountered by the fluid. 1 p = 1 = 1
cm 2
cm ⋅ s
A prime example of viscosity in action is water and
crude oil. An identical amount of each liquid poured
at similar angles shows that water will flow at a higher N⋅s kg
1 p = 0.1 Pa − s = 0.1 † 2 = 0.1 †
rate than oil. This is due to the high viscosity in oil; m m⋅s
there is greater resistance to motion, thus, it will take
longer to pour. lb ⋅ s
1 p = 0.0020885 † f 2 †
ft
8.2.4.1 Absolute Viscosity
Under careful observation of fluid flow in contact with 8.2.4.2 Kinematic Viscosity
a solid surface, a moving fluid will see no net motion
More often than not, viscosity is expressed as a ratio
at the surface. This phenomenon is known as the “no-
of absolute viscosity to density of the fluid, in the
slip condition.” As seen in Figure 8.1, the velocity profile
form of
shows that the flow velocity is dependent on the dis-
tance from the surface. The closer the fluid is to the solid
surface, the lower the velocity will be and vice versa. µ
ν= (8.5)
With this understanding, we can relate the shear stress ρ
encountered by the fluid to the viscosity and volumetric
flow rate in the form of
where ν is the kinematic viscosity in m2/s or ft2/s.
The reason that kinematic viscosity is tabulated is
dV due to the frequency in which the term appears in
τ=µ (8.4) fluid calculations. Kinematic viscosity is most useful
dy
when analyzing systems involving liquids in which
the density is unlikely to change significantly with
ambient conditions and temperature. In the case of
gases, compression, atmospheric pressure change due
to elevation, and fluid temperature changes will affect
the density, and thus the kinematic viscosity. For these
V
reasons, absolute viscosity should be utilized for any
gas analysis.
More commonly used units for kinematic viscosity,
depending on application, are the stoke, centistokes,
dy and Saybolt universal seconds (SSU). A centistoke is
y equivalent to 10−2 stokes or
dV
m2
1 St = 10 −4
s
m2
1 cSt = 10 −6
s
ft 2
FIGURE 8.1 1 cSt = 10.76 †
Velocity profile of fluid along a solid surface. s
TABLE 8.2 Example 8.1

Specific Weight and Viscosity of Various Fluids How high should a water tank be located above
at Atmospheric Pressure grade to produce 100 psig of static pressure?
Specific Weight Kinematic Viscosity Assume the tank is exposed to standard tempera-
Fluid (lbf/ft3) (ft2/s) ture and atmospheric pressure (0 psig).
Since the tank will be stationary, one can apply
Air 0.075 1.58 × 10−4
hydrostatics to solve for the height. Isolating for
Mercury 847 1.3 × 10−6
the height difference will yield:
Glycerin 78.6 6.48 × 10−4
Water 62.4 1.13 × 10−6
P
Ethyl alcohol 49.3 1.6 × 10−5 ∆z =
γ
Gasoline 42.5 8.8 × 10−7
At standard pressure, the specific weight of

lb  N
On the other hand, SSUs are a measure of time, in sec- water is 62.4† 3f †9790†
 .
ft  m3 
onds, required for 60 mL of crude oil to flow through
a calibrated tube under a carefully controlled tempera-
ture. The conversion factor of kinematic viscosity from lbf in.2
100 × 144 2
SSU to centistokes is provided herewith: ∆z = in.2
ft = 231 ft (70.3 m )
lbf
62.4 3
 195  ft
1 cSt = ( 0.266 × SSU ) −   for 32 ≤ SSU ≤ 100
 SSU 
 135 
1 cSt = (0.220 × SSU ) −   for SSU > 100 8.3.1 Pressure
 SSU 
Pressure in a fluid is measured by collisions due to mole-
While there are numerous conventional units of cules acting on a surface and expressed in force per unit
kinematic viscosity, such as Redwood second, Degrees area. Depending on the application and convention, the
Engler, and Saybolt seconds Furol to name a few, the units of pressure may vary. In English units, the com-
most commonly used units in industrial practices are mon units of pressure are pounds per square inch (psi)
centistokes and SSU. Table 8.2 shows a handful of fluids and inches of water column (inches WC), while SI units
and their associated specific weight and kinematic vis- use kilopascals (kPa) and millimeters of water column
cosity values. (mm WC). Some useful units and conversions are in the
following.
N dynes
1 Pa = 1 = 10 = 9.869233 × 10 −6 atm
m 2
cm 2
8.3 Hydrostatics lbf
= 2.089 × 10 −2
Within the study of fluid dynamics lies a subset topic ft 2
known as hydrostatics. Hydrostatics focuses on fluids
that are at rest and in hydrostatic equilibrium. The con- lbf N kg
1 = 6895 2 = 5.171 cmHg = 0.0703
dition for equilibrium is met when the net force due to in.2
m cm 2
the weight of the fluid is balanced by an equal pressure.
= 27.68 in.WC
Using hydrostatic equations, one can solve for changes
in pressure as elevation changes. The differential equa-
tion for computing this is N kg lb
1 atm = 101, 3258 2 = 1.03323 = 14.696 f2
m cm 2 in.
dP 1 atm = 1.01325 †kPa = 76 cm Hg = 406.8 †in.WC

= −γ (8.6)
dz
lb
where2 1 bar = 0.9869 atm = 100 kPa = 14.5038 † f2
in.
dP is the differential pressure
dz is the elevation difference = 29.53 in. Hg
8.3.2 Absolute and Gage Pressure Restriction
Oftentimes, pressure is measured relative to a zero refer- Air flow

ence. The two notable scales of measuring pressure are A B
absolute and gage pressure. Absolute pressure measures
pressure relative to zero molecular interaction. Similar to
U-tube manometer
the Kelvin temperature scale, absolute pressure of zero
would mean no molecular motion. Therefore, absolute Air
pressure is always measured to be positive. The abso-
lute pressure scale is prominent in calculations involv- h3 Water
ing outer space, due to absence of gases in space. The
units for pressure in the absolute scale is denoted by a
trailing “a,” such as psia or psfa. h2
More commonly, pressure is measured relative to the
local atmosphere, or gage pressure. A net positive gage h1
pressure represents the pressure above that of the local
atmospheric pressure. For standard temperature and a
pressure, the gage pressure is measured at 1 atm, equiv- FIGURE 8.2
alent to 0 psig or 14.7 psia. The relationship between U-tube manometer. (Adapted from Baukal, C.E., Jr., The John Zink
absolute and gage pressure is expressed next as Combustion Handbook, CRC Press, Boca Raton, FL, 2000, Fig. 4.7.)
Pabs = Patm + Pgage (8.7) of manometers. For a U-tube manometer, the hydrostatic
equation is
Unlike absolute pressure, gage pressure can be n
egative.
A negative gage pressure indicates a pressure that is PA − PB = γ H2 O (h 2 − h1 ) (8.8)
lower that that exerted by the atmosphere. A negative
gage pressure is said to be in vacuum.
where
PA − PB †is†the pressure †drop †across†the†restriction††in
8.3.3 Pressure Measurement N/m 2 or lbf /ft 2
Many instruments exist that are used to measure γ x is the specific†weight†of†fluid†in†N/m 3 †or†lbf /ft 3
pressure, barometers, piezometers, manometers, and h 2 − h1 is the liquid†column†height†difference in†m†or†ft †
Bourdon-tube gages. Two of the more commonly used
pressure measurement instruments are manometers Note that the height of the gas is rarely included in the
and bourdon-tube gages. hydrostatic equation due to gas’s low specific weight.
However, if the fluid flowing from point A to point B
has a substantial specific weight, the height difference
8.3.3.1 Manometers of the fluid must be included.
Similarly, an inclined manometer will have a similar
A manometer is simple and one of the most useful hydrostatic equation, in the form of
devices for measuring pressure for gases. This is done
by filling a column with a liquid at a given height. The
PA − PB = (L sin φ)γ H2 O (8.9)
manometer is then connected at two given points and
the pressure drop is determined by the displacement of
the column of liquid in the manometer. Alternatively, where
one point can be open to atmospheric air and the pres- L is the displacement distance of fluid in m or ft
sure is measured with respect to atmosphere or gauge ϕ is the angle of declination in degrees or radians
pressure. Manometers are usually used to measure low
pressures and are commonly used in the combustion The main advantage of using an inclined manometer
industry. is increased accuracy, especially with small pressure
Manometers can take any given shape and can be differences. As ϕ approaches 0, the displacement of the
oriented at any given angle; however, the standard fluid increases, therein making subtle pressure drops
manometers are the U-tube and Inclined manometers more noticeable. This is especially useful when measur-
with the acting fluid as water, mercury, or oil. Figures ing minute pressure differences such as draft pressure,
8.2 and 8.3 show typical configurations for these types which will be covered later in the chapter.
Restriction
Air flow
A B
Inclined manometer
0
Water or oil
L
FIGURE 8.3
Inclined manometer. (Adapted from Baukal, C.E., Jr., The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2000, Fig. 4.8.)
Tube movement
Bourdon tube
Gear and
Spiral spring linkage
system
Indicating (movement)
needle
Socket
Fluid in
FIGURE 8.4
Bourdon-tube gage. (Adapted from Hydraulic Energy Transmission and Control, Hill Learning Systems, 2001–2007, http://www.hetacfluidpower.
com/screen_shot_gallery.htm)
Example 8.2 8.3.3.2 Bourdon-Tube Gage

A U-tube manometer filled with mercury is con- The second type of common pressure measurement
nected to a pressure vessel with one end open to device is a Bourdon-tube gage. A Bourdon-tube gage mea-
atmosphere. The difference in height of mercury sures pressure using a coiled tube, which is fixed at one
reads 60 in. (1.52 m). What is the pressure in the
end. At standard pressure, the tube remains unperturbed.
vessel?
The specific weight of mercury can be Once pressure is applied to the metal tube, the force will
found in Table 8.2 and is determined to be cause the tube to straighten out and move a pointer needle,
( )
847 lbf /ft 3 1.33 × 10 5 N/m 3 . Using Equation 8.8, which represents the pressure difference. Bourdon-tube
gages are sized for ranges 1 psig all the way to 10,000 psig
one can substitute for the variables:
(69 Mpag), depending on application. Figure 8.44 illustrates
the major components of a Bourdon-tube gage.
lb  1 ft 
∆P = 847 3f ×  60 in. ×
ft  12 in. 

8.4 Fluid Flow
For centuries, fluid flow has fascinated scientists.
lb 1 ft 2 Visual observation of fluid flow has shown that there
∆P = 4235 2f × = 29.4 psi†( 203 kPa )
ft 144 in.2 are two distinct broad classifications of flows. At low
velocities, flow undergoes a smooth and streamline like turbulent flow. In addition to determining the type
motion with little variation over time. At high veloci- of flow, the Reynolds number also shows the intensity
ties, the flow is associated with chaotic fluctuations. of fluid mixing and vortex generation. Figure 8.5 shows
These fluctuations result in eddies caused by the the experiment used to study the effects of laminar and
wake, which increase the rate of mixing. These two turbulent flow using colored dye.
velocity dependent flows are classified as laminar
and turbulent flow respectively. While this naturally
occurring phenomenon was observed for centuries, Example 8.3
it was not until 1883, when Osborn Reynolds was Determine the velocity of water at 70°F (21°C)
able to quantify the parameters, which dictated the through a 2 in. (5 cm) pipe to achieve turbulent
behavior of fluid flow.3 This parameter is written in flow. Repeat calculations for air at 70°F (21°C).
the form of The fluid properties of water at 70°F are
ρVL c lb  kg 
Re = (8.10) ρwater = 62.4 † 3f † 998 † 3 
µ ft  m 

where lb ⋅ s  N ⋅s
µ water = 2.05 × 10 −5 † f 2 † 1.00 × 10 −3 2 
Re is the dimensionless Reynolds number ft  m 
Lc is the characteristic length relative to the flow
The density and dynamic viscosity will yield
The Reynolds number can also be represented as a func- the kinematic viscosity.
tion of kinematic viscosity, in the form of
µ 2.05 × 10 −5 × 32.2 ft 2
VL c νwater = = = 1.06 × 10 −5 †
Re = (8.11) ρ 62.4 s
ν
In pipes, the characteristic length is equal to the diam- The correlation of viscosity to velocity in the
eter of the pipe. For flow in pipes, a Reynolds number form of Reynolds number is
less than 2300 is considered to be laminar and Reynolds
numbers greater than 4000 to be turbulent. Any number VD Re × ν
in between and the flow is said to be in transition flow, Re = ⇒V=
ν D
where the flow is no longer laminar, but does not behave
Dye
Dye in laminar flow
Low velocity water
Dye in laminar flow
High velocity water
FIGURE 8.5
Osborn Reynold’s experimental apparatus used to study the transition from laminar to turbulent flow. (Adapted from Baukal, C.E., The John
Zink Hamworthy Combustion Handbook, Vol. 1, CRC Press, Boca Raton, FL, 2012, Fig. 9.6.)
Under the assumption of turbulent flow, the ∂P ds dA

P + ——
minimum Reynolds number will be 4000. Solving ∂s
for V:
ST
ft 2
4000 × 1.06 × 10 −5 †
Vwater = s = 0.25 †ft ds
(2 in. × (1 ft / 12 in.)) s

Now repeating the problem for air, the fluid dz

properties are θ
lb  kg 
ρair = 0.075† 3f † 1.20 † 3  dAds
ft  m  ρg ———
gc
PdA
lb ⋅ s  N ⋅s
µ air = 3.79 × 10 −7 † f 2 † 1.81 × 10 −5 
ft  m2  FIGURE 8.6
Force components on fluid particle in direction of streamline.
ft 2  m2 
νair = 1.63 × 10 −4 † 1.51 × 10 −5 † 
s  s 
Dividing through by ρdAds and substitution of
cos θ = dZ/ds will yield:
Following the Reynolds number relationship,
the velocity of air in the pipe will be

1 ∂P g ∂Z as
4000 × 1.63 × 10 −4
ft  m + = (8.13)
Vair = = 3.91 † 1.19  ρ ∂s g c ∂s g c
(2 in. × (1 ft/12 in.)) s  s

Note that due to the density and viscosity The acceleration as has a tangential and normal term
of the air, the velocity must be almost 10 times in the form of
larger than the velocity of water to achieve tur-
bulent flow. dV ∂V ds ∂V

as = = + (8.14)
dt ∂s dt ∂t
where (∂V/∂t ) = 0 at steady state. Substitution of the

8.5 Euler’s Equation acceleration term and moving them to one side will
result with
When evaluating for pressure in a fluid, the hydrostatic
equation discussed in Section 8.3 will only apply for
fluids in equilibrium. Fluids undergoing motion are not 1 ∂P g ∂Z V∂V
in equilibrium; and therefore, they cannot be evaluated + + = 0 (8.15)
ρ ∂s g c ∂s g c∂s
using hydrostatic equations. Using Newton’s second
law, we consider a one-dimensional steady state flow
along a streamline, as seen in Figure 8.6: Since all the terms are a function of s only, the partial
derivatives can be replaced with total derivatives.
∂P ρg
∑ F = ma ⇒ PdA −  P + ∂s ds dA − g c
dAds cos θ

1 g V
dP + dz + dV = 0 (8.16)
ρ gc gc
ρdAds
= as (8.12)
gc
This equation is known as Euler’s equation and rep-
resents the pressure necessary to accelerate a liquid.
where Euler’s equation can be used if the flow follows the
as is the acceleration of fluid in direction of streamline assumptions of steady state, no viscous effects, incom-
gc is the gravitational constant pressible, and along a streamline.
2
8.6 Bernoulli Equation
P2 = ?
If one integrates Newton’s second law, they can deter-
mine the work done on a particle. Similarly, if one inte-
grates Euler’s equation along a path “s,” one will get a
similar relation. Starting with Euler’s equation
1 g V
dP + dz + dV = 0 (8.17)
ρ gc gc
100 ft
Integration will yield
P g V2
+ z+ = C (8.18)
ρ gc 2g c
where C is the constant. 1

γg c
Substitution of ρ with yields
g P1 = 100 psig
γV 2 FIGURE 8.7
P + γz + =C Water flowing in pipe from point 1 to 2.
2g
(8.19)
γV 2 γV 2
P1 + γz1 + 1 = P2 + γz 2 + 2 Assuming standard temperature and pressure,
2g 2g the specific weight of water is
lbf  N
This equation is known as the Bernoulli equation and is γ = 62.4 † 9790 † 3 
ft 3  m 
valid for all flow that meets the following criteria:
According to the Bernoulli equation, the pressure
• Steady flow and velocity head of a fluid will remain constant
• Incompressible throughout a conduit. If there are no losses due
to friction, then the Bernoulli equation applied at
• Along streamline points 1 and 2 must also be constant and equal to
• No viscous effects one another. Doing this will net
P1 V 2 P V2
Another form of the Bernoulli equation involves + + gz1 = 2 + + gz 2
γ 2g γ 2g
dividing the entire equation by the specific weight to get
Since the flows from both points are internal and
P V2 constant cross section, the velocities at points
+ +z = C (8.20)
γ 2g 1 and 2 are equal.
P1 P
+ gz1 = 2 + gz 2
Given enough parameters, the Bernoulli equation can γ γ

solve for problems to find pressure, velocity, or height at
any location as long as the flow meets the criteria listed Isolating for P2 yields
earlier.
P2 = P1 − γ(Z 2 − Z1 )
Example 8.4
Water flows in a 3 in. (8 cm) pipe up a hill to 100 ft lbf lb   12†in.   1†ft 3  
(30 m) above the grade (see Figure 8.7). If the supply P2 = 100 − 62.4 3f (100 ft − 0 ) ×  
in.2
ft  1 ft   123 in.3  
 ††
pressure is 100 psig (700 kPa), what is the pressure
at the top of the hill? Neglect any friction between
= 56.6 psig ( 390 kPa )
fluid and pipe wall.
Example 8.5
Air at standard temperature and pressure flows
through a pipe at 20 ft/s (6 m/s). The pressure in Flow
the pipe is measured to be 0.5 psig (3.4 kPa) with
the exit nozzle at a reduced area exposed to open
air. Assume a straight pipe with negligible friction.
What is the maximum velocity at the nozzle exit?
Since the flow is along the streamline, incom-
pressible, and steady with no viscous effects, the
Bernoulli equation can be used. The flow in and
at the exit of the nozzle can be determined from
P1 V12 P V2
+ + gz1 = 2 + 2 + gz 2 Static pressure + Velocity pressure = Total pressure
γ 2g γ 2g

FIGURE 8.8
The specific weight of air at STP is 0.075 lbf /in.2 Static, velocity, and total pressure measurements inside a pipe. (From
Because the nozzle exit is exposed to open air, the Baukal, C.E., Jr., The John Zink Combustion Handbook, CRC Press,
gage pressure at the exit will be zero. Likewise, a Boca Raton, FL, 2000.)
straight pipe would assume no height difference;
therefore, the height terms can be neglected and
The two remaining terms P/γ and V 2 /2g are known
is the usual case for gases. The simplified equa-
tion will then be as the static and velocity pressures, respectively. The
static pressure is defined as the actual pressure of the
P1 V12 V22
fluid and is dependent on the state, but not the flow
+ = associated with the fluid. Static pressure is easily mea-
γ 2g 2g
sureable using any pressure measurement device. The
velocity pressure is a pressure drop associated with the
The exit velocity can be isolated to get velocity in a moving fluid and is inversely proportional
to the static pressure. That is, a drop in velocity pres-
sure will result in an increase in velocity pressure and
lb ft in.2
2P1g 2 × 0.5† f2 × 32.2† 2 × 144 † 2 2 vice-versa. Therefore, the total pressure is the sum of
V2 = + V12 → in. s ft +  20 ft 
lb   the static and velocity pressure (shown in Figure 8.8)
γ 0.075† 3f
 s
ft in the form:
ft  m PT = PS + PV (8.22)
= 249  76 
s s
where
When calculating for the exit velocity, note that PS is the static pressure
the velocity in the pipe does not contribute greatly PV is the velocity pressure
to exit velocity. In many applications, the up-
stream flow velocity can be neglected in n ozzle Example 8.6
calculations.
Air passes through a conduit and the static pres-
sure is measured at 2 psig. If the total pressure
8.6.1 Static, Velocity, and Total Pressure across the Pitot tube is 60 in. WC, what is the
velocity of the air?
From the previous section, Euler’s equation was inte-
The relationship between the total pressures
grated to derive the Bernoulli equation. Due to integra- in a Pitot tube is expressed in Equation 8.22.
tion, the Bernoulli equation contains a constant, which Therefore, solve for the velocity pressure in the
is commonly referred to as the total pressure when ele- tube. Since the total pressure is in inches water
vation can be ignored as in most gas flows. This would column, one must convert from in. WC to psf.
result in a simplified Bernoulli equation in the form
 1 ft lb 
PV = PT − PS =  60 in.WC × × 62.4 3f 
P V2 γ †V 2  12 in. ft 
+ = PT or P + (8.21)
γ 2g 2g
 lbf in.2  lb
− 2 × 144 2 
= 24 2f
 in.2
ft  ft
where PT is the total pressure.
The velocity pressure is then calculated to be where

ρi is the density of fluid at i
lbf  ft  Ai is the cross-sectional area of flow region at i
24 × 2  32.2† 2 
lb γV 2 ft 2  s  Vi is the fluid velocity at i
24 2f = →V=
ft 2g lb
0.075 3f
ft For steady, incompressible flow, the continuity equation
simplifies to
ft  m
= 143  43.7 
s s

Q = A1 V1 = A 2 V2 (8.28)
This example shows the relationship between
static and total pressure. In practice, a Pitot tube
usually measures velocity pressure directly using where Q is the flow rate in m3/s or ft3/s
the difference between total and static pressure.
Example 8.7
Air flows in a 1 m2 (10.7 ft2) duct at 10 kg/s
(22 lbm /s) , find the flow rate and air velocity at the
8.7 Continuity duct neck with an area of 0.2 m2 (2 ft2).
Usually in duct flows, the density does not
When evaluating flow passing through two distinct change much so this can be assumed and the
points, the law of conservation of mass states that the simplified continuity equation can be used
time rate of mass coming into a system is equal to the
net rate of mass leaving the system. Using this law, a
Q = A1V1 = A 2 V2 †
closed system may be taken with a zero rate of change
in mass flow. The pipe has a differential area of dA with
length step increments of ds. Therefore the differential The rate of discharge, or volumetric flow rate, can
mass flow will be be found by taking the mass flow rate of the fluid
divided by the density. For air, the density of air
dm = ρdsdA (8.23) ( )
is 1.2 kg/m 3 0.075 lbf /ft 3 .
where kg
dm is the differential mass flow rate . 10 †
m s = 8.33 †m
3
Q= =
dA is the differential area ρ 1.2†kg s
m3
Taking the derivative of dm will yield the time rate of
change in mass flow.
Now with the volume flow rate, the air velocity at
ds the neck can be determined.
dm
 = ρ dA (8.24)
dt
The change of length with respect to time can be sim- m3
8.33 †
plified as the velocity of the fluid. Since the system is a Q1 s = 41.7 m
Q1 =A 2 V2 → V2 = =
closed system, the rate of change for mass flow is zero. A1 0.2†m 2 s

ds
ρ = ρVda = 0 (8.25)
dt
Example 8.8
Integration results in
Water flows in a 2 in. (5 cm) diameter pipe at
2 10 ft/s (3 ft/s) before necking to 1 in. (2.5 cm)
∫ρVdA = C
1
(8.26) diameter. Find the velocity at the neck in ft/s
(m/s) and flow rate in GPM (m3/s).
Once again, using the continuity equation,
Evaluation for the following integral gives the equation neglecting changes in density for water since it is
for continuity. not very compressible yields:
ρ1A1 V1 = ρ2 A 2 V2 (8.27) Q = A1V1 = A 2 V2 †

The volumetric flow rate, Q, will be Example 8.9
π  ft 1 ft 2 An 80,000 lb/h boiler requires a fan with a 75%

Q = A1V1 =  22 in.2  × 10 × mechanical efficiency to provide air at a rate of
4  s 144 in.2
20,000 CFM. If the pressure drop is designed to
gal s be 22 in. of water column, what is the power of
× 7.51† 3 × 60 † = 98.3 GPM
ft min the fan in hp?
To determine the fan power, take the measured
Next, isolating for V2 will provide pressured drop and convert to lbf/ft2. Since the
pressure is measured in inches of water, take the
specific weight of water to be 62.4 †lbf /ft 3 .
 π 2 2 ft
 4 × 12  × 10 s
AV   ft lbf 1 ft
V2 = 1 1 = = 40 ∆P = γ H2O × in. WC = 62.4 × 22 in. × †
A2  π 1 2 s ft 3 12 in.
 4 × 12 
  lbf
= 114
ft 2
Therefore,
8.8 Fan and Pump Sizing lbf ft 3 1 min

∆PQ = 114 × 20, 000 ×
In industrial and especially boiler applications, a deliv- ft 2
min 60 s
ery system for air or fuel usually includes pumps and 1 s ⋅ hp
× = 69.3 hp
fans to increase liquid and gas pressures in order to 550 ft ⋅ lbf
operate properly. The power of a fan or pump is the
force applied over a distance per unit time and can be
Taking the mechanical efficiency, the actual
derived by differentiating Newton’s second law. power of the fan will be
F ⋅ ds Work
Power = ∑ ∆T
=
Unit†time
(8.29)

hpActual =
hpTheoretical 69.3
η
=
0.75
= 92.4 hp
The force exerted on a fluid can be measured by eval-

uating the pressure drop multiplied by the pipe/duct Converting to SI units, the answer will be
area. Likewise, the change in distance with respect to
time can be expressed as the flow velocity. Therefore, 746 †W
Power = 92.4 hp × = 69.0 kW
the aforementioned equation can be simplified to 1 hp

( P2 − P1 ) A ⋅ V (8.30) Another useful tool for sizing fans and pumps is the
similarity law, which relates the speed with the flow
or rate, pressure, and power of fans or pumps operating at
different speeds. The similarity laws are as follows:
Power = ∆PQ (8.31)
N2 Q2
where ΔP is the pressure drop across the fan or pump in = (8.32)
N1 Q1
N/m2 or lbf/ft2.
The power of a fan/pump is measured in units of 2
 N2  P2
N ⋅ m/s or†ft ⋅ lbf /s, or more commonly in watts or  N  = P (8.33)
horsepower, where 1 1
N⋅m J  N2 
3
HP2
1 = 1 = 1 W
s s  N  = HP (8.34)
1 1
ft ⋅ lbf where
550 = 1hp
s Ni is the fan or pump speed in RPM
Qi is the volume flow rate in m3/s or ft3/s
1 hp = 746 W Pi is the pressure head in cm WC or in. WC6

While the equations are stated as the pump similarity γV12 γV 2

laws, the aforementioned expressions are still viable for P1 + + γ z1 = P2 + 2 + γ z 2 + ∆PLosses (8.35)
2 2
fans and blowers.
where ΔPLosses is the total pressure losses in equivalent
Example 8.10 units.
The total pressure losses through a pipe must account
The same boiler in Example 8.9 needs a feed water for both frictional losses and fitting losses. Pressure
pump to supply water at a flow rate of 160 GPM
drop due to frictional factors are often known as pipe
(0.27 m3/h) with an estimated pressure rise of
180 psig (1.24 MPa). Calculate the power of the losses or major losses, while pressure drops through
motor, in hp (kW), for a 70% efficient pump. bends and fittings are known as component losses or
minor losses. Although these are named by conven-
A pump behaves similar to a fan; therefore, the power tion, many times the “minor losses” are greater than
of a pump is determined using the same equation. the “major losses.” The total pressure loss is equal to
the sum of these two major and minor losses.
lb gal 1 min
Power = ∆PQ = 180 † f2 × 160 † × †
in. min 60 s 8.9.1 Major Losses
ft3
144 in 2
lb ⋅ ft One of the most useful equations for solving problems
× 0.133 † × 2 = 9192 f
gal 1 ft s relating to fluid mechanics is known as the Darcy–
Weisbach equation. It is used to determine major losses
Converting to hp will yield a fluid encounters when travelling down a pipe or tube
for fully developed steady flow.
lbf ⋅ ft When a fluid first enters at an inlet, the fluid profile
9192
s = 16.7 hp varies in shape but, slowly changes into a parabolic pro-
hp ⋅ s file. Once the flow profile has assumed a fully parabolic
550
lbf ⋅ ft shape, the flow is said to be fully developed. Any fluid
with the exception of high viscous fluids, which has
Now, for a 70% efficient pump, the actual hp will be traveled a sufficiently long distance can be said to have
fully developed flow.
hpTheoretical 16.7 To determine the associated head loss for fully devel-
ηpump = = oped flow, the derived Darcy–Weisbach equation is
hpActual HPActual
used, as presented herewith
hpActual = 23.8 †hp†(17.7 †kW )
fL γV 2
∆PMajor = (8.36)
D 2g
Note that for combustion systems, generally the pump
power required for liquid flow is much lower than the where
power necessary in a fan. This is mainly due to the incom- γ is specific weight
pressibility of liquids. f is the Darcy friction factor
L is the length of pipe in m or ft
D is the diameter of pipe (circular) in cm or in.
V is the average fluid velocity in m/s or ft/s
g
is the acceleration†due†to†gravity†in†m/s or ft/s
2 2
8.9 Pressure Losses through Pipes

The Darcy friction factor “f” is a dimensionless number,
Whenever a fluid flows through a straight section of which represents the ratio of the shear stress to kinetic
pipe with constant area, there is a drop in the down- pressure that a fluid experiences.
stream pressure due to frictional losses from the pipe For turbulent flow, the Darcy friction factor is depen-
wall. There can also be pressure losses when a fluid dent on the associated Reynolds number and the “rel-
flows through fittings, bends, pipe size changes, and ative roughness.” The relative roughness term first
inlets. The derived Bernoulli equation assumed an ideal appeared when J. Nikuadse conducted experiments with
case without friction, but in real world scenarios, even sands of known size were fitted onto the inner walls of
a smooth pipe will exert friction on a fluid. Hence, a a pipe to determine the effects of roughness in a pipe
revised Bernoulli equation that takes losses into account wall. The relative roughness is a ratio of the height of the
from point 1 to point 2 can be written as: grains of sand to the diameter of the pipe, measured in
TABLE 8.3 that for smooth pipes (ɛ ≈ 0), the friction factor does not
Equivalent Roughness for New Pipes equal zero. This implies that no matter how smooth a
pipe may appear, there is always pressure loss due to
Relative Roughness, ɛ
roughness on a macroscopic level. Another method of
Pipe (ft) (mm) finding the friction factor for turbulent flow is the equa-
Riveted steel 0.003–0.03 0.9–9.0 tion developed by Swamee and Jain.6 The equation is
Concrete 0.001–0.01 0.3–3.0
0.25
Wood stave 0.0006–0.003 0.18–0.9 f= 2
(8.37)
  ε 5.74 
log  3.7 D + Re0.9
Cast iron 0.00085 0.26

Galvanized iron 0.0005 0.15   
Commercial steel 0.00015 0.045
or wrought iron Nikuadse’s experiments have also shown that for lami-
Drawn tubing 0.000005 0.0015 nar flow, the Darcy friction factor is independent of the
Plastic, glass 0.0 (smooth) 0.0 (smooth) wall roughness and only depends on the Reynolds num-
ber as seen in Figure 8.9. The relationship between friction
Source: Roberson, J.A. and Crowe, C.T., Engineering Fluid
Mechanics, Houghton Mifflin, Boston, MA, 1980. factor and Reynolds number for laminar flow is as shown
terms of length.3 Table 8.3 shows the equivalent rough- 64

f= (8.38)
ness for the inner walls of typically used materials. Re
With a given Reynolds number and relative rough-
ness value, the friction factor can be found using what Many of the aforementioned equations assume fluid
is called the Moody diagram, as seen in Figure 8.9. Note flow in a circular pipe. In order to apply these equations
0.1
0.09 f = 64/Re
0.08 z/D
0.07 0.05
0.04
0.06
0.03
0.05
0.02
0.015
0.04
0.01
Friction factor f
0.006
0.03
0.004
0.002
0.02 0.001
0.0006
0.0004
0.0002
0.0001
0 (smooth) 0.00005
0.01
0.009
0.000005
0.00001
2 3 4 6 8 2 3 4 6 8 2 3 4 6 8 2 3 4 6 8 2 3 4 6 8
1000 104 105 106 107 108
Reynolds number
FIGURE 8.9
Moody diagram. (Adapted from Roberson, J.A. and Crowe, C.T., Engineering Fluid Mechanics, Houghton Mifflin, Boston, MA, 1980.)
for a noncircular cross section, such as a rectangular duct, With a Reynolds number in the 105 range, the
the “D” term must be replaced with a hydraulic diameter. flow is considered turbulent, so the Moody dia-
The equation for hydraulic diameter is seen next as gram or Equation 8.37 should be used. Since the
flow is turbulent, the pipe roughness will be
4 × Ac needed to be taken into account. Based on Table
Dh = (8.39) 8.3, commercial steel has a relative roughness of
Pw 0.00015 ft (.045 mm), thus the value ɛ/D will be
where 0.0009. On the Moody diagram, the Reynolds
Ac is the cross-sectional area number and relative roughness correspond to a
friction factor of approximately 0.021.
Pw is the wetted perimeter8
If Equation 8.37 were used, the friction factor
will be
The wetted perimeter term refers to the perimeter
of the conduit that is in physical contact with the fluid. 0.25
Using the hydraulic diameter will give a friction fac- f= 2
  ε 5.74  
tor that is within a reasonable range for turbulent flow. log  3.7 D + Re0.9  
However, using the hydraulic diameter for laminar flow  
will result in significant error differences. 0.25
= 2 = 0.021
  
Example 8.11 log  0.0009 + 5.74 

( )
0.9 
 3.7 1.57 × 10 −5 
   
Find the pressure drop in a 100 ft (30 m) straight
( )
pipe flowing at 100 GPM 22 m 3 /h . Water flows
in the 2 in. (5 cm) circular at 70°F (21°C). Assume The two friction factors are identical, proving the
commercial steel pipe. accuracy of Equation 8.37. Evaluating for the pres-
To evaluate for the pressure drop through the sure drop will result in
pipe, consider the Darcy–Weisbach equation for
pressure loss from Equation 8.36. lbf 2 ft
2
0.021(100 ft) 62 .4 † × 10 .2 †
∆P = × ft 3 s2
fL γV 2
∆P = ( 2 in. × (1 ft/12 in.)) 2 × 32.2
2
ft
D 2g s

lbf 1 ft 2
The friction factor is correlated to the flow type, = 1245
×
= 8.65
psi
(60
kPa )
which is a dependent on the Reynolds number. ft 2
144 in.2
Therefore, find the velocity for a flow rate of
100 GPM. In industrial application, height can normally be
ignored for gases, but not so for liquids.
Q
Q = AV or V =
A
8.9.2 Minor Losses
gal ft 3 1 min
100 × 0.133 ×
min gal 60 s ft While the previous section involved solving for major
V= = 10.2
π 1 ft 2
 s losses in a straight pipe of constant cross section, an
 4 × 2 in. × 144 in.2 
2 2
 
actual boiler comprises several components such as
bends, elbow, tees, and other fittings. A fluid passing
For water at 70°F, the properties of water are through these fitting will experience pressure drops
via additional turbulence and flow separation through
lbf  N sudden shifts in momentum. Pressure drops associated
γ = 62.4 † 9790 3 
ft 3  m  with fittings are known as minor losses, which contrib-
ute to the overall pressure loss.
ft 2  m2 
ν = 1.059 × 10 −5 † 1.00 × 10 −3 †  Unlike major losses, equations for minor losses are not
s  s  as convenient. Therefore, a commonly adapted method

for determining minor losses is from a loss coefficient.
Then, the Reynolds number will be The loss coefficient is defined as
ft 1†ft
10 × 2†in. × dP
VD s 12 in. KL = (8.40)
Re = = = 157 , 000 = 1.57 × 10 5 1 γV 2
−5 ft
2
ν
1.059 × 10
s 2 g
Notice that the loss coefficient KL is a dimensionless Because the velocity of CH4 is not given, one
number which is a ratio of the pressure difference to must find the fluid properties of CH4 and mass flow
velocity pressure. Solving Equation 8.40 for pressure rate needed to provide 250 × 10 6 Btu/h (73 MW)
will result in to the system. The density and specific weight of
CH4 will be
γV 2
dp = K L (8.41) 30 + 14.7 lb  kg 
2g ρ = 0.075 (0.55) = 0.125† m3 † 2.00 † 3 
14.7 ft  m 
The loss coefficient KL is strongly dependent on g 32.2 lb  N

γ =ρ = 0.125 × = 0.125 † 3f † 19.6 † 3 
the geo metry of the fitting and Reynolds number. gc 32.2 ft  m 
Fitting losses for a few cases are presented herewith in
Figure 8.10, which shows the variation in KL due to sev-
The LHV value of CH4 is 21,495 Btu/lbm, which
eral factors. More detailed information on the topic is
will yield a mass flow rate of
discussed further by Idelchik7 and Crane.8
Btu 1 lbm 1 h lb
250 × 10 6 × × † = 3.23 † m
Example 8.12 h 21, 495 Btu 3, 600 s s
Methane flows through a 250 ft (76 m) straight
supply line to a boiler with supply pressure at Many texts will employ an iterative approach that
30 psig (210 kPa). The fluid also passes by two requires using a solver program to iterate for a
valves (KL = 3.0). What diameter should the pipe be diameter. However, this method requires an edu-
such that the heat supplied will be 250 × 106 Btu/h cated initial guess based on normal velocities and
(73 MW) and not exceed a 5 psig (34 kPa) pressure a scaling method to refine the result. Instead, for
drop? Assume fully developed steady flow with this example, start with a typical flow velocity of
CH4 on a LHV basis. 100 ft/s, which is the typical fluid velocity for
KL
D2/D1 θ = 60° θ = 180°
Contraction 0.0 0.08 0.50
ρV2 0.20 0.08 0.49
dP = KL 2 0.40 0.07 0.42
D1 θ D2 V
0.60 0.06 0.32
0.80 0.05 0.18
0.90 0.04 0.10
KL
Expansion D1/D2 θ = 10° θ = 180°
ρV2 0.0 — 1.00

dP = KL 2 D1 0.20 0.13 0.92
V θ D2
0.40 0.11 0.72
0.60 0.06 0.42
0.80 0.03 0.16
Pipe entrance
ρV2
dP = KL R /D KL
2
D V 0.0 0.50
0.10 0.12
>0.20 0.03
R
FIGURE 8.10
Coefficient loss for various fittings. (Adapted from Roberson, J.A. and Crowe, C.T., Engineering Fluid Mechanics, Houghton Mifflin, Boston, MA, 1980.)
gas flow applications. At 100 ft/s, the continuity Therefore, a 6.88 in. pipe meets both conditions;
equation gives however, since 6.88 in. is not a standard pipe
size, a smaller or larger pipe diameter will need
 = AV
Q to be selected. Instead of guessing, a more direct

method would be used to correlate the net pres-
The volume flow rate relates to the mass flow rate sure drop to the major and minor losses using the
by a factor of density of the fluid or equations provided herewith.
m
  =m

ρ=  →Q 1
Q ρ ∆P ∝ for†major†losses
D g5

m lb 1 ft 1 s 3
A= = 3.23 † m × × = 0.258 ft 2 1
ρV s 0.125 lbm 100 ft ∆P ∝ for†minor†losses
D g4

Therefore, the diameter of the pipe at 100ft/s
will be Since the major and minor losses each contribute
to half the total head loss, one can calculate ΔP as
π 2 144 in.2 4
A= D ⇒ D = 0.258 †ft 2 × × = 6.88 in.
4 1 ft 2 π ∆PD g4.5 = Constant

While this pipe diameter will be able to sup-
ply the 250 × 10 6 Btu/h heat needed, the pres- This will result in
sure drop still needs to be determined to ensure
the total losses are no more than 5 psig. Since
1.69 psig × (6.88 )
4.5
= 5 psig × ( D )
4.5
the supply line is a straight section with valves,

there are both major and minor losses. Summing
Equations 8.36 and 8.41 will give the total pres-
D = 5.4 in.
sure drop through the system:
γV 2  fL So a 6 in. pipe should be used. Double checking

∆PTotal =
2g  D
+ ∑K valves

 for the pressure drop will provide
lb
3.23 † m
For the major losses, evaluate for the Reynolds s ft
V= = 132†
number to solve for the friction factor. lb π 1 ft 
2
s
0.125† m3 ×  6 2 ×
ft 4 144 in 2 
 6.88 
100 × 
 12 
Re = = 3.33 × 10 5
1.72 × 10 −4
ft  6 
131.6 † ×  †ft
Using commercial steel, the relative roughness s  12 
Re = = 3.82 × 10 5 ⇒ f = 0.016
−4 ft
2
will be 0.00015 ft, which gives an ε/D of 0.0002. 1.72 × 10 †
According to the Moody diagram, with the s
given Reynolds number and relative roughness,
the friction factor will be approximately 0.016.
Likewise, Equation 8.37 can be used to solve for lb ft 2  
friction factor. Now the total pressure drop in the 0.125† 3f × 1312 † 2  0.016 ( 250 †ft ) 
∆P = ft s  + 6
system will be ft   6 †ft 
2 × 32.2† 2
s   12  
 
  lbf
lb ft   = 466.33 = 3.2psig
0.125† 3f × 100 2 † 0 .016 ( 250 ft ) ft 2
∆PTotal = ft s  + 2 ( 3 )
ft  1 ft  
2 × 32.2† 2   6.88 × 
s   12 in.   This method shows that by guessing a typical
fluid velocity of 100 ft/s, it will provide a rea-
lbf 1 ft 2
= 243.4 × = 1.69 psig sonably close answer to the actual pressure
ft 2
144 in.2 drop for gas flows.
Example 8.13 With this diameter, solve for the velocity of

water, which is found to be
Determine the diameter for a 100 ft (30 m) water
pipe with a supply flow rate of 1200 GPM (327
m 3 /h ). The water passes through two valves π 2 2 1 ft 2
A= 6 in. × = 0.196 ft 2
(KL = 1.5). 4 144 in.2
Similar to the previous example, start with a
typical fluid velocity for liquids of 15 ft/s (4.6 m/s). gal ft 3 1 min
Combined with the volume flow rate, solve for the 1200 × 0.133 † ×
Q min gal 60 s ft
pipe area. V= = = 13.5
A 0.196 †ft 2 s
gal ft 3 1 min The Reynolds number will then be

1200 × 0.133 † ×
Q min gal 60 s
Q = AV → A = =
V ft ft 6.0
15 13.5 × ft
s Re = s 12
2 = 5.15 × 10
6
−6 ft
= 0.177 ft 2 1.31 × 10
s
The specified pipe diameter for this flow will be The friction factor will remain at 0.015 and the
pressure drop will be
in.2  
4 × 0.177 ft 2 × 144
4A ft 2 = 5.70 in. 62.4 × 13.52  0.015 × 100 
D= = ∆P =  + 2 (1.5)
π π 2 × 32.2  6

 12 
Next solve for the pressure drop to ensure ΔP 1 ft 2
× = 7.35 psi
does not exceed 10 psig. For this, solve for the 144 in.2
Reynolds number.
The pressure drop does not exceed the required

ft 5.7 pressure, so a 6 in. (15 cm) diameter is acceptable.
15 × ft
Re = s 12
2 = 5.44 × 10
6 For many situations, it is completely acceptable
−6 ft to use typical fluid velocities when not explicitly
1.31 × 10
s given. Regardless of the case, a velocity of 100 ft/s
(30 m/s) for gases and 15 ft/s (4.5 m/s) for liquids
are appropriate first guesses and will be close to
Assuming a commercial steel pipe, the rough- what the actual flow velocities will be.
ness factor will be 0.00015 ft and ε/D of 0.0003.
Using the Moody diagram or Equation 8.37, the When evaluating minor losses for valves, a regularly
friction factor will equal 0.015. Evaluating for the
used term to rate a valve is the CV value. Although the
pressure drop will yield
term is similar looking to the specific heat under con-
stant volume conditions, CV is a measure of volume flow
lbf ft 2 rate of water, in GPM, at a pressure of 1 psig. CV can
62.4 × 152 2
γV 2  fL  ft 3
s be related to the minor losses by applying the Darcy–
∆P = + K Valves  =
2g  D   ft  Weisbach equation.
2  32.2 2 
 s 
γV 2
  ∆P = K
 .015 × 100 ft  1 ft 2 2g
× + 2 (1.5) × = 9.62 psi
5.7
 144 in.
2

 12 
The CV relates to the velocity, via the continuity equa-
tion, Q = AV, so
The pressure drop does not exceed 10 psi
(0.7 bar), so a 5.7 in. (14 cm) diameter will suffice. gal ft 3 1 min
CV × 0.13368 † ×
However, the standard pipe sizes are not sized for min gal 60 s C
5.7 in. (14 cm) pipes; therefore a 6 in. (15 cm) pipe V= = 0.4085 V2
π 2 2 1 ft 2
D
should be used. The pressure drop for a 6 in. pipe D in. ×
may be checked if so desired. 4 144 in.2
Substitution into the Darcy equation will provide K eq = 14.9

2
lbf  0.4085 C V 
62.4 ×  The equivalent loss coefficient will then be used
lb ft 3  D2 1 ft 2 in Equation 8.42 to find the necessary CV for the
1 f2 = K ×
in. ft 144 in.2 valve.
2 × 32.2 2
s
KC 2V K=
891D 4
⇒ CV =
(891) 3 4 = 69.6
⇒ 1 = 0.0011 C2V 14.9
D4
891D 4
K eq = (8.42)
C 2V 8.9.3 Series and Parallel Flow
where Keq is the equivalent loss coefficient. When evaluating flow passing through several compo-
The equivalent loss coefficient is useful because it nents with varying cross-sectional area, calculating the
relates the associated CV value with the loss coefficient pressure drop across multiple elements may prove to be
commonly used to determine minor losses. quite time-consuming. Taking this into mind, an equa-
tion can be derived, similar to a thermal circuit, which
Example 8.14
computes total losses across series and parallel flow.
If one evaluates a conduit with several valves and fit-
Compute the minimum CV required for a 3 in. natu- tings, one can correlate that the total pressure drop will
ral gas valve supplying gas at a rate of 40,000 SCFH. be equal to the partial pressure drop through each indi-
The supply pressure is at 10 psig and the pressure
vidual element (see Figure 8.11).
drop across the valve must not exceed 2 psig.
Taking the standard volume flow rate of natu-
ral gas, convert to an actual flow rate by taking
the specific gravity of CH4 and average molecular
∆P = ∑ dP (8.43)
weight. The equation herewith gives the actual

gas mixture flow rate: The losses through each element are a function of
the loss coefficient, flow rate, and cross-sectional area.
ft 3 1 h Hence, the series pressure drop will be
40, 000 † ×
h 3600 s
lb lb
 Q2 Q2 Q2  γ
0.075 † 3f †(0.55) † m ∆P1→ 2 = K A 2A + K B 2B + K C 2C  (8.44)
ft SCF ft 3  AA AB A C  2g
× = 6.6 †
 lbf   10 + 14.7  ACF s
 0.075† 3  (0.55)  
ft 14.7  lbm If one assumes constant specific weights throughout the
conduit, then the volumetric flow rates for each element
Using the continuity equation, the velocity of nat- must also be equal.
ural gas will be
Q A = Q B = Q C (8.45)
ft 3
6.6 † 144 †in.2
Q s ft
V= ⇒ × = 134
A  π 2 2 ft 2 s
 3  in.
4
Since the pressure drop should not exceed 2 psig, the

maximum loss factor, Keq can be determined to be
γV 2
∆P = K eq
2g
lb ft A B C
0.069† 3f × 134 2 †
lb ft s
⇒ 2 f2 = K eq
in. ft
2 × 32.2† 2
s
1 ft 2
× FIGURE 8.11
144 in.2 Series flow in a conduit.
This then can be simplified to fl 0.018 × 50 †ft 1

∑ DA 2
=
 3    π 2 2 1 ft 2 
2 = 1494
ft 4
γQ 2  k fl  ft   3 †in.  × †
∆P1→ 2 =
2g  ∑ A + ∑ DA  (8.46)
2 2
12  4 144 in.2 
This equation is applicable for when the flow is incom- while the minor losses resulting from the valves
will be
pressible and steady. As long as the γ1–γ2 ≤ 10%–15%,
one can assume γ1 = γ2. Otherwise, an average specific
   
γ1 + γ 2    
weight γ = should be used.
2 K  4   2  1
∑ A 2
= 2
2
 
2 +
 π  2     π  3    2
2 = 17 , 666 † 4

ft
Example 8.15  ×      ×    
  4  12      4  12   
A 50 ft (15 m) natural gas line with 3 in. (7.5 cm)
pipe, supplies CH4 at a rate of 600 SCF/min
(17 Nm3/min) and 20 psig (1.4 barg). The gas With series and friction losses, the pressure drop
passes through two 2 in. (5 cm) valves (K = 4) and will be
one 3 in. (7.5 cm) valve (K = 2). Compute the pres-
sure drop in the line. 2
lb  ft 3 
The specific weight of natural gas at the given 0.097 † 3f ×  4.2† 
pressure is determined by the specific weight and ft  s   1 1 
gravity of air, expressed as
∆P =
ft 1, 494 †ft 4 + 17 , 666 ft 4 
2 × 32.2† 2
s
lb  20 + 14.7 †psi  lb lbf 1 ft 2

γ = 0.075† 3f (0.55)  = 0.097 3f = 509 × = 3.53 psig
ft 14.7 †psi ft ft 2 144 in 2

Now, since the flow rate is in SCF, first convert to

get the actual volumetric flow rate. 8.9.3.1 Parallel Flow
In some cases, a single flow can diverge into two sepa-
STDft 3 0.075 × .55 1 min ft 3
600 × × = 4.2 rate flows with valves or fittings running parallel with
min 0.097 60 s s
one another before finally converging into a single flow.
A parallel flow configuration complicates calculations
The volumetric flow rate will give the velocity of since the flow is split between two conduits. For these
the fluid to be problems, one can establish equations to determine the
flow through each component. Thus, the pressure drop
ft 3 or other variables in a parallel flow arrangement (see
4.2†
Q s ft Figure 8.12) can be computed.
V= = 2 = 85 †
A π 2 1 ft s Similar to electronic circuits, the pressure drops
 × 3 †in.  × †
2
4 144 in.2 between two components put in parallel to one another

are equal:
The Reynolds number for the natural gas will
then be ∆PA = ∆PB (8.47)
ft  3  Since the pressure drop is equal across both compo-

85† ×   ft
s  12  nents, the component losses across each element as a
Re = 2 = 1.24 × 10
5
4 ft
function of flow rate will be
1.72 × 10 †
s 2 2
Q  Q 
KA  A  = KB  B  (8.48)
Assuming the line is composed of commercial  AA   AB 
steel, the relative roughness will be 0.00015 ft.
Given the Reynolds number and relative rough- The ratio of flow rate for A and B is
ness, the friction factor, as given through the
Moody diagram or Equation 8.37 will yield QA K B AA
f = 0.018. The sum of losses resulting from friction = (8.49)
is then
QB K A AB
KA
AA AA KA
P1 P2 P1 P2
AB
AB KB
KB
FIGURE 8.12
Common arrangement for parallel flow problems.
The Law of Conservation dictates that flow in must Example 8.17

equal the flow out; therefore, the sum of flow between In a flow pipe (Figure 8.14) with a 30 in. (76 cm)
both components must equal the total flow: diameter, it is desired to place a spinner in the
pipe (K = 5). What size should the spinner be to
QA + QB = QT (8.50) yield 15% of the flow to the spinner?
Combining Equations 8.49 and 8.50 after eliminating QA The goal is provide—15% of the total flow to
the spinner—therefore, the remaining flow must
will result in
bypass the remaining pipe area:
K B AA
QB + QB = QT (8.51)
Q T = 0.15†Q A + 0.85†Q B
K A AB
The flow across component B, QB, will then be

B
Ports
QT
QB = (8.52) K = 1.5
K B AA
1+
K A AB QT
Solving for QA will then yield
QT
QA = (8.53)
K A AB
1+
K B AA
A
Example 8.16 Burner
A complex nozzle (shown in Figure 8.13) contains A = 7 ft2
K=4
a 7 ft2 (0.65 m2) burner tip (K = 4) running in par-
allel flow to burner ports (K = 1.5). It is desired to FIGURE 8.13
bypass 20% of the total flow via ports. Estimate Complex parallel nozzle.
the area of the ports.
Since 20% of the flow must bypass the ports,
the remaining flow must pass through the burner
tip. The sum of these two will equal the total flow,
expressed as
Q T = 0.8 †Q Burner + 0.2†Q Ports

The port area should then be
Q Burner KP AB 0.8 1.5 7

= → =
Q Ports K B AP 0.2 4 AP

FIGURE 8.14
A P = 1.07 †ft 2 Parallel flow with spinner.
π 2 The sum of individual component flow rates equal the

AB = 30 − A A total flow rate, therefore, substitution will give the rela-
4
tion between the two in the form of
Applying the ratio of flows to the spinner and
pipe will give K Eq A1 K Eq A 2
QT = QT + QT (8.58)
K1 AR K 2 AR
Q A 0.15 K B AA
= =
Q B 0.85 K A AB Dividing by QT and isolating KEq on one side results in

Equation 8.59.
Substitution of the flow area will give the equa-
tion as a function of spinner area only. 1 1 A1 1 A2
= + (8.59)
K Eq K1 AR K 2 AR
1.1 AA
0.176 = † Lastly, the equivalent loss coefficient for parallel flow
5 π 2 
 30 − A A  can be simplified for all parallel flow cases as
4
2
AA  
0.37 =
706.85 − A A  1 
K Eq =  (8.60)
A A = 191 in. 2  1 A1 + 1 A 2 + … + 1 A n 

 K 1 A R K 2 AR K n A R 
With the area, the diameter of the spinner can be

calculated:
where
D=
4AS
= 15.6 †in.
AR = ∑A + A 1 2 + … + An
π
Making the ratios of flow to the total flow to be
Q1 K Eq A1 (8.61)
8.9.3.2 General Parallel Flow =
QT K1 AR
For general parallel pipe flow, it is oftentimes more con-
venient to lump all parallel branches into a single ele- Q2 K Eq A 2 (8.62)
=
ment with an equivalent loss coefficient. Doing so will QT K 2 AR
reduce the need to iterate for flow rates. Instead, the
equivalent parallel element will only evaluate the pres-
sure drop as a function of the total flow and total area. Example 8.18
First define the total area to be Determine the discharge coefficient for the given
burner configuration seen in Figure 8.15.
A R = A1 + A 2 (8.54) For the given burner, there is a single flow
that converges into the throat, A2, which then
bypasses a spinner before finally exiting at A5. In
Then, the relationship of the pressure losses will be this scenario, the flow must be evaluated in both
series and parallel. Figure 8.16 illustrates what the
Q12 Q2 Q2
K1 = K 2 22 = K Eq 2T (8.55) A1
A12
A2 AR A5
A2 A3
The individual flow rates through each component will
relate to KEq via
K Eq A12 2
Q12 = Q T (8.56)
K 1 A R2 A4
K Eq A 2
2
Q 22 = Q T2 (8.57) FIGURE 8.15
K2 A 2
R Combined series parallel flow in a burner.
A4 K4 the loss coefficient for the burner. For a general

A1 A2 A5 parallel flow situation, KEq for the parallel section
will equal
2
K1 K2 K5  
A3 K3
 1 
K Eq = 
 R3 + R4 
FIGURE 8.16  K3 K4 
Series and parallel flow circuit diagram. (Adapted from Baukal, C.E.,  
The John Zink Hamworthy Combustion Handbook, Vol. 1, CRC Press,
Boca Raton, FL, 2012, Fig. 9.85.) Therefore, the overall loss coefficient through the
burner will be
flow circuit will be. For simplification, take the
ratio of area of each section to a reference ratio,
2
ARef. ARef will be the area measured through the    
throat section:    
 K1 1 K
K Burner = + K2 +   + 25 
 R 12  R3 + R4  R5 
R 1 = A1 A
  K3  
A1 = Blade†area   K4 
r

A 2 = Register†area R 2 = A2 A
r In a typical burner, the values of K and ratios
for each area are calculated via engineering
A 3 = Spinner†area R 3 = A3 A drawings. The K and ratio of the areas to ARef for
r
this configuration are approximately
A 4 = Annulus R 4 = A 4 A
r
K 1 = 1.2 R 1 = 1.3
A 5 = Exit†throat R 5 = A5 A
r K 2 = 0.2 R 2 = 1.0
K 3 = 5.0 R 3 = 0.5
The standard form of the pressure drop is
K 4 = 1.1 R 4 = 0.5
γ K 2

∆P =
2g ∑ A2
QT
K 5 = 0.3 R 5 = 1.6
When expanded, the sum of the loss coefficient  2


K, will be  
   
1 .2 1 0.3 
K Burner =  2 + 0.2 +   +
∆P =
γ 2  K 1 K 2 K Eq K 5 
QT  2 + 2 + 2 + 2 
 1.3  0.5 + 0.5  1.6 2 
  5 1.1  
2g  A1 A 2 A R A 5   

Define the areas in terms of Aref; by doing so, one = 0.71 + 0.2 + 2.03 + 0.11 = 3.05

not only limits the variables, but also provides an
equation relating to Aref, which is commonly used
The relationship between K and Cd is such that
to size burners.
2
γ 2  K1 K 2 K Eq K5   1 
∆P = QT  2 + + 2 +  K= = 3.05 (8.63)
 Cd 
 (R 1A R )
2g A R A R ( R 5 A R )2 
2
Since AR is a common denominator, one can sim- Hence, the discharge coefficient, Cd, will be 0.57.
plify the equation to
γ Q T2  K 1 K 
8.9.4 Discharge Coefficient
∆P = + K 2 + K Eq + 25 
2g ( A R )2  R 12 R5  In boiler applications, nozzles are designed to control

the amount of fuel delivered to a boiler. When engineers
The discharge coefficient is a function of the loss calculate for flow rate through a nozzle, many assump-
coefficient K; therefore, pressure drop and flow tions are made, which do not necessarily translate to
rate can be neglected. Instead, simply solve for the actual flow. These calculations do not account for
Vena contracta Vena contracta
Thin, square Thick, square Tapered Rounded

edge entrance edge entrance entrance entrance
Cd ~ 0.7 Cd ~ 0.8 Cd ~ 0.9 Cd ~ 0.95
FIGURE 8.17
Discharge coefficients for different internal nozzle designs. (Adapted from Baukal, C.E., Industrial Combustion Testing, CRC Press, Boca Raton,
FL, 2010.)
pressure losses, heat transfer, or reduction in cross- L

sectional area. To compensate for the results based off
ideal cases, a correction factor is applied, called the ori-
fice discharge coefficient. The discharge coefficient is
the ratio of actual mass flow rate to ideal mass flow rate, α
expressed as
d V
m D
Cd = actual (8.64)
m  ideal
where Cd is the coefficient of discharge.

Likewise, the coefficient of discharge can also be
related to the area of the nozzle via
Length-to-diameter ratio = L/d
A ideal Reynolds number = V × d/ν
Cd =
A actual Beta ratio = d/D
Port angle =α
FIGURE 8.18
As seen in Figure 8.17, even though the port areas for Factors affecting discharge coefficient. (Adapted from Baukal, C.E., Jr.,
each case are the same, the different interior designs will The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2000,
yield a different discharge coefficient. In most cases, the Fig. 4.32.)
discharge coefficient will vary over a range of 0.6–1.0,
where 1.0 is an ideal nozzle. 8.9.5 Draft
The coefficients of discharge are oftentimes deter-
When dealing with boilers, oftentimes it is important to
mined experimentally. Some factors that affect the coef-
evaluate the pressure in a stack with respect to atmo-
ficient, presented in Figure 8.18, are
spheric pressure. The measure of pressure difference in
a stack is also known as draft. Draft is generally mea-
l
= Length†to†diameter†ratio sured in mm WC or in. WC and often measured with
d an inclined manometer, due to the small pressure dif-
Re = Reynolds number ferences from atmosphere. When evaluating draft, boil-

ers are typically designed such that the pressure in the
α = Port angle stack can be operated under zero gage pressure or less,

d where the stack pressure is less than the atmospheric
= Beta†ratio pressure. Using the Bernoulli equation, the stack draft
D
can be expressed as
A complete listing of discharge coefficient equations

∆PDraft = ( γ Amb − γ FG ) ∆Z (8.65)
can be found in Miller.9
where The aforementioned example assumes no losses due

ΔPDraft is the draft pressure in mm WC or in. WC to friction; however, in real cases, there will be friction
γAmb is the specific weight of atmospheric air losses. Similar to the pressure drop in a straight pipe, the
γFG is the specific weight of flue gas Dary–Weisbach equation can be applied at the stack to
ΔZ is the height of stack in m or ft determine friction losses. When applied to the Bernoulli
equation, the equation will be
The specific weight of the flue gas is a function of both
γV12 γV 2
the temperature of the gas in the stack and the com-
position. An increase in exhaust flue gas temperature
P1 +
2g
+ ∆PDraft = P2 + 2 +
2g ∑Losses (8.67)
decreases the specific weight, causing an increase in the
draft. The specific weight of a hot flue gas can be calcu- Since the draft measures the pressure losses due to
lated as follows: height difference, it would be redundant to include
the height term in the Bernoulli equation, and can be
ignored. Likewise, since the exit of the stack is exposed
 520   288  to atmospheric pressure, P2 will be zero. As explored in
γ FG = γ Amb  460 + T (°F )  † or† γ Amb  273 + T (°C ) 
 FG   FG  previous examples, the inlet velocity generally does not
(8.66) contribute greatly to the overall frictional losses, there-
fore, V1 can be neglected and assumed to be zero. The
simplified equation will then be
where γFG at ST is the specific gravity of flue gas at 60°F
(15°C).
 fL  γV 2
P1 + [ γ air − γ FG ] ∆Z =  1 +  2 (8.68)
 D  2g
Example 8.19
Find the stack draft of flue gas at 300°F (150°C) Example 8.20
with a stack height of 50 ft (15 m). Assume the
molecular weight of flue gas to be 28. Size a stack for P1 = 0.1 in. WC (.25 mbar) to be
To find the stack draft, determine the properties slightly positive with a flow rate of 100,000
for atmospheric air and flue gas, mainly specific lb/hr (45,000 kg/hr) of flue gas at 400°F (200°C).
weight. The specific weight of air at standard tem- The height of the stack is 70 ft (21 m). Assume
perature is γ air = 0.075†lbf /ft 3 †(1.2†kg/m 3 ) and the the molecular weight of flue gas to be 28.
specific weight of flue gas will be Since the Bernoulli equation uses units of lbf /ft 2 ,
first convert P1 from in. WC to lbf /ft 2.
MW  520 
γ FG = 1 †psi 144 †in.2 lb
MWST  460 + Tstack  P1 = 0.1†in. WC × × = 0.52† 2f
27.7 in. WC 1†ft 2 ft

28  520  lb
= γ air = 0.0498 † 3f
28.85  460 + 300  ft The specific weight of the ambient air is
0.075 lbf /ft 3 , hence the specific weight of flue gas
at 400°F will be
where the molecular weight of flue gas at standard
temperature is 28.85 lbm /mol . Application of lbf  28.0   520  lbf
Equation 8.65 will give the stack draft of 0.075 ×   = 0.044 † 3
ft 3  28.85   460 + 400  ft
With the pressure and specific weights, solve for

 lb lb 
∆Pdraft = ( γ air − γ FG ) ∆Z =  0.075† 3f − 0.0498 † 3f  50 †ft the total losses.
 ft ft 
lbf 1 ft 2 ∆P1 + ( γ air − γ FG ) ∆Z = Losses

= 1.26 × = 0.00875psi
ft 2
144 in.2 lbf  lb lb  lb
→ 0.52 +  0.075 3f − 0.044 3f  70 ft = 2.69 2f
ft 3
 ft ft  ft
Converting to in. WC will yield
Using Equation 8.68, the losses are expressed as
in. WC lbf  fL  γV 2
0.00875psi × 27.7 = 0.24 in. WC 2.69 = 1 +  2
psi ft 2
 D  2g

However, since the major losses have yet to be

determined, make a guess. For typical boilers,
the stack losses are usually ≤1. For this example, 8.10 Compressible Flow
make a guess of fL/D to be 0.5. Solving for the The topics and equations presented in the earlier sec-
flue gas velocity will then result in
tions have all made basic assumptions on a variety of
properties, such as steady or fully developed flow, no
lbf γV 2
2.69 = (1 + 0.5) 2 viscous effects, or along a streamline. In several cases,
ft 2
2g evaluation of a problem could only be solved by assum-

ing the state of flow to be incompressible. A flow can be
lbf  ft  designated as incompressible when the density of a fluid
2.69 × 2  32.2 2  will not vary as a function of pressure. For liquids like
ft 2  s  ft
V2 = = 51.2 water, one can assume an incompressible flow due to the
lbf s
1.5 × .044 3
ft relatively small change in volume, even under high pres-

sure loads. Therefore, for engineering purposes, it is fre-
With the mass flow rate given, take the velocity of quently assumed that a flow is incompressible as long as
the flue gas and evaluate for the stack area. the density change is less than about 5%. For flows that
vary drastically with pressure gradients, fluid properties
lb 1 h must take into account the compressibility of the flow.
100 × 10 3 † m ×
m h 3600 s = 12.33 ft 2
m
 = ρAV ⇒ A = =
ρV lbm ft
0.044 † 3 × 51.2
ft s 8.10.1 Mach Number
The most important factor in determining the type of
Therefore, the diameter of the stack will be
flow a fluid is undergoing is a nondimensional number
known as the Mach number, named after the Austrian
π 2 4A 4 in.2 physicist Ernst Mach. The Mach number is the ratio of
A= D ⇒ = × 12.3 †ft 2 × 144 † 2 = 47.5†in.
4 π π ft actual velocity of a fluid to the speed of sound in the
same fluid, expressed as
Since the friction factor was assumed to be 0.5,
check the answer to test the validity. V
M= (8.69)
c
ft
51.2† × 3.9†ft where c is the speed of sound in fluid in m/s or ft/s.
Re = s
2 = 4.38 × 10
5
−3 ft
Flow is evaluated into distinct regimes depending on
0.455 × 10 † the value of the Mach number and is separated as such
s
Using steel for the stack material, the roughness M < 1: Subsonic flow
factor, ε/D , will be 0.0000384. With the follow- M ≈ 1: Transonic flow
ing friction factor and Reynolds number, the M > 1: Supersonic flow
friction factor, as dictated by the Moody dia-
gram is approximately 0.013. The friction factor For all flow with a Mach number greater than 0.3 or
will then be flows with significant pressure variations, the fluid
fL 0.013 × 70 must be assessed using the principles of compressible
= = 0.23 flow. Note that because the speed of sound is measured
D 3.9
with respect to a fluid, c will vary depending on the
Since the friction factor is lower than the initial fluid state. The derived form for c as a function of fluid
guess, evaluate for the new velocity and diameter. properties is presented as
 0.044 × V22  ft 1545

2.62 = (1 + .23 )  ⇒ V = 55.8 † c = kRT or †† k ††Tr †32.2 (8.70)
 2 × 32.2  s MW

The stack diameter will then be where
k is the ratio of specific heats of fluid, Cp/Cv
lb R is the gas constant in J/(kg · K) or Btu/(lbm · °R)
4 × 100, 000 † m in.
D= h × 12 = 45.5
in. T is the temperature of fluid in K or °R
h lbf ft ft Tr = °R
3, 600 π × 0.044 † 3 × 55.8 †
s ft s MW is the molecular weight of the gas
8.10.2 Quasi–One-Dimensional Isentropic Flow One common practice to induce compressible flow
would be to use a converging–diverging nozzle. A con-
In fluid flow through devices such as nozzles or diffusers,
verging–diverging nozzle consists of a wide opening,
the flow can be thought of as quasi–one-dimensional. In
which decreases in area to a throat section and subse-
a one-dimensional flow regime, the cross-sectional area
quently increases again, similar to a venturi nozzle. The
must be constant; quasi–one-dimensional flows can be
main purpose of the nozzle is to accelerate a fluid to
assumed one-dimensional due to gradual expansion or
supersonic speeds.
contraction of the flow area. In this type of flow, the vari-
ations in the normal directions to the flow are neglected.
Variables such as P, ρ, and T are treated as functions of x Example 8.21
only. Isentropic flow implies an adiabatic flow with no Natural gas supplied at 10 psig (0.7 barg) and
heat transfer. The following sections will assume isen- 100°F (37°C) flows through a nozzle, which is
tropic flow. exposed to open air and supplies a heat input
of 50 × 10 6 Btu/h (15 MW) on a LHV basis. The
coefficient of discharge is 0.75. Determine the
8.10.3 Mach Number Relationships required flow area.
For natural gas, the properties of fluid are
For a one-dimensional isentropic compressible flow, the
Mach number has a large impact on the fluid proper- Btu  J 
ties. Using thermodynamic relations for ideal gases, the R = 3098 † 518 †
lbm ⋅ °R  kg ⋅ K 
pressure, temperature, and density can all be related
as a function of the Mach number. Without going into k = 1.31
much detail, application of the ratio of specific heat and
ideal gas laws, equations relating to fluid properties are The first step to approaching this problem is to
expressed as calculate for the Mach number to determine the
flow regime. Using the total pressure relationship,
 k −1 2  solve for the Mach number. Since the nozzle exit
Tt = T  1 + M  (8.71) is open to atmospheric air, the static pressure is
 2 
found to be 0 psig (0 barg) or 14.7 (1.013 bara) psia.
where
k/(k − 1)
Tt is the total temperature  k − 1 2
PT = P  1 + M 
T is the static temperature  2 
1.31/(1
1.31 − 1)
In this equation, the static temperature refers to the  1.31 − 1 2 
→ 24.7 = 14.7  1 + M 
temperature of the fluid as measured by a static device  2 
with respect to the flow. For example, a thermometer
moving with the fluid will measure the static tempera- Solving for the Mach number will yield M = 0.91,
ture of said fluid. The total temperature is the temperature therefore the flow is subsonic. With the Mach num-
ber, the density of natural gas at the nozzle will be
the fluid will have if the velocity were reduced to zero
adiabatically. 1/(k − 1)
 k − 1 2 lbm
Pressure can be expressed similarly with respect to ρT = ρ  1 + M  → 0.06560 †
the Mach number as  2  ft 3
1/(1.31 − 1)
k/(k −1)  1.31 − 1  lbm
 k −1 2  = ρ1 + 0.912  = 0.04121
Pt = P  1 + M  (8.72)  2  ft 3
 2 
Likewise, the Mach number can be used to deter-

where Pt is the total pressure. mine the fluid velocity via the Mach number relation.
The total pressure also refers to the pressure associ-
ated with a fluid decelerated to zero velocity in an adia- V
M=
batic manner. c
Lastly, the density as a function of Mach number is The speed of sound is determined to be c = kRT ,
making the equation
1/( k −1)
 k −1 2 
ρt = ρ  1 + M  (8.73)
 2  V ft
0.91 = → V = 1371†
1.31 × 3098 × (100°F + 459.67 ) s
where ρt is the total density.
From the given rate of heat transfer for the nozzle, medium. Since the Mach number for choked flow
first calculate for the mass flow rate of natural is 1, the velocity of the fluid will be equal to the
gas. speed of sound.
Btu 1 lb 1 h lb ft
 = 50 × 10 6
m × † m × = 0.65 m V = kRT = 1.31 × 3098 × 632.83 = 1603†
h 21, 495 Btu 3600 s s s
With the density, velocity, and mass flow rate of Again, the continuity equation will provide a
natural gas, one can find the nozzle area using the flow rate of
continuity equation:
lb  1 ft 2  ft
  = ρAV = 0.04912† m3 × 3 ×
m × 1603†
Q = AV → m
 = ρAV ft 
 144 in.2  s

lb lb
0.65† m = 1.64 † m
m s in.2 s
A= = × 144 † 2 = 1.65†in.2
lb
ρV 0.04121† m × 1371†ft ft
ft 3 s
The actual flow rate, through the nozzle will be
The following area corresponds to an ideal flow
m
 actual m
area. When the discharge coefficient if applied the Cd = → 0.8 = actual
actual area will be m ideal 1.64
lb
A 1.65†in.2  actual = 1.31† m
m
A actual = = = 2.20 in.2 s
Cd 0.75
The heating input can be found using the mass
flow rate and the LHV for methane and is found
Example 8.22 to be
If natural gas enters a converging nozzle at 20
psig (1.4 barg) and 200°F (95°C), calculate the lb Btu s
 actual × LHV = 1.31† m × 21, 945†
Q=m × 3, 600 †
mass flow rate and total heat input on a LHV s lbm h
basis for a 3 in.2 (20 cm 2) nozzle with a Cd of 0.8,
Btu
which is exposed to open air. = 103 × 10 6 †
First, start by evaluating the flow and determin- h
ing the type of flow associated with the gas.
k/(k − 1)
 k − 1 2
PT = P  1 + M 
 2 
1.31 − 1 2 
1.31/(1
1.31 − 1) 8.11 Free Jet

→ 34.7 = 14.7  1 + M 
 2  Figures 8.1910 and 8.20 show a fluid jet emerging into
a quiescent, surrounding fluid; this is referred to as a
M = 1.03 free jet. The white streaks in the photograph are small
particles that have been released into the surrounding
When the Mach number exceeds a value of 1, the fluid, just upstream of the jet. A low pressure region cre-
flow can be considered to be choked. Given that ated within the free jet causes the particles to flow into
the flow has become choked, the Mach number
the path of the jet and mix, oftentimes referred to as free
will be evaluated at M = 1. The static temperature
and density are then determined to be jet entrainment.
The structure of a free jet can be separated into three
 1.31 − 1 2  regions: potential core, transition, and fully devel-
200°F = T  1 + 1  = 173°F
 2  oped, as illustrated in Figure 8.21.11 The potential core
is located just downstream of the nozzle exit and rep-
lbm 1.31 − 1 2 
1/(k − 1)
lbm
resents the region where the centerline of the fluid jet

0.07819 = ρ1 + 1 = 0.04912 has not mixed with the ambient fluid. Mixing at the jet
ft 3  2  ft 3
centerline does not occur until about three to five nozzle
The velocity of natural gas is a function of the diameters downstream of the nozzle exit; at this loca-
Mach number and the speed of sound for a given tion, the mixing spreads to the centerline, and the core
region no longer exists. Beyond the core region, the flow

enters the transition region. Within the transition region
the jet continues to spread as the velocity decreases at
a rate necessary to conserve axial momentum. Within
the fully developed region, the velocity profile of the jet
begins to take on a self-similar shape. That is, the center-
Gas nozzle line velocity, within the fully developed region of a cir-
cular, free jet (as seen in Figure 8.2112) can be described
by the following equation14:
Free jet
Vo x
= 0.16 − 1.5 (8.74)
Vm do
where
Vo is the initial velocity at nozzle exit
Vm is the maximum velocity of jet downstream
do is the nozzle diameter
FIGURE 8.19
Notice the aforementioned equation is dimensionless;
Photograph of a gas exiting a nozzle. The white streaks are small par-
ticles that have been released in the vicinity of the jet. (Courtesy of the only requirement for accuracy is that consistent units
Onera-The French Aerospace Lab.) for velocity and length be used. The velocity profile in
FIGURE 8.20
Mixing downstream of a free jet. (From Van Dyke, M., An Album of Fluid Motion, The Parabolic Press, Stanford, CA, p. 97, 1982.) (Courtesy of
Paul E. Dimotakis.)
Potential core Transition Fully-developed
Gas jet
Centerline of jet
Velocity profile
FIGURE 8.21
General structure of a turbulent free jet. (Adapted from Saddington, A.J. et al., Aeronaut. J., Paper No. 2825, 2004.)
the radial direction can be described by the following realize that while many of the equations have power-
equation: ful applications, they are constrained by assumptions
regarding flow, flow states, and fluid properties. The
V   r  information provided in this chapter only provides a
= exp  −K u    (8.75)
Vm   x  handful out of numerous topics regarding the subject.
Those interested should explore further texts relating
where
to the topic.
V is the actual velocity within the free jet
Vm is the maximum velocity at x − location
r is the radial distance from jet centerline
K u is the Gaussian constant equal to about 92
Equations for concentration decay are similar to the References

velocity equations, and are given by
1. G.A. Tokaty, A History and Philosophy of Fluid Mechanics,
Co x Dover Publications, New York, 1994.
= 0.22 − 1.5 (8.76)
2. C.T. Crowe, Engineering Fluid Mechanics, 9th edn., John
Cm do
Wiley & Sons, Inc, Denver, CO, 2009.
3. J. Nikuradse, Stromungsgesetze in rauhen Rohren,
C   r 
= exp  −K c    (8.77) VBI-Forschungsheft, 1933, 361. Translation available in
Cm   x  N.A.C.A. Tech. Memorandum, 1292.
where13 4. Hydraulic Energy Transmission and Control, Hill Learning
C is the actual concentration Systems, 2001–2007, http://www.hetacfluidpower.com/
Co is the initial concentration screen_shot_gallery.htm
Cm is the maximum concentration of the jet downstream 5. E.G. Pita, Air Conditioning Principles and Systems: An
x is the distance from nozzle exit Energy Approach, 4th edn., Prentice Hall, Upper Saddle
River, NJ, 2002.
do is the nozzle diameter
6. P.K. Swamee and A.K. Jain, Explicit equations for pipe-
r is the radial distance from jet centerline
flow problems, Journal of Hydraulic Division of the ASCE,
Kc is the Gaussian constant equal to about 55.5 102(HY5), 657–664, May 1976.
7. I.E. Idelchik, Handbook of Hydraulic Resistance,
It is important to note that the concentration of a free Hemisphere, New York, 1986.
jet is independent of velocity. Intuition would lead one 8. Engineering Division Crane, Flow of Fluids through Valves,
to believe that the faster a jet exits an orifice, the faster it Fittings, and Pipe, Crane Co., New York, 1969.
mixes with the surrounding fluid. While it is true that 9. R.W. Miller, Flow Measurement Engineering Handbook,
it entrains more mass, it is not mixing at a faster rate; McGraw-Hill, Inc., New York, 1989.
that is, downstream concentrations are not affected by 10. ONERA Photograph, Werlé, An Album of Fluid Motion,
the free jet exit velocity. The Parabolic Press, Stanford, CA, 1982, p. 99.
11. M. Van Dyke, An Album of Fluid Motion, The Parabolic
Press, Stanford, CA, 1982, p. 97.
12. A.J. Saddington, N.J. Lawson, and K. Knowles, An
experimental and numerical investigation of under-
expanded turbulent jets, The Aeronautical Journal, March
8.12 Conclusions 2004, Paper No. 2825.
13. M.A. Saad, Compressible Fluid Flow, Prentice Hall,
As discussed in the introduction, the study of fluid Englewood Cliffs, NJ, 1985.
dynamics plays a pivotal role in the design and oper- 14. J.A. Roberson and C.T. Crowe, Engineering Fluid
ation of combustion equipment. It is important to Mechanics, Houghton Mifflin, Boston, MA, 1980.
9
CFD-Based Combustion Modeling
Michael A. Lorra and Shirley X. Chen
CONTENTS
9.1 Introduction.................................................................................................................................................................... 183
9.2 CFD Model Background............................................................................................................................................... 185
9.2.1 Transport Equations.......................................................................................................................................... 185
9.2.2 Turbulence Models............................................................................................................................................ 186
9.2.3 Algebraic Models............................................................................................................................................... 186
9.2.4 One- and Two-Equation Models...................................................................................................................... 186
9.2.4.1 The k−ε Turbulence Model................................................................................................................ 187
9.2.4.2 k−ε Turbulence Model Boundary Conditions................................................................................. 189
9.2.5 Other Turbulence Modeling Approaches....................................................................................................... 191
9.3 CFD-Based Combustion Submodels........................................................................................................................... 192
9.3.1 Regimes of Turbulent Combustion................................................................................................................. 192
9.3.2 Reaction Kinetics............................................................................................................................................... 193
9.3.3 Eddy Breakup Model......................................................................................................................................... 194
9.3.4 Eddy Dissipation Combustion Model............................................................................................................. 194
9.3.5 Eddy Dissipation Concept................................................................................................................................ 195
9.3.6 Mixture Fraction Approach for Equilibrium or Finite Rate Chemistry.................................................... 195
9.3.7 Pollutant Chemistry Models............................................................................................................................ 197
9.4 Radiation Models........................................................................................................................................................... 198
9.4.1 P-1 Radiation Model.......................................................................................................................................... 198
9.4.2 Discrete Ordinates Radiation Model.............................................................................................................. 198
9.4.3 Monte Carlo Method......................................................................................................................................... 200
9.4.4 Gas-Radiation Properties.................................................................................................................................. 200
9.4.5 Weighted Sum of Gray Gases........................................................................................................................... 201
9.4.6 Effect of Soot on Thermal Radiation............................................................................................................... 201
9.5 Solution Methodolgy..................................................................................................................................................... 201
9.5.1 Problem Setup: Preprocessing......................................................................................................................... 202
9.5.2 Solution Convergence........................................................................................................................................ 204
9.5.3 Analysis of Results: Postprocessing................................................................................................................ 204
9.6 Summary......................................................................................................................................................................... 206
Nomenclature.......................................................................................................................................................................... 206
References................................................................................................................................................................................. 209
in combustion equipment, see also Figure 9.1, to a

generic tool that is theoretically capable of simulating
9.1 Introduction
complex flow phenomena with respect to geometry
Computational fluid dynamics (CFD) has become a and physics.
widely accepted tool to help in the design and opera- Simulation of reactive flows with complex reaction
tion of equipment in various industries. It evolved mechanisms, convective and radiative heat transfer,
from a tool to simulate isothermal flow distribution in phase changes of liquid material, and volatilization of
simpler geometries, as for example flow distribution coal particles are now capabilities of commercially
183
of full-scale combustion chambers in a way that

is efficient enough to be achievable for computa-
tions on a reasonable amount of computers in a
reasonable amount of time.
4. Uncertainty of boundary conditions
a. For a range of applications the exact bound-
ary conditions are unknown and can only
be estimated or given as a “temporal snap-
shot.” One example here would be a CFD
model of a ground flare exposed to ambi-
ent conditions with varying conditions for
wind speed and wind direction. Data can
be obtained for median values as recorded
by discrete measuring stations like weather
stations on airports; but the local conditions,
FIGURE 9.1
also affected by the surrounding terrain,
Investigation of an isothermal flow field.
remain unknown. This prompts the need for
an extended domain to capture topographi-
available CFD packages. However, even today there are cal data in order to obtain a solution that is
still some caveats that limit the use of a CFD model for a more representative for the impact of any
certain application or problem at hand: given wind condition to the object of interest.
An example is shown in Figures 9.2 and 9.3,
1. Limited theoretical understanding of the physics showing the original topographical data and
involved in reactive flow systems. the representation in a CFD model, respec-
tively. The blue color in Figure 9.2 depicts
a. This includes uncertainty in fundamental
lower ground elevation; the red indicates
chemical reaction kinetics for complex fuels
higher ground elevation.
(higher hydrocarbons), the understanding
and mathematical description of the chaotic
A CFD code or a specific submodel for this code has a
nature of turbulent mixing, the more complex
three-fold basis:
heterogeneous chemistry related to liquids or
solids (e.g., soot formation, catalysis, evapora-
1. The development of a CFD code or specific sub-
tion of multicomponent droplets, etc.).
model requires the understanding of the physi-
2. Uncertainty of thermodynamic properties cal phenomena involved in any fluid dynamic
a. One problem of using CFD models for com- process.
plex combustion systems such as incinera- 2. The “real world” physics of the flow phenom-
tors stems from the lack of thermodynamic ena must be translated into mathematical
data of species involved in the combustion
process. The datasets available and accepted
worldwide include a variety of hydrocarbon
species ranging from C1 to C10, including all
intermediate radicals. As soon as the process
involves more “exotic” species, as or example
various sulfur compounds, thermodynamic
data become unavailable or various sources
might offer differing data.
3. The large disparity in both spatial and temporal
time scales in typical reacting flow systems.
To capture the flow characteristics responsi-
ble for momentum transfer one needs to resolve
the flow to the smallest characteristic size (i.e.,
Kolmogorov scale), typically in the order of
magnitude of micrometers for most laboratory FIGURE 9.2
flows. This poses a major problem for the model Original topographic data.
CFD-Based Combustion Modeling 185
FIGURE 9.3
Representation of topographical data in a CFD model (blue showing
lower elevation, red showing higher elevation).
FIGURE 9.5
Representation of a burner, colored by temperature (blue showing
low temperatures, red high temperatures).
to determine if the results are correct within its context.

When used appropriately, CFD can help reduce devel-
opment cycle time, potentially improves efficiency of a
process or can extend the lifetime of the equipment.
Figures 9.4 and 9.5 are examples of the use of CFD for
a new burner development. In a project like this, results
from the real world, established through measurements in
a test furnace and the results from a CFD model can com-
plement each other leading to shorter development times.
9.2 CFD Model Background

FIGURE 9.4
Close-up view of a burner in a test furnace. All CFD codes have evolved from similar backgrounds
being based on the Semi-Implicit Pressure Linked
Equations (SIMPLE) algorithm, which solves a set
relationships. The developer of a CFD model (or of nonlinear coupled partial differential equations
submodel) would have to find the functional or describing the conservation of mass, momentum, and
parametric relationship between cause and effect. energy, described later. To illustrate the mathematical
3. Due to the complexity of nonlinear mathematics, basis from which CFD codes are derived, a brief review
the governing equations describing the flow phe- of the general make-up of a CFD code is given in the
nomenon must be solved using numerical algo- following section.
rithms to make those digestible for a computer.
Error in any one translates into erroneous predic- 9.2.1 Transport Equations
tions, and failing to understand the limitations of each Nonreactive turbulent fluid flow is modeled using the
can lead to incorrect conclusions and potentially cata- steady-form of the Navier–Stokes equations by assum-
strophic results. For optimal use, an engineer utilizing ing a continuous flow field described locally by the gen-
CFD code must understand the physics, the mathemat- eral conservation of mass and momentum1:
ics, and the numerical approach to determine if a simu-
Dρ
lation produces sound results. In addition, knowledge = −ρ(∇ ⋅ v) (9.1)
about the application that has to be modeled is crucial Dt
Dv These represent the turbulent transport equation set

ρ = −∇p − ∇ ⋅ τ + ρ g (9.2) for nonreacting flow. However, as a result of the averag-
Dt
ing procedure, several additional variables called Favre-
where D/Dt is the total (or substantial) derivative. stresses ( ρui′′u′′j ) and Favre-fluxes (ρφ′′u′′j ) have been
These equations may be simplified by assuming introduced. These stresses and fluxes represent the mean
steady-state flow of a Newtonian fluid. If the fluid  is momentum transport and the mean scalar transport by
assumed to be noncompressible, the dilatation (∇ · v ) can turbulent diffusion. Additional equations are required
also be used to further simplify the equations. A non- to solve for these new turbulent transport variables.
compressible assumption is applicable to low Mach
number flows (i.e., Mach No. <0.3) and for incompress-
ible flows. Given these simplifying assumptions, the 9.2.2 Turbulence Models
general conservation or transport equations for mass There are generally four types of turbulence models
and momentum for steady flow can be written in commonly used in engineering practice.
Cartesian tensor notation as
• Algebraic models
∂(ρu j )
= 0 (9.3) • One-equation models
∂x j
• Two-equation models
∂(ρuiu j ) ∂P ∂τ ij • Second-order closure models
=− + + ρf i (9.4)
∂x j ∂x i ∂x j
Wilcox2 and Rodi3 provide good reviews of various
 ∂u ∂u j   2  ∂uk methods of turbulence modeling. Most of the models,
τ ij = µ  i +  +  3 µ − µB  ∂x δ ij (9.5) which fall into one of the first three categories, have
∂
 j x ∂ x i k
an underlying commonality, which is the Boussinesq
Similarly, a transport equation can be written for any approximation. The Boussinesq approximation assumes
conserved scalar, Φs: that the Reynolds-stress tensor is proportional to the
mean strain-rate tensor, for every location in a turbu-
lent flow. The constant of proportionality between the
∂(ρΦu j ) ∂  ∂Φ 

∂x j
=
∂x j  Γ φ ∂x  + SΦ (9.6) Reynolds-stress tensor and the mean strain-rate tensor
 j is the eddy viscosity (μT).
These equations are valid for all flow regimes from
laminar to turbulent. However, for all practical pur- 9.2.3 Algebraic Models
poses, these equations cannot be solved directly for tur-
Many algebraic models calculate an eddy viscosity
bulent flows. Hence, the need for a statistical approach
from the Prandtl mixing-length hypothesis, given by
to model turbulent flow regimes. The approach is rather
Equation 9.11:
simplistic; a simple decomposition into mean values
and fluctuations with a zero mean value.
Φ = Φ + φ′′ (9.7) ( ) ∂∂Uy

µ T = ρ lm 2 (9.11)
where Φ = ρΦ/ρ and ρφ′′ = 0, but φ′′ ≠ 0. where

Applying this decomposition to all variables except lm is the mixing length
density and pressure, the conservation equations U is the mean velocity
(Equations 9.3 through 9.6) are transformed into the
mass-averaged or Favre-averaged transport equations: Algebraic models are not very general, because the
mixing length depends on the flow scenario (i.e., jet,
∂(ρu
j ) boundary layer, pipe flow, etc.). Therefore, an expression
=0 (9.8)
∂x j for the mixing length must be obtained for each type of
flow when using an algebraic model.
(
∂ ρu
iu
j ) = − ∂P +

∂x j ∂x i
∂
∂x j
( )
τij − ρui′′u′′j + ρf i (9.9)
9.2.4 One- and Two-Equation Models

(
∂ ρΦu j )= ∂  ∂Φ
Γφ

− ρφ″ u′′j  + SΦ (9.10)
One- and two-equation models attempt to overcome the
∂x j ∂x j  ∂x j 
difficulties with the Prandtl mixing-length hypothesis by
solving transport equations for quantities that are related to
the type of flow. In the algebraic model mentioned earlier,

there is a direct link between the fluctuating velocity scale
Measured quantity in a turbulent flow

and the mean velocity gradients, as shown by Equation
9.11. In one- and two-equation models, the link between
the fluctuating velocity scale and the mean velocity gra-
dients is found by solving one or more transport equa-
tions. Wilcox2 describes two, more recent, one-equation
turbulence models, which appear to show promise. The
Baldwin–Barth4 model employs a transport equation for
the turbulent Reynolds number and the Spalart–Allmaras5 Time-averaged value
model employs a transport equation for the eddy viscosity.
Prandtl chose the turbulent kinetic energy as the basis
for a turbulent velocity scale.
1 ′ ′ 1  ′2  Time
k= ui ui =  u + v′ 2 + w′ 2  (9.12)
2 2  
FIGURE 9.6
The eddy viscosity can be computed using the Point measurement of a scalar in a turbulent flow. (From Baukal,
Kolmogorov–Prandtl expression: C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca
Raton, FL, 2001.)
µ T = C1ρ kl (9.13)
where of interest such as pressure and velocity fluctuate in an

k is the turbulent kinetic energy apparently random fashion. Analysis shows that these
l is the turbulence length scale quantities are not truly random.6 Information revealed
C1 is the constant of proportionality by spectral analysis of point measurements indicates that
there are ranges of temporal and length scales, which con-
Wilcox2 provides a derivation of the Reynolds-stress tain significant energy (the large or integral scales) and
equation, obtained by taking moments of the Navier–Stokes smaller scales where this turbulent energy is dissipated
equation. The Reynolds-stress equation is a tensor equa- by viscous processes. The energy cascade is the mecha-
tion, and by taking the trace of it, a transport equation is nism by which energy is moved from the large scales to
obtained for turbulence kinetic energy. This transport the small scales. For more information on the physics of
equation for the turbulence kinetic energy forms the basis turbulent flows, the reader should refer to Libby.6
for a number of one- and two-equation turbulence mod- Prediction of turbulent flow from the Reynolds-
els. The turbulence kinetic energy transport equation has averaged (time-averaged) or Favre-averaged conser-
various terms, which have been given physical interpreta- vation equations requires closure approximations.
tions. Included in the terms is a dissipation term (ε), which This is because the time-averaged conservation equa-
represents the rate at which turbulence kinetic energy is tions contain terms that are not known. In the case
converted to thermal internal energy. From the turbulence of the momentum equations, the time- averaging of
kinetic energy equation, ε is the product of the viscosity the convection terms leads to the Reynolds stresses:
and the square of the fluctuating vorticity. Therefore, based ρu′ 2 , ρv′ 2 , ρw′ 2 , ρu′ v′ , ρu′ w′ , ρv′w′. Closure approxima-
on dimensional arguments, the dissipation (ε) should be tions are required to solve the Reynolds averaged con-
related to the turbulence kinetic energy and length scale: servation equations.
The workhorse turbulence model used in furnace sim-
ulations is the standard k−ε model.7 The popularity of
C 2 k 3/ 2
ε= (9.14) this model can be ascribed to its relative simplicity (com-
l pared to a Reynolds’ stress model, for example) and its
good performance in a variety of engineering flows. Its
where C2 is the constant of proportionality.
weaknesses include its performance in unconfined flows,
in rotating and swirling flows, and in flows with large
strains, such as curved boundary layers. The Reynolds
9.2.4.1 The k−ε Turbulence Model
stress model (RSM) addresses some of these performance
A turbulent flow is a flow with a wide range of tempo- issues. RSM is much more computationally demanding
ral and length scales. Figure 9.6 is an example of a typi- because it involves seven extra partial differential equa-
cal point measurement (e.g., pressure or velocity) within tions rather than the two of the k−ε model. However, in
a turbulent flow. Within a turbulent flow, the quantities a typical combustion calculation, the number of PDEs
solved is typically quite large, so adding five more may be Given these definitions, additional transport equations
easily justified if the quality of the prediction improves. for the turbulent kinetic energy k and the dissipation rate
A number of variants of the classical k−ε model exist. of turbulent kinetic energy, ε can be written, respectively,
The classical k−ε model uses a single eddy viscosity in
all directions. The nonlinear k−ε model by Speziale8 ∂k  1 ν 
addresses this assumption, which is known to be poor + v ⋅ ∇k = ∇ ⋅  t ∇k  + G − ε
∂t ρ  σk 
even in relatively simple flows. Another development in
νt  2   ∂ui ∂u j  
k−ε modeling is the renormalization group (RNG) k−ε 3 3
model of Yakhot et al.9 Its performance in complex flows

has been promising, so much so that several of the com-

G=
σk 
2∇ v + ∑∑  ∂x + ∂xi  
 j 
(9.18)
 i = 1, i ≠ j j =1 
mercial CFD code vendors have implemented the RNG
k−ε model. The realizable k−ε10 represents yet another ∂ε  1 ν   ε  ε2 
variant recently introduced. The advantages and limita- + u ⋅ ∇ε = ∇ ⋅  t ∇ε + f1c1G   − f 2c2   (9.19)
∂t ρ  σε   k  k
tions of these turbulence models are discussed in more
detail in Veersteeg and Malalasekera.11 with the eddy viscosity defined as
Perhaps the most commonly used turbulence model
for practical flow problems is the k−ε model. The earliest f µ cµ k 2
developments related to the k-ε model were by Chou,12 νt = (9.20)
ε
Davidov,13 and Harlow and Nakayama.14 The most
well-known paper on the k-ε model is that of Jones and Several key “empirical” constants are required by
Launder,15 which according to Wilcox2 has almost reached the k−ε turbulence model. An accepted set of constants
the status of the Boussinesq and Reynolds papers within is shown in Table 9.1. These values are those originally
the turbulence modeling community. The k−ε model is proposed by Launder and Spalding,7 but differ slightly
based on the turbulence kinetic energy transport equa- from those reported by other researchers.15,18,19 This may
tion discussed earlier, and a second transport equation be due to the fact that these constants are based on simple
for the dissipation rate. The equation for the dissipation two-dimensional flows and adjustment may be required
rate (ε) is derived by taking a moment of the Navier– to simulate more complex flows. Regardless, this fact
Stokes equation using the fluctuating vorticity. and the other simplifying assumptions suggest that the
Although several turbulence models have been pro- flow results be closely scrutinized when applying any
posed,16 the k−ε turbulence model, originally proposed CFD code using this turbulence model to simulate com-
by Harlow and Nakayama,17 remains the most widely plex flow systems. The application of the k−ε turbulence
used model to describe practical flow systems.8 The k−ε model requires boundary conditions for both k and ε.
turbulence model employs a modified version of the Boundary layer theory could be used to derive the
Boussinesq hypothesis: equations for flow near the wall, but to reduce computer
 ∂u ∂uj  2 storage and run times the k−ε turbulence model uses
−ui′′u′′j = − νt  i +
∂ x ∂ x  − 3 ( νt∇ ⋅ v + k ) δ ij (9.15) wall functions instead. The Van Driest hypothesis on
 j i
turbulent flow near walls is used to derive wall func-
where νt is known as the eddy diffusivity or turbulent tions consistent with the logarithmic law of the wall.7
viscosity. This approach allows the molecular viscosity These functions relate the dependent variables near the
to be replaced with the eddy (or turbulent) diffusivity, wall to those in the bulk flow field. Given the turbulence
which allows the instantaneous transport equations model with the necessary boundary conditions, the full
(Equations 9.3 through 9.6) to be modeled using the equation set may be written (see Tables 9.2 and 9.3).
mean-value equations (Equations 9.8 through 9.10). A As shown, each equation is conveniently cast into a
disadvantage to this approach is the need to assume iso- general convection-diffusion form with the off terms
tropic eddy diffusivity. However, given this assumption collected on the right-hand side (RHS), the specific
and the specific velocity- and length-scales (u′, l′) terms depend on the coordinate system selected.
u′ ≈ k
TABLE 9.1
3 (9.16) Universal “Empirical” Constants Used in the
k 2
l ′ ≈ Cµ
Standard k−ε Turbulence Model
ε
Constant Cμ C1 C2 σk σe κ
where the turbulent kinetic energy k can be defined as
Value 0.09 1.44 1.92 1.0 1.3 0.4187
1 1
2 2
(
k = uiui = u1u1 + u2u2 + u3u3 ) (9.17) Baukal, C.E. (ed.), The John Zink Combustion
Source:
Handbook, CRC Press, Boca Raton, FL, 2001.
TABLE 9.2
Cartesian Differential Equation Set
∂(ρu
φ ) ∂(ρvφ) ∂(ρw
 φ) ∂  ∂( φ )  ∂  ∂( φ )  ∂  ∂( φ ) 
+ + −  Γφ  −  Γφ  −  Γφ  = Sφ
∂x ∂y ∂z ∂x  ∂x ∂y  ∂y ∂z  ∂z
Equation φ Γφ Sφ
Continuity 1 0 0
X-momentum u
 ∂p ∂  ∂u
 ∂  ∂v  ∂  ∂w
 2 
μe − +  µ e +  µ e +  µe  + ρg x − ρk
∂x ∂x ∂x  ∂y ∂x  ∂z  ∂x  3
∂p ∂  
∂u ∂  ∂v  ∂  
∂w 2 
Y-momentum v μe − +  µ e  + ∂y  µ e  + ∂z  µ e  + ρg y − ρk
∂y ∂x ∂y  ∂y  ∂y  3
∂p ∂  ∂u
 ∂  ∂v  ∂  ∂w
 2 
Z-momentum w
 μe − +  µ e +  µ e +  µe  + ρg z − ρk
∂z ∂x ∂z  ∂y ∂z  ∂z  ∂z  3
µe
Mixture fraction Z 0
σf
Cg 1µ e
 ∂f 2  ∂f  2 ∂f 
2

g µe    ε
Mixture fraction variance − +   +   +    − C g 2ρg
σg σg  ∂x   ∂y   ∂z   k
 
µe
Turbulent energy k G − ρ ε
σk
µe  ε 
Dissipation rate ε    (c1G − c2ρε)
σε k
Where
2 2 2
   2 2
 2

  ∂u   ∂v   ∂w    ∂u ∂v   ∂u ∂w   ∂v ∂w  
G = µ e 2   +   + 
   ∂y   ∂z    ∂y ∂x   ∂x ∂x   ∂x ∂y  
 + +  + +  + + 
  ∂x  
Examining the θ-momentum equation (see Table 9.3) fifth, and sixth terms represent the corresponding diffu-
helps illustrate the meaning of each term: sion terms. When the turbulence model solves for the indi-
vidual Reynolds stresses, the diffusion terms do not strictly
∂(ρuw
  ) ∂(rρvw
  ww
 ) ∂(ρ  ) ∂  ∂w

r + + − r  µe  represent molecular diffusion, but also include momen-
∂x ∂r ∂θ ∂x  ∂x  tum contributions due to the turbulent motion of the fluid.
∂  ∂w
  ∂  µ e ∂w
 ∂p ∂  µ ∂u
 The first term on the RHS of Equation 9.21 represents the
−  rµ e  −   = − +r  e  pressure force on the volume element. All other terms on
∂r ∂r ∂θ r ∂θ ∂θ ∂x r ∂θ 

the RHS of the equation represent either a source or sink
∂  ∂v ∂   µ e   ∂w
 ∂v   term for momentum (gravity force, centripetal forces, etc.).
+  µe   +
− µew + 2r  
∂r  ∂θ  ∂θ  r   ∂θ ∂r  
9.2.4.2 k−ε Turbulence Model Boundary Conditions
 ∂w
 ∂v w

+ µe  + −  −ρvw
  + rρ
 θ (9.21)

 ∂r r ∂θ r  Establishing boundary conditions for the k−ε tur-
bulence model results from an order of magnitude
The first three terms of Equation 9.21 represent the net analysis of the boundary layer momentum equation
rate of momentum addition to a volume element by con- in the log layer. The log layer is an overlap region
vection from the three direction components. The fourth, between the viscous sublayer and the defect layer,
TABLE 9.3
Cylindrical Differential Equation Set
∂(ρu
φ ) ∂( rρvφ) ∂(ρw
 φ) ∂  ∂( φ )  ∂  ∂( φ )  ∂  Γφ ∂( φ ) 
r + + −r  Γφ  −  rΓφ  −   = Sφ
∂x ∂r ∂θ ∂x ∂x ∂r ∂r ∂θ  r ∂θ
Equation φ Γφ Sφ
Continuity 1 0 0
∂p ∂  ∂u
 ∂  ∂v  ∂  ∂w
 2
X-momentum u
 μe −r +r  µ e +  rµ e +  µe  + rρg x − rρk
∂x ∂x ∂x  ∂r ∂x  ∂θ  ∂x  3
∂p ∂  ∂u
 ∂  ∂v  ∂  ∂w
 w
 2 µ e ∂w

Y-momentum v μe −r +r  µe  +  rµ e  +  µe −  −
∂r ∂x  ∂r  ∂r  ∂r  ∂θ  ∂r r r ∂θ
2vµ e 2 2 
− + ρw
 + rρg r − ρk
r 3
∂p ∂  µe ∂u  ∂  ∂v 
Z-momentum w
 μe − +r  +  µ e − µew

∂θ ∂x r ∂θ  ∂r ∂θ 
∂  µ e   ∂w   ∂w 1 ∂w
 w
 2
− ρvw
 +   − 2v  + µ e  + −  + rρgθ − rρk
∂θ  r   ∂θ   ∂r r ∂θ r 3
f µe
Mixture fraction 0
σf
µe Cg 1µ e r  ∂f  2  ∂f  2  1 ∂f  2

ε
g +   +   + 
Mixture fraction −   − Cg 2rρg 
variance σg σg  ∂x   ∂r   r ∂θ 

k
Turbulent energy k µe r(G − ρ ε )

σk
µe  ε 
Dissipation rate ε  r   (c1G − c2ρε)
σε k
Where
  ∂u 2 ∂v 
2
∂w v 
2
  ∂u ∂v  2  1 ∂u ∂w  2  1 ∂v ∂w w
2

   1   
G = µ e 2   +   +  +   +  +  +  +  +  + −  
 ∂x   ∂r  r r    ∂r ∂x  r ∂θ ∂x  r ∂r r  
2
 
∂θ ∂θ
 
where the law of the wall applies. An order of magni- Standard wall functions for ε:
tude analysis of the boundary layer momentum equa-
tion suggests that convection, pressure gradient, and
molecular diffusion terms can be neglected. Applying k 3/2Cµ3/4
ε= (9.23)
these simplifications to the transport equations for k κy
and ε (9.18 and 9.19) results in relations for k and ε
at the boundaries, which are referred to as wall func- where
tions. The standard wall functions are given by the fol- uτ is the friction velocity
lowing two equations. τw
uτ =
Standard wall functions for k: ρ
uτ2 τw is the wall shear stress

k= (9.22) κ is the Kármán constant (κ ≈ 0.4)
Cµ
y is the distance measured normal from the wall
The wall functions (9.22 and 9.23) are applied to the node the computational demands are far in excess of current
adjacent to the wall, which should lie within the log layer computational resources as an applicable tool for furnace
region. The law of the wall Equation 9.24 is used to find simulations.23,24 DNS looks at solving the Navier–Stokes
the friction velocity (uτ) at the node adjacent to the wall, equations over the complete range of time and length
and then the wall functions are used to specify k and ε. scales present in a particular system. Perhaps the greatest
issue with DNS is that the computational domain must
 1  u y  be big enough to accommodate the largest length scales
U = uτ  ln  τ  + B1  (9.24) present, with a grid small enough to allow for resolution
κ  υ  
of the smallest scales, where kinetic energy dissipation
where occurs. This criterion insures that the number of grid
U is the mean velocity points required to solve a problem of practical size will
υ is the kinematic viscosity be quite large as compared to more conventional RANS
B1 is constant type calculations. By way of example, the number of grid
points required in each direction of a three-dimensional
If the surface is smooth, Equation 9.24 should be calculation is proportional to Re¾, the Reynolds number
used, however if the surface is not smooth, an equiva- based upon the magnitude of the velocity fluctuations
lent equation to 9.24 is given by Equation 9.25, and was and the integral scale. Hence, for Re = 106 (which is not
developed by Nikuradse. Equations 9.24 and 9.25 are unusual for flows in typical engineering systems), the
given in most engineering fluid mechanics texts, such resulting grid requirement for a three-dimensional simu-
as Roberson and Crowe20: lation would be on the order of 1013. While the results of
such calculations would contain very useful information,
1  y   the current limits of available computer hardware make
U = uτ  ln   + B2  (9.25)
 κ  k s   routine use of this approach to solve practical engineer-
ing problems prohibitive. In addition to the large grid
where requirements, the time scales required for this approach
ks is the surface roughness height is constrained to very small time steps to capture the fluc-
B2 is the constant tuations at the very small end of the range, which further
increases the computational load. Currently, DNS is used
Wilcox2 notes that numerical solutions are sensitive to to solve small, low-Reynolds number flow circumstances
the location of the node adjacent to the surface and recom- based upon simple geometries.
mends using near-wall grids. Another potential problem Large-eddy simulation is positioned between DNS
with using wall functions is that theoretically they do not and the RANS approaches. This approach separates the
apply for flows that separate from the wall. Finally, Wilcox range of length scales (eddies) into two groups. The first
points out that the standard k−ε model with standard wall consists of relatively large eddies that can be numerically
functions does not perform well for boundary layers with simulated and represent the majority of the turbulent
adverse pressure gradients. Kim and Choudhury21 have energy. These are generally constrained more by bound-
proposed modifications to the standard k−ε wall func- ary conditions and domain geometries. The second group
tions to account for adverse pressure gradient. comprises small eddies, which are more influenced by
molecular viscous forces. The second group tends to be
more isotropic and, therefore, it is easier to justify the use
9.2.5 Other Turbulence Modeling Approaches
of numerical approximations to model them. Hence, the
The vorticity that characterizes turbulent flow has a large simulation rationale focuses upon the direct resolution
range of length and time scales that can exist in rela- of the large eddies and filtering out eddies smaller than
tively close proximity to one another. When Reynolds- some lower limit or smaller than the local grid resolu-
averaged Navier–Stokes (RANS) equations are utilized tion and allowing them to be modeled. A significant
to simulate turbulent flow, all of the turbulent fluctua- issue related to the utilization of the LES approach is the
tions are modeled using algebraic approximations rather “communication” that must occur between the simula-
than being directly calculated, which results in a signifi- tion of the two groups as large eddies degrade to smaller
cant reduction in computational requirements. Various ones. The impact this has on grid requirements is that
other approaches exist that are used to compute the tur- much coarser grids than are required by DNS can be
bulent relationships that exist in flowing systems. Three used, although the grid requirements for LES are still
of the more popular methodologies are known as direct much greater than for the Reynolds-/Favre-averaged
numerical simulation (DNS), large-eddy simulation approaches. The recent improvements in computer hard-
(LES),22 and detached eddy simulation (DES). DNS can be ware performance have allowed this approach to be con-
categorically neglected for this class of problems because sidered for engineering analysis.
When the laminar flame time is less than the

Kolmogorov scale (i.e., Ka < 1), the flame is considered
9.3 CFD-Based Combustion Submodels
to be a laminar flame stretched by a turbulent flow.
This section discusses the modeling of combustion Combustion in this regime is referred to as flame-
chemistry in petrochemical applications. The focus of let combustion. When the DamkÖhler number is less
this section is on methods for modeling the interaction than one, the time scale of larger turbulent eddies has
of turbulence with combustion chemistry. This is an area become smaller than the chemical time scale. In these
of intense current research, and some of this research is conditions, the combustion process is described as a
briefly discussed as it pertains to current CFD calcula- well-stirred reaction zone. For intermediate values of Da
tions as well as near-future CFD calculations. There are and Ka, combustion is said to occur in distributed reac-
several relatively new turbulence/chemistry interaction zones. This term indicates that the turbulent flow
tion models (such as CMC and joint-probability density can affect the structure of the reaction zone, in contrast
function [pdf] transport models), which are not cur- to the flamelet regime, but the turbulent mixing is not so
rently available for use in any of the commercial CFD fast that the reaction can be considered to occur under
packages. It is hoped this situation will change soon and well-stirred conditions.
these models will be available for more widespread use. Using the relationships for the DamkÖhler number as
described earlier, two physical limits have been identi-
9.3.1 Regimes of Turbulent Combustion fied.26 The first, referred to as the “Frozen” limit occurs
when
DamkÖhler numbers are ratios of a fluid dynamical time
scale to a chemical time scale.25 In a turbulent flow, there Da → 0 (9.28)
are a variety of time scales, such as the integral scale (a con-
vective scale) and the Kolmogorov scale (a viscous scale).
There are also a variety of chemical time scales because of In this case the reaction time (tL) is much larger than
the many chemical reactions that accompany the combus- the flow time (tT), and kinetic effects are negligible com-
tion of even a simple molecule such as CH4. Frequently, pared to mixing effects.
combustion problems are described as being in the high Conversely, the second limit, referred to as the “Fast-
DamkÖhler or flamelet regime. The term flamelet is used Chemistry” limit occurs when
because of the notion that within a turbulent nonpre-
mixed flame, the actual combustion reactions take place Da → ∞ (9.29)
within small layers termed flamelets. These flamelets
are so small that they are not affected by the turbulent
In this case, the reaction time is very short (fast reac-
motions within the fluid, instead molecular diffusion
tions) relative to the mixing time. Many diffusion flames
effects dominate and the structure of the reaction zone is
in environments that shift the chemical equilibrium to
that of a laminar flame (albeit a strained laminar flame).
the product side are approximated well by the latter
Following Bray,23 the DamkÖhler number is defined as
limit. For high temperature combustion chemistry, it
can be assumed that the homogeneous reaction kinet-
tT kuL0
Da = = 0 (9.26) ics are sufficiently fast so that gas mixing is controlling
tL0 εlL factor for the conversion of reactants to the products.
This is commonly referred to as the “mixed-is-burnt”
assumption.
where
Given the large disparity between short reaction
L and 0 refer to an unstretched laminar flame
time scales and long mixing time scales, chemi-
T refers to the scale of the turbulence
cal activity may be confined to an infinitesimally
thin layer commonly referred to as a “flamelet” or
In cases where nonpremixed combustion is studied,
“flamesheet.”27 This assumption allows flame chem-
it is common to use the velocity and length scales (the
istry to be approximated using local thermodynamic
laminar premixed flame speed and thickness) as repre-
equilibrium without significant error. This assump-
sentative of the relevant chemical scales. The Karlovitz
tion is not valid for reactions that are considered slow
number is
as, for example, the formation of thermal NOx. In these
cases the reaction time scale and the mixing time scale
tL0 tL
Ka = = (9.27) are of similar magnitude:
tκ ν/ε

Da ≈ O(1) (9.30)
where the subscript κ refers to the Kolmogorov time scale.
In order to get a satisfactory solution, the finite-rate most commercial CFD codes only include limited
chemistry must be coupled with the turbulent fluid support for premixed flame modeling.
mechanics calculations. Because turbulent effects must
be included in the kinetic scheme, global mechanisms
9.3.2 Reaction Kinetics
are generally used to avoid solving individual trans-
port equations for each species in a detailed kinetics The general form for a reaction with educts A and B and
mechanism. the products C and D can be written as
In general, a differentiation has to be made for types
of flames as there are ν′A A + ν′BB ↔ ν′′CC + ν′′DD (9.31)
• Premixed where νi is the stoichiometric coefficient.

• Partially premixed The reaction rate can formally be written as the prod-
uct of the reaction rate coefficient and all participating
• Nonpremixed or diffusion flames species.
Depending on the Reynolds, DamkÖhler, and  dc 
ri =  i  = kc1a1c2a 2c3a 3  cNaN (9.32)
Karlovitz numbers, a further separation has to be made  dt 
in laminar, turbulent, wrinkled, corrugated flames,
and homogeneous reaction zones.28 Since almost every with the reaction coefficient k, the reaction order ai and
flow regime in a technical combustion application is the species ci, where the concentration of each species i
highly turbulent, the laminar flame modeling will not is defined as
be included in this discussion. Further information ρiYi
on laminar reaction systems can be found in Kee,29,30 ci = (9.33)
M
Glarborg,31 and Rogg.32–34
The type of turbulent flame to be modeled determines with M as the mean molar mass of the mixture. The rate
the combustion submodel to be used. coefficient k can be written in general form as
For the premixed flames the models available are
 E 
k = AT m e  − A  (9.34)
• Progress variable C  RT 
• Transport equation for the flame front G with R as the gas constant (8.3143 kJ/mole K). The
• Eddy break up (EBU), Eddy dissipation com- parameters here are the pre-exponential Arrhenius fac-
bustion model (EDX), and the eddy dissipation tor A, the temperature exponent m, and the activation
concept (EDC) energy EA.
The final formulation for a concentration change of
For nonpremixed combustion the models available are species i, over all elementary reactions k can be formu-
lated as
• Presumed pdf with equilibrium chemistry (adia-
batic and nonadiabatic) M N νi′,k
∂ci Ek   ρYj 
• Presumed pdf with finite rate chemistry
∂t
= ∑ ( ν′′i , k − ν′i , k )( AT )k exp  − RT
α
  ∏J =1
W 
 j
(9.35)
(flamelet-model) k =1
• EBU, EDX, and EDC or written as reaction rate
Turbulent premixed flames have proven to be much K

∂ci
more difficult to model than their nonpremixed
counterparts.35 In a turbulent, mixing-limited non-

∂t
Wi = Yi = Wi ∑ν
k =!
r
IK k
(9.36)
premixed flame, the flame structure is governed by

turbulent mixing, a reasonably well-understood phe- And finally the additional source term for the equa-
nomena. The ideal turbulent premixed flame consists tion for energy conservation. The total released energy
of a flame sheet propagating at some flame speed with can be written as the sum of the product of reaction
respect to the fluid around it, which is itself undergo- enthalpy H and the concentration change of species i.
ing turbulent motions. The consequence of superpos-
K K N
ing flame propagation and turbulent fluid motions  ∆H k   dci 
is that premixed flame modeling is much more chal- q= ∑ k =1
qk = ∑
k =1
− ⋅
 νik   dt  k
=− ∑hω
i =1
i i (9.37)
lenging than nonpremixed flames.24 For this reason,
These equations are generally valid for all reaction where

mechanisms, either for a model with a detailed reaction A is the pre-exponential factor
mechanism including hundreds of reactions or a simpli- EA the activation energy
fied global reaction as, for example, the conversion of R is the gas constant
methane:
The actual rate of fuel consumption can be determined
xCH 4 + 2xO 2 = xCO 2 + 2xH 2O (9.38) by
1
The difficulty at hand for the formulation of a global Y F , actual = (9.42)
(1/τYkin ) + (1/Yeddy )

reaction mechanism lies in the fact that the Arrhenius
parameters as described earlier become unavailable for
a global reaction, since a global reaction tries to describe The major drawback of this approach is the assump-
a process that includes a large number of elementary tion of homogeneous distribution and perfect mixing of
reactions and intermediate species. fuel and oxidant. The formulation would only be valid
However, there are well-documented sources36 for the for a turbulent, premixed flame with homogeneous dis-
formulation of Arrhenius parameters of global reactions tribution of fuel/air and flue gas pockets.
for the combustion of hydrocarbon fuels.
The implementation of a reaction mechanism into
a CFD model requires the formulation of a transport 9.3.4 Eddy Dissipation Combustion Model
equation of the mass fraction of any species involved in A development of the EBU-model is the eddy dissipation
the combustion model setup. combustion model (EDX) formulated by Magnussen.38
The equation governing the species transport is He takes into account the unmixedness of fuel and oxi-
dizer and postulates that the rate of combustion will be
∂ determined by the turbulent intermixing of fuel and
ρYi + ∇ ⋅ (ρuYi ) = ∇ ⋅ (ρDi∇Yi ) + Yi , (9.39)
∂t oxygen eddies on a molecular scale, or in other words
by dissipation of these eddies. The reaction rate can be
∙
with Yi the source term of the species i. The determina- written down in general form as
tion of that source term can be done according to the
aforementioned equations, or due to the complexity and
ε
number of involved species using approaches, which Yi = CYi (9.43)
reduce the computational effort. k
where
9.3.3 Eddy Breakup Model C is a constant depending on the structure of the flame
Yi is the time-averaged concentration
The eddy breakup model (EBU) was developed by
Spalding37 for the calculation of turbulent, premixed This approach acknowledges the distribution and
flame with irreversible one-step reaction. The assump- unmixedness of fuel oxidizer and hot products, and is
tion was that the rate of reaction is totally controlled by therefore valid also for the calculation of nonpremixed
dissipation of eddies either containing unburned fuel flames. The actual reaction rate is determined by the
or burned hot gases. The reaction rate of the unburned minimum of these criteria.
∙
mixture YF, as to be used in Equation 9.39 is determined The reaction rate is determined by the following:
by the decay rate of these eddies.
1. The concentration of fuel
du
Y F , eddy = C(1 − τ)ρ (9.40) ε
dy YF = AYF (9.44)
k
following the mixing length hypothesis. where A = 4.0 and YF is the time-averaged fuel
In order to estimate the actual rate of fuel consumption, concentration
we have to take into account the kinetically controlled
reaction rate as defined by 2. The reaction rate is determined by the concen-
tration of oxygen
 E  Y ε
Y F , kin = AYFYO 2 exp  − A  (9.41) YO2 = A O2 (9.45)
 RT  rO2 k
with YO2 the local mean oxygen concentration where

and rO2 the stoichiometric oxygen requirement. ν is the kinematic viscosity
3. The reaction rate is limited by the presence of ε is the rate of dissipation of turbulent energy
hot products
Assuming isotropic turbulence, the mass fraction of the
Y ε fine structure and the mass transfer between the fine
YPr = AB Pr (9.46) structure and the surrounding can be written as
1 + rO2 k
with A = 4.0, B = 0.5, YPr the local mean product 3/ 4

 νε 
concentration and rO2 the stoichiometric oxygen γ * = 9.7  2  (9.51)
k 
requirement.
and for the mass transfer
9.3.5 Eddy Dissipation Concept
1/ 4
νε ε
The disadvantage of the EDX approach is the fact that still  = 23.6  2 
m (9.52)
k  k
global reactions have to be used. An extension of these
models is the eddy dissipation concept EDC.39 This model
If we assume that the reactions inside of the fine struc-
is still based on the eddy dissipation along the turbulent
tures are infinitely fast, the mass transfer between the
energy cascade, but differentiates between areas of chem-
surrounding fluid and the fine structure itself limits
ical reaction, the so-called fine structure region and their
the reaction rate. So the rate of reaction can be written
surrounding nonreactive regions. These fine structures
depending on the mass transfer
are situated at the end of the turbulent energy cascade.
There length scale is that of eddies in Kolmogorov scale, 1/ 4
where the species are mixed on a molecular level and  νε  ε
YF = 23.6  2  Ymin (9.53)
react as the energy level meets their respective activation k  k
energy. The chemistry inside of these fine structures can
be treated as perfectly stirred reactor. where Ymin is the smallest concentration of YF or YO2, lim-
This model takes into account that the dissipation is iting the reaction rate. It cannot be assumed that all fine
not homogeneously distributed in the computational structures are on the same temperature level, so the con-
domain, but takes place mainly in strained areas, the centration of hot products has to be taken into account.
separating area from the reacting and nonreacting vol- A correction factor
umes. Magnussen proposes the following definition for
the mass fraction contained in these fine structures: YPr /(1 + rO2 )
χ= (9.54)
3
YPr/(1 + rO2 ) + YF
 u* 
γ =   (9.47) is introduced and the final expression for the reaction
 u′ 
rate is
where
1/ 4
u* is the characteristic velocity of the fine structure  νε  ε
YF = 23.6  2  χYmin (9.55)
u′ is the turbulent velocity k  k
Any mass transfer from surrounding nonreacting fluid The main advantage of this approach is the possibility
into these fine structures follows the expression to include elementary reactions to describe the kinetics
within the fine structures. Since any given set of reac-
u* *
m
 =2 γ (9.48) tions has only to be solved within these small regions,
L* the computational time can be reduced significantly.
where the characteristic velocity u* and the characteris-
tic length L* can be expressed with 9.3.6 Mixture Fraction Approach for
Equilibrium or Finite Rate Chemistry
u* = 1.74( νε)1/4 (9.49)
Another modeling approach is the pdf-approach with
and finite rate or equilibrium chemistry. The mixture fraction
concept plays a central role in reducing a turbulent non-
ν 3/ 4
L* = 1.43 1/4 (9.50) premixed flame to a mixing problem. The mixture frac-
ε tion is a conserved scalar, meaning that it is convected
and diffused by fluid motions and gradients, but it is nei- For the case of a stoichiometric mixture with νYF = YO2
ther created nor destroyed. The mixture fraction, Z repre- the stoichiometric mixture fraction Zst can be calculated
sents the mass fraction of fluid at a particular location that via
originated with the fuel stream. The pure fuel stream then
−1
will have Z = 1, while the oxidant stream will have Z = 0. YO2 , 2  νY 
In a turbulent flow, the mixture fraction Z fluctuates Zst = = 1 + F ,1  (9.60)
νYF ,1 − YO2 , 2  YO2 , 2 
at a given point with time. A pdf for these fluctuations
can be defined so that the probability of Z lying between
In order to include this approach into a CFD model,
some value x and x + dx is P(x)dx. This approach is often
two additional equations have to be solved. The first one
applied to turbulent flames using a presumed pdf for
is the conservation equation for the mixture fraction Z,
the mixture fraction. The schematics for an adiabatic
the second one is the conservation equation for the mix-
calculation are shown in Refs. [40,41].
ture fraction variance Z″.
The mixture fraction Z is defined as an element mass
The transport equation for the mixture fraction Z as it
fraction of matter originating from the fuel stream. It
is represented in a flow simulation
is a so-called conserved scalar, because elements are
conserved during combustion and it is not influencing
∂(ρuZ) ∂(ρvZ) ∂(ρwZ)
directly the underlying fluid mechanics. The element + +
∂x ∂y ∂Z
mass fraction of an element i in an N-species mixture is
N  µ   ∂ 2Z ∂ 2Z ∂ 2Z 
M atom =  Df + t   2 + 2 + 2 
Zi = ∑
j =1
a jiYj i
Mj
(9.56)  Sc f   ∂x ∂y ∂z 
(9.61)
and the transport equation for the mixture fraction vari-

where
ance Z″
aij is the number of atoms i in species j
Mj the molecular weight of species j
Miatom the atomic mass of atom i ∂(ρuZ′′ ) ∂(ρvZ′′ ) ∂(ρwZ′′ )
+ +
∂x ∂y ∂z
The mixture fraction Z is defined as linear combination
of Zi, where it is 0 in the oxidizer stream and 1 in the  µ   ∂ 2Z′′ ∂ 2Z′′ ∂ 2Z′′ 
−  Dg + t   + +
fuel. In a two-feed system with a fuel stream mass flow  Sc g   ∂x 2 ∂y 2 ∂z 2 
m∙ 1 and an oxidizer stream mass flow m∙ 2, the mixture
fraction represents the mass fraction of the fuel stream µt  ∂Z ∂Z ∂Z  ε
locally in the unburned mixture =2
Sc g  ∂x + ∂y + ∂z  − 2ρ k g (9.62)
m1 Z Z
Z= = F = 1 − O (9.57)
m1 + m2 YF ,1
  YO2 , 2 The pdf integration is commonly done by a presumed
pdf, using the β-pdf. Other pdf shapes such as a clipped
where Gaussian function and a double delta function are dis-
Yf,1 is the mass fraction of fuel in the fuel stream cussed in Jones and Whitelaw.19 The equilibrium chem-
YO2 , 2 is the mass fraction of oxygen in the oxidizer istry assumption is poor in flames that are lifted or
stream flames near extinction.
ZF and ZO are the fuel element mass fraction coming The β-function is defined as
from the fuel and, respectively, coming from the
oxidizer stream Γ(α )Γ(χ)
Beta(α , χ) = , (9.63)
Γ (α + χ )
The fuel element mass fraction ZF is equivalent to the
sum of fuel atoms. In the case of hydrocarbon-combus- where
tion it is equal to
ZF = ZC + ZH (9.58)  Z(1 − Z) 
α = Z − 1 (9.64)
 Z′′ 2 
The mixture fraction in the presence of combustion and
can be written as
νYF − YO2 + YO2 , 2 (9.59)  Z(1 − Z) 
Z= β = (1 − Z)  − 1 (9.65)
νYF ,1 − YO2 , , 2  Z′′ 2 
method: its assumption that the reaction layers are

bounded by pure fuel on one side and pure oxidizer on
the other. Their strained dissipation and reaction layer
(SDRL) model is based on the one-dimensionality of the
Z = 0.1, Z΄2 = 0.1 reaction layer, but does not assume the reaction layer to
be thin relative to the dissipative scales.
p(Z)
The classical flamelet model’s assumption that the

Z = 0.3, Z΄2 = 0.1
reaction zones are bounded by pure fuel on one side and
pure oxidizer on the other is severe in light of the NOx
control strategies used in practical combustion systems.
Z = 0.3, Z΄2 = 0.4 NOx control is predicated on entraining cooled combus-
tion products into the reaction zone, and the proportion
of these gases entrained varies along the length of the
0 0.2 0.4 0.6 0.8 1 flame. The effect of this flue gas entrainment is to reduce
Z flame temperatures and dilute the reactants. Both of
these effects are effective at reducing NOx formation.
FIGURE 9.7 Research of models of nonpremixed combustion
Plot of the β-function for several values of Z and Z″. (From Baukal, continues at a fervent pace. Pope’s35 joint pdf methods
C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton,
FL, 2001.)
appear promising because they have the ability to treat
finite rate kinetics and eliminate the closure problems.
Bilger’s46 conditional moment closure (CMC) method
with also is a promising model for nonpremixed combus-
∞ tion modeling. Both of these models are applicable to

∫
Γ(φ) = t φ −1e − t dt (9.66)
0
premixed combustion as well.23 These models are still
subjects of active research and academic debate. It does
not appear that these models have been implemented in
The shape of the β-function is shown in Figure 9.7. This any of the commercial CFD packages at this time.
graph shows the distribution for some values of mixture
fraction and several values for the mixture fraction vari-
ance Z″. 9.3.7 Pollutant Chemistry Models
The mass fraction of any species and the temperature Pollutant emissions are among the most important driv-
distribution after integration over ß-function can be ers for the development of an improved burner design or
determined with furnace layout. The permitted emission levels of NOx and
SOx from petrochemical plants and refineries continue to

∫
(Z , Z′′ ) = Y(Z)P(Z , Z′′ )dz (9.67)
Y decrease. To respond to this challenge, burner manufac-
turers strive to develop burners that produce lower and
and
lower emissions. In addition, furnace manufacturers and
∫
T(Z , Z′′ ) = T (Z)P(Z , Z '')dz (9.68) other vendors develop postcombustion technologies such

as selective catalytic reduction (SCR) and selective noncat-
alytic reduction (SNCR) to reduce NOx in the stack. Sulfur
A graph of the calculated temperatures of a meth- scrubbers are used to reduce SOx levels in stack gases.
ane-air flame for different variances Z″ can be found Two different models for the turbulence/chemistry
elsewhere.42 interaction are proposed by Pope (see Pope,35 for exam-
An alternative to the equilibrium chemistry discus- ple) and Bilger.46 The pdf model described in this paper
sion in the previous paragraph is the laminar flamelet is not the assumed-pdf discussed at length earlier. This
model. In this model, the relationship between the state method solves for the transport and production of
of the mixture and the mixture fraction f is determined the scalar joint pdf and is extremely computationally
by a laminar diffusion flame calculation. Peters43 intro- expensive since a Monte Carlo solution algorithm must
duced this idea, which assumes that the reaction length be used. There is a particularly interesting comment
scale, LR, is much smaller than the Kolmogorov length in the introduction of that article that says, “a realistic
scale, LK. Bilger44 has criticized the classical flamelet target for agreement between experiment and predic-
method, claiming that for most nonpremixed flames of tion might be ±20% ± 30%.” The flame studied in this
interest, the flamelet criterion, LR < LK, is violated. Bish paper is a simple diluted hydrogen jet flame. If the most
and Dahm45 discuss the concept further and attempt sophisticated turbulence/chemistry models currently
to eliminate what they view as a key limitation of the under research applied to a very simple flame in a very
simple geometry can only be expected to yield an accu- to the major methods and especially those having close
racy of ±30%, then how accurately can NOx emissions relationships with CFD simulations in petrochemical
be predicted? furnaces.
9.4.1 P-1 Radiation Model

The P-1 radiation model is basically the general P-N
approximation62 when N is set to one, which is the low-
9.4 Radiation Models
est-order form of P-N formulation. Although there are
Radiative transport from a nongray gas (the products higher-order applications like P-361 in relatively simple
of combustion) to the furnace walls and tubes (with the geometries, P-1 is the most popular method since it
process fluid flowing inside) whose emissivity is tem- reduces the RTE equation to a relatively simple partial dif-
perature dependent is the primary mode of heat trans- ferential equation by expressing radiative intensity with
fer. The interaction between these physical processes is generalized Fourier series and spherical harmonics.65
of considerable importance. The turbulence/chemistry The P-1 model may take into account scattering effects
interaction has been well studied for many years, par- and could produce reasonable results with optically
ticularly for nonpremixed systems. More recently, the thick media. Fortunately, most combustion applications
interaction between turbulence and radiative emission are in this optically thick range. But the P-1 method may
from a nongray gas with properties has been studied.47 have significant inaccuracy in optically thin media and
Typical petrochemical furnaces consist of a radiant at the same time with strongly anisotropic characteris-
section and a convection section. These regions are so tics, especially when surface emission is much stronger
named because of the dominant mode of heat transfer. In than media emission.
the radiant section, refractory surface temperatures can For example, it was pointed out by Sazhin et al.58 that
be higher than 2200°F (1200°C). Radiant heat is incident the P-1 model overestimates the radiative heat fluxes
on the process tubes both from the high temperature from localized heat sources in optically thin media.
surfaces and directly from the flame. Accurate model- On the other hand, they also concluded that P-1 model
ing of the heat delivered to the process fluid requires is capable of accounting for the radiative exchange
an accurate prediction of the radiant intensity inside the between gas and particles. The results of coal combus-
furnace. In addition, accurate prediction of radiation tion simulation in an industrial furnace using the P-1
from the flame is necessary to accurately predict emis- model are shown to be in reasonable agreement with
sions. For example, Barlow48 notes that the different the experimental observations.
radiation models can affect NOx predictions as much as
the different turbulence/chemistry interaction models
9.4.2 Discrete Ordinates Radiation Model
evaluated in the paper.
Thermal radiation transport presents a difficult prob- The discrete ordinates (DO) method, also called
lem because of the number of independent variables. SN-method was first proposed and developed by
The radiation transport equation (RTE) describes radia- Chandrasekhar50,51 as an analytic approach in his
tion transport in an absorbing, emitting, and scattering research on atmospheric radiative transfer. The method
media. The equation is (notation of Modest49). was then extensively investigated and applied to neu-
tron transport field by Lathrop etc.52 With the fast devel-
∂I λ σ   
∂s
= κ λ I bλ − β λ I λ + sλ
4π ∫ I (s )Φ(s , s )dΩ
4π
η i i i (9.69) opment of petroleum and power generation industries,
accurate calculation of radiative heat transfer in emit-
ting, absorbing, and scattering media became more and
Radiant intensity is a function of location, (three coor- more important. This led to early studies and applica-
dinates in a three-dimensional problem) direction, (two tions of the discrete ordinates method in radiative heat
angular independent variables), and wavelength (one transfer problems during 1970s to 1980s, mainly by the
independent variable varying from 0 to infinity). This works of Fiveland et al.53–57 as well as Truelove and his
means that the problem of radiative transport is a six- coworkers.58,59
dimensional problem. A common approach is to remove The discrete ordinates method transforms the RTE
the wavelength dependence by making gray media equation into a set of simultaneous partial differential
approximations (see Chapter 7). equations; Modest provided detailed evolution of the
There are many solution methods in solving the gen- theory and extensive discussions in his book. To sim-
eral RTE equation. Here, we provide a brief introduction plify the problem, consider a gray medium where all
equations are valid with spectral dependent cases as Please note that the heat flux is a vector with magni-
well. Equation 9.69 then takes the following form: tude and direction. In Cartesian coordinates, the three
 components are expressed as
dI ( r , s ) 
= k(r )I b (r ) − β(r )I ( r , s )
ds   
σ(r )   
qx ( x , y , z ) =
∫ I(x, y, z, s ) ⋅ cos(s , x) ⋅ dΩ
∫ I (r , s′) Φ (r , s′, s ) dΩ′
4π
+ (9.70)
4π n
∑ w I (x, y, z) ⋅ cos(s , x)
4π  
≅ i i i (9.75)
where i =1
the radiative intensity I is a function of r and direc-
  
tion vector →s
r represents the position including three space coor-
qy ( x , y , z ) =
∫ I(x, y, z, s ) ⋅ cos(s , y) ⋅ dΩ
4π
dinates (x, y, z)
→ n
s is the unit vector of the intensity direction including
∑ w I (x, y, z) ⋅ cos(s , y)
 
two direction coordinates (θ, φ) in 4π solid angle ≅ i i i (9.76)
i =1
β is the extinction coefficient
σ is the scattering albedo of the medium   
Φ is the scattering phase function qz ( x , y , z ) =
∫ I(x, y, z, s ) ⋅ cos(s , z) ⋅ dΩ
4π
Using the discrete ordinates method, Equation 9.70 n
∑ w I (x, y , z) ⋅ cos(s , z)
will be solved along a number of specified directions  
→ ≅ i i i (9.77)
s i (i = 1, 2, 3, …, n) and the whole 4π solid angle integral i =1

term will be replaced by numerical quadrature.
This results in the following equations instead of Usually, the more directions that are chosen or the
Equation 9.70: higher the order of discrete ordinates, the more accurate
 the results will be, but at a much higher cost of com-
dI ( r , si )  putational resources and slower convergence speed.
= k(r )I b (r ) − β(r )I ( r , si ) Several different sets of directions and weights based
dsi
on different principles can be found in the literature.60,61
n
σ(r ) Generally, there are three criteria of the 0th, first, and
∑ w I (r , s ) Φ (r , s , s ),
  
+ j j i j i = 1, 2, 3, … , n second moments to satisfy. Table 9.4 showing S2, S4, and
4π j =1
(9.71) S6 approximations is reproduced from Modest.62
The beauty of the discrete ordinates method is that it
where, wj in this equation is the quadrature weight. The can be carried out to any higher order and accuracy. But as
direction →
s i in 4π solid angles could be expressed as a consequence of the spatial discretization technique, sim-
          ilar to the “numerical diffusion” in CFD calculations, the
si = cos( si , x ) ⋅ x + cos( si , y ) ⋅ y + cos( si , z ) ⋅ z discrete ordinates method may cause “false scattering”
   leading to physically unrealistic solutions. Also, as the
= ξ i ⋅ x + ηi ⋅ y + µ i ⋅ z (9.72) consequence of angular discretization approach, discrete
ordinates can have the so-called “ray effect.” Between two
Here x→, y→, and →
z ; are principle unit vectors. Once the ordinate directions, there may be some control volumes
intensities in specified ordinate directions in 4π solid that do not receive any energy from distant emission
angle are obtained, the radiative heat flux and incident sources. Both false scattering and the ray effect may lead
radiation could be calculated by their definition formu- to physically impossible results. By refining the control
las as follows: volume mesh and increasing discrete ordinate directions
(higher order), both numerical errors can be reduced.
n
∑ w I ( r )s
    So far, a brief introduction to the standard form of
q( r ) =
∫
4π
I (r , s )sdΩ ≅
i =1
i i i (9.73) discrete ordinates method has been presented. This
standard form is hard to implement into fluid dynamics
n flow and heat transfer computations. One of several rea-
∑ w I (r )

G(r ) =
∫
4π
I (r , s )dΩ ≅
i =1
i i (9.74) sons is that it is difficult to apply to irregular geometries,
because it is hard to find a set of directions and weights
TABLE 9.4
Discrete Ordinates for the SN-Approximation (N = 2, 4, and 6)
Ordinates Weights
Order of
Approximation ξ η μ W
S2 (symmetric) 0.5773503 0.5773503 0.5773503 1.5707963
S2 (nonsymmetrical) 0.5000000 0.7071068 0.5000000 1.5707963
S4 0.2958759 0.2958759 0.9082483 0.5235987
0.2958759 0.9082483 0.2958759 0.5235987
0.9082483 0.2958759 0.2958759 0.5235987
S6 0.1838670 0.1838670 0.9656013 0.1609517
0.1838670 0.6950514 0.6950514 0.3626469
0.1838670 0.9656013 0.1838670 0.1609517
0.6950514 0.1838670 0.6950514 0.3626469
0.6950514 0.6950514 0.1838670 0.3626469
0.9656013 0.1838670 0.1838670 0.1609517
Source: Modest, M.F., Radiative Heat Transfer, McGraw-Hill, New York, 1993.
to calculate heat fluxes at irregular wall boundaries the solution effort increases much less rapidly than with
without errors. In order to overcome those difficulties, it conventional techniques. The Monte Carlo method may
is logical for people to use the finite volume method, or be the only method that can satisfactorily deal with the
a modified discrete ordinates method in radiation cal- effects of irregular radiative properties (nonideal direc-
culation for reacting flow simulations. The finite volume tional and/or nongray behavior). The disadvantage of
method is easier to implement into the control volume the Monte Carlo method is the unavoidable statistical
CFD scheme and combined heat transfer problems. error it may have. The Monte Carlo method can also
The finite volume method actually calculates the solid be computationally intensive depending on the size
angle integral term by exact integration through a fully of samples and accuracy requirement. Many research-
finite volume approach in direction. By doing so, the ers are trying to improve the efficiency of Monte Carlo
energy is conservative. The directional domain of 4π is programs.59,64
divided into n solid angles. Assuming radiative inten- The Monte Carlo method takes a much different
sity in the finite solid angle is a constant or defined by approach from other methods. In the Monte Carlo
simple formulas. Numerical errors can be incurred due method, individual photons of radiant energy are emit-
to solid angle overhang; however, this can be improved ted, reflected, and absorbed by both solid surfaces and
by pixilation. participating media using ray tracing algorithms. This
method provides a very elegant approach to treating
nongray radiation as well as the directional dependence
9.4.3 Monte Carlo Method
of radiation. Its use is limited by its computational cost
Monte Carlo is a statistical method, which is described if high accuracy is required.
as solving a mathematic problem through a statistical Siegel and Howell65 and Modest62 provide exten-
sampling technique. When dealing with radiative heat sive discussion of the solution methods for radiation
transfer, the Monte Carlo method means solving ther- in participating media. These texts discuss the accu-
mal radiation problems by tracing the history of bun- racy, computational effort, and limitations of the vari-
dles of photons (or rays) from their points of emission ous models. The reader should consult these books for
to their points of absorption. Depending on how a ray more discussion on these solution methods.
is traced, the Monte Carlo ray tracing method has two
popular algorithms: forward Monte Carlo ray tracing
9.4.4 Gas-Radiation Properties
(FMCRT) and reverse (or backward) Monte Carlo ray
tracing (RMCRT). Both FMCRT and RMCRT trace a ray Molecular gas radiation is an important mode of heat
in the same way, but in FMCRT, the tracing is on how transfer in gas-fired furnaces (see Chapter 7). Radiative
much energy is lost along the path of the ray; while in emission from nonluminous hydrocarbon flames is
RMCRT, the tracing is on how much intensity contrib- mostly due to the H2O and CO2 species present in the
utes to the point of interest from the reversed paths.63 products of combustion. Radiation from these gases is
One of the advantages of the Monte Carlo method is fairly well understood, but a rigorous treatment of this
that it can solve even the most complicated problem with radiation requires significant computational resources.
relative ease. As the complexity of the problem increases, For instance, Mazumder and Modest47 considered 10
radiative bands in modeling emission from a hydro-

carbon flame. This means that they solved the RTE for
10 different intensities. In a large-scale furnace calcula-
tion, such a model would be extremely computationally
demanding.
Quantum mechanics postulates that molecular gases
emit and absorb gases only at distinct wave numbers,
called spectral lines. However, in reality these distinct
lines are broadened by several mechanisms includ-
ing collision broadening, natural line broadening, and
Doppler broadening. These individual lines are charac-
terized by a line strength and a line width. These lines
are caused by quantum transitions in the vibrational or
rotational state of a molecule. Frequently, vibrational
and rotational transitions occur simultaneously, leading
to a tightly clustered array of lines around a given vibra-
tional transition. This subject is beyond the scope of the
present chapter. The intent is to illustrate the complexity
of modeling a radiating gas.
9.4.5 Weighted Sum of Gray Gases

The weighted-sum-of-gray-gases model65 provides for- FIGURE 9.8
mulas for computing the emissivity of a gas volume as a Representation of a luminous flame utilizing a soot model.
function of its temperature and partial pressures of CO2
and H2O. The model assumes the gas is a mixture of than the gas radiation emitted by the flame.49 In some
radiating gases that is transparent between the absorp- applications, (oil-firing, in particular) soot emissions
tion bands. The weighted-sum-of-gray-gases model from the flame are regulated by environmental agencies.
is probably the most widely used method to calculate In flaring applications, smokeless (smoke results from
radiation within combustion gases. The computational unoxidized soot particles leaving the flame) operation
cost of radiation transport can be very high compared is frequently guaranteed by the flare vendor for some
to the flow solver portion of a simulation because of the range of conditions. In petrochemical applications, the
large number of independent variables in the RTE. In gases flared are a wide range of hydrocarbons, typically
practice, it is usually reasonable to lag the calculation ranging in molecular weight from 16 to 40. These gases
of the RTE for a number of flow solver iterations, with have components such as ethylene and acetylene, which
the actual number dependent on the solver in use and are known precursors to soot formation. However, the
stability requirements. calculated soot concentration from those models could
be used for the qualitative representation of a luminous
flame, either in a combustion furnace, or as shown in
9.4.6 Effect of Soot on Thermal Radiation
Figure 9.8 for a flare burner under ambient conditions.
The presence of soot in a flame can significantly increase Current CFD codes (limited by physical model avail-
the flame emissivity. Predicting soot formation within a ability) cannot predict smoking from these large, buoy-
flame is very difficult because soot is formed in fuel-rich ant flare fires, but current LES work in this area appears
regions of a flame when the temperature is high. Models promising.
such as Khan and Greeves66,67 allow the prediction of
soot concentrations, but these models are very empirical
and cannot be expected to provide accurate quantitative
results.
Soot within a flame is caused by the combustion of
hydrocarbons under fuel-rich conditions. Soot is visu-
9.5 Solution Methodolgy
ally observed as a yellow-red brightness in the flame. There are many schemes used to discretize the par-
C2 hydrocarbons and higher have more tendency to tial differential equations of fluid flows onto different
soot, while methane does not normally produce a sooty types of meshes. Since the primary focus of this chapter
flame. Soot has a strong impact on flame radiation. is applied CFD where mesh types by necessity include
Emission from soot in flames is frequently much larger tetrahedral cells, the two important discretization
schemes are the finite volume method and the finite ele-
ment method. The finite volume method is clearly the
method of choice in the industry today for large-scale
computations of turbulent flows. The dominant soft-
ware products commercially available for these prob-
lems almost exclusively use the finite volume method.
There are occasions when other methods, such as the
finite difference method, are used.
9.5.1 Problem Setup: Preprocessing

The preprocessing phase of a problem includes all the
steps from the initial problem definition through the
beginning of computations. In typical problems, this
includes geometry creation, mesh generation, model
selection, fluid property specification, and enabling and
setting up the appropriate submodels. Z
At this stage, the scope of the geometry to be studied
X
should be considered. In many cases, it is difficult to deter-
Y
mine where to place the outside boundaries of a CFD
model. Flow conditions must be known at all the inlets.
So, for instance, putting a flow inlet just downstream of an
elbow would probably be a poor choice because it would
be difficult to know the velocity and pressure profiles at FIGURE 9.9
such a location. This issue is particularly important if heat Discretized geometry of a typical burner.
transfer has to be considered. Thermal boundary condi-
tions are typically difficult to specify, requiring consider- treat complex geometries to accurately model the per-
able physical insight into a problem. formance of the modeled burner.
It is also important to consider the capabilities of the A simplified burner model may be used to reduce the
software and computer hardware to be used when speci- overall complexity of the CFD model. Here, experience
fying a problem. For instance, if the software’s only tur- has shown that a model must resolve each fuel jet to cap-
bulence model is the standard k−ε model, then studying ture the governing physics that result in flame shape and
a high swirling isothermal flow (where standard k−ε spread inside a furnace. In all cases, experience is required
is known to perform poorly) may generate erroneous to understand what level of detail is required for a spe-
results. On the other hand, if one is aware of this limita- cific burner/furnace system. In some cases, one may avoid
tion and recognizes that the turbulence model will not analyzing the full system geometry. On the other hand,
accurately predict the axial and tangential momentum of one should include as much of the domain as is practically
the swirl, a conscious decision can be made to neglect the reasonable, given available computer resources.
portions of the solution that are expected to be poor and Recent developments of software vendors of CFD soft-
only use the results that are expected to be meaningful. ware now allow the discretization of those details with
The capability of a CFD package to treat complex very high accuracy, without the downside of creating
geometries is an important consideration for industrial huge numbers of cells and, therefore, increase accuracy
applications; see also Figures 9.9 and 9.10 as examples. of the results without increasing computational time.
The geometries encountered in low emissions burners Generating a computational mesh for this type of
frequently employ complicated shapes and jet angles. The geometries is a well-known bottleneck in a CFD anal-
purpose of these geometries is to precisely control when ysis. Improvements in mesh generation technology
and where the fuel is oxidized. These combustion control greatly benefit industrial CFD users as they allow more
strategies are critical to the performance of the equipment. and more of the actual geometry to be included in the
CFD models must be able to accurately capture the effect CFD model. In addition, mesh generation improvements
of these complex geometries in order to be useful. frequently simplify the process of modifying an existing
Figure 9.11 shows a rendered view of a CFD model geometry. To obtain a solution, the mathematical expres-
of a burner tip with a local mesh refinement. The drill sions describing physical phenomena must be reduced
port sizes are resolved without the use of tetrahedral to simplified but analogous discrete algebraic equations
cells, which would increase the computational effort that require solution at discrete points representing spe-
disproportionally. However, such detail is necessary to cific subvolumes of the overall computational domain.68
FIGURE 9.10
Discretized geometry of a typical boiler burner.
For definition, if the partial differential equation form

of the conservation relationships is used, the resulting
methodology is known as a finite difference scheme. If
the integral form of the equations is used, the resulting
technique is known as the finite volume method.
The arrangement of the discrete points mentioned ear-
lier is referred to as the computational mesh or grid and
provides a segmented approximation of the system geom-
etry to be analyzed or simulated by a CFD model. The
density or number of grid points in any one portion of the
computational domain should be sufficient to resolve the
important geometric features, as well as capture the sig-
nificant physics (e.g., boundary layer, shear layer, recircula-
tion zone, reaction layer, etc.). The density of grid points
typically varies from one region of the domain to another
to help resolve these features. In practice, many times it
FIGURE 9.11 is not feasible or even necessary to resolve the computa-
Close-up view of primary and secondary tips. The view shows fuel tional domain in this way, since some details might be of
jets (indicated by the vectors in the image) issuing from those tips. less importance for the sought solution. If, for example, the
engineer is only interested in the general flow field and
This replacement of “continuum” equations with those temperature distribution in a larger combustion cham-
utilizing discrete numbers is referred to as discretiza- ber, the resolution of the boundary layer on the walls of
tion. Hence, the simulation becomes possible with the that combustion chamber might become less important.
associated computational load related to the number of This also means that whoever generates a model has to
discrete points where values of conserved scalars are define the scope and purpose of this CFD model. For an
calculated in the general domain. These discrete points aeronautical engineer developing a new design of an air-
are referred to as grid points and the collection of grid plane wing, it is more important to take care of resolving
points represents a computational mesh. Once the cal- the exact flow including the behavior in the boundary
culation within the domain is broken down into these layer across that wing. Similarly, a combustion engineer
discrete points, the closed-form of the mathematical must resolve the exact flow at a wall of a large furnace.
expressions can be expressed as a series of algebraic This might change, however, if the purpose of the model
equations. This allows for numerical solution of the is the calculation of heat transfer to the tubes in a radiant
various flow-field variables at each of the grid points. section of a furnace. Then the boundary layer becomes
important again to obtain a more accurate result for the smaller and convergence is achieved if RΦ falls below a
convective heat transfer to those tubes. With today’s com- (to be defined) threshold. Comparison of errors from each
putational capabilities, a CFD model is still a compromise of the seven equations is difficult because of the relative
between the real world physics and the results of the CFD magnitude of the coefficients (Ai’s) for each equation.
model. With increasing computational capacities, that gap Normalization is also difficult due to the range of vari-
becomes decreasingly smaller. able and source term magnitudes within each equation.
Generally, it is acknowledged that greater grid den- Without comparison of the convergence of each equation,
sities (i.e., larger number of elements describing the it is impossible to determine when “overall” convergence
domain), produce more accurate the simulation results. is achieved or which equation is slowing the convergence
process. This phase of the CFD analysis does not usually
require a lot of effort unless severe convergence problems
9.5.2 Solution Convergence
are encountered. Normally, all that is required of the ana-
After generating the mesh and setting up the problem for lyst is to observe the progress of the CFD code toward
solution, the calculations begin. Typically, the momentum convergence and perhaps adjust under-relaxation factors
equations are solved for each velocity component, the tur- and adapt the grid. It is always the stated goal to obtain a
bulence equations are solved for the respective turbulence solution that is grid-independent, but in practice, it usu-
variables, and the continuity equation is left for the pres- ally is too time-consuming to refine the mesh such that
sure field. A key issue in CFD is solving for the pressure the obtained solution can be proven grid-independent.
gradient source terms, found in the momentum equations,
since the pressure fields for enclosed flows are usually
9.5.3 Analysis of Results: Postprocessing
unknown. CFD codes use traditionally the SIMPLE algo-
rithm for solving the equations of motion and continuity A typical CFD simulation provides on the order of 108–
in a decoupled fashion, by transforming the continuity 1010 discrete numerical outputs. For example, a simula-
equation into a pressure correction equation. The codes tion with 5,000,000 nodes and 11 variables per node
use a tri-diagonal algorithm to solve the PDE’s for each (pressure, density, three velocity components, k, ε, tem-
variable along a line on each plane of the computational perature, mixture fraction, variance of mixture fraction,
space. The variables are solved in succession, starting with and irradiation) would generate 55,000,000 numbers. If
a velocity component and ending with one of the turbu- the various chemical species are considered as well as
lence variables. Since the PDE’s are solved in a decoupled the detailed results of a discrete ordinates model, the
fashion, only—four to five “micro-iterations” are required number of variables per node could easily exceed 50,
per variable. A complete cycle through the equation set, leading to 250,000,000 numerical results. The generation
termed a macro-iteration, resolves the nonlinear cou- of two-dimensional plots (for instance, temperature vs.
pling between equations to a prespecified convergence position along the burner centerline), contour plots, veloc-
criterion. Overall convergence typically requires between ity vector plots, streamline plots, and combinations and
2,000 and 10,000 macro-iterations. CFD codes iterate on animations of these outputs are necessary for the analyst
each equation individually, the equation coupling necessi- to understand the results of a simulation. The production
tates simultaneous convergence of the entire equation set. of these different sorts of outputs becomes very impor-
Various methods have been used to measure convergence, tant in communicating the results of a simulation. This
compare convergence rates of each equation, and deter- is especially true when the intended audience is not com-
mine when the required level of convergence is obtained. posed of CFD specialists. Current postprocessing pack-
Typically, the error used to track convergence represents ages have the ability to add lighting to a model, which
the residual for each FDE as shown: makes the images more realistic to the viewer.
Figure 9.12 shows an example of using the rendering
Rφ0 = AEφE + AW φW + AN φ N + ASφS capabilities of a CFD package to generate an image with
photo-realistic qualities. Images such as Figure 9.12 can
+ AT φT + ABφB + SU − AP φ P (9.78)
take anywhere from several seconds to several minutes
where for current generation scientific workstations to render,
Apϕp represents the computational node and the other depending on the number of lights applied, the number of
Aiϕi ’s represent the neighboring nodes surfaces in the scene, and the complexity of these surfaces.
Su represents the source term (RHS of equations High performance virtual reality environments must be
shown earlier) able to regenerate these scenes many times per second.
Rϕ represents the residual or relative error in the equation In addition to still images, animations can be effectively
used to illustrate CFD results. Animated velocity vectors
In other words, a balance around each computational cell and streamlines illustrate the path of fluid flow in internal
is made. As the solution converges, the residual is getting and external flow problems very well. Sweeping planes
X Z
FIGURE 9.12
Rendered view inside an ethylene cracker showing flow patterns near the premixed radiant wall burners. (From Baukal, C.E. (ed.), The John
Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
showing either velocity vectors or filled contour maps Pictures like these require the CFD engineer to under-
of a scalar result can quickly present information about stand the underlying physics of a CFD model and its inter-
an entire three-dimensional simulation. Some output pretation. Relationships like these can also be used for
requires knowledge about the process, or the underlying the improved understanding of mixing mechanisms. The
physics, like for example the picture shown in Figure 9.13. following two Figures 9.14 and 9.15 show an iso-contour,
This figure is a representation of a flame shape. Flame similar to the one in Figure 9.13. However, the value of CO
shapes are not a direct output (quantitative solution) of a is biased here toward much smaller values to illustrate
CFD model, but must be derived from other calculated the mixing of the oxidizer, here coming from a gas tur-
quantities such as the concentration of CO in this example. bine, with fuel released from the duct burners. A proper
FIGURE 9.13
Illustration of a flame envelope defined as an iso-contour of 2500 ppm CO.
interpretation of these pictures leads to the conclusion that an incorrect conclusion. In addition, in an industrial set-
the mixing of oxidizer with fuel for the two top-most burn- ting, the audience will frequently not have the expertise
ers is insufficient leading to a larger combustion product required to assess the quality of a simulation.
envelope. A CFD model helped investigating this issue
and offer a solution, as shown in Figure 9.15.
Generating effective presentations like these, including
still images and animations, is a time-consuming task, 9.6 Summary
since the CFD engineer has to understand the problem CFD modeling of industrial furnaces is a valuable tool,
and find a way to convey the results to the “customer” in which can be used profitably. CFD modeling can help
a way that is descriptive and easy to understand. Creating identify the cause of problems and it can be used to test
a suitable image to make a specific argument frequently solutions. In addition, CFD modeling can be a valuable
requires the analyst to look at and reject a large number design tool for combustion equipment in the petrochemi-
of candidate images. It also requires significant exper- cal industry. It is also clear that CFD has not achieved
tise from the CFD analyst. It is certainly true that CFD the status of stress analysis in terms of ease of use. In
results can be misinterpreted or misapplied to lead to many cases, engineers without advanced understanding
of the physics do stress analysis of mechanical designs
and obtain reasonable results. With CFD analysis, espe-
cially the study of combustion systems, this is not the
case. Understanding and interpreting the results of a
CFD model requires a thorough understanding of the
underlying physics. In a typical furnace model, the sci-
ence involved is multidisciplinary, involving heat trans-
fer, fluid flow, and combustion kinetics.69 For additional
examples of CFD output see Figures 9.16 through 9.20.
Nomenclature
Z
A (−) difference coefficient composed of convec-
Y X
tion/diffusion terms
Da (−) Damköhler number: ratio of reaction time
to flow time
FIGURE 9.14
Illustration of combustion products indicating poor mixing between
f (−) general body force in momentum equation
fuel and oxidizer. g (kg m/s2) gravity
I (W/m2 sr) radiation intensity
k (m2/s2) turbulent kinetic energy
Ks (−) scattering coefficient
Ka (−) adsorption coefficient
lF (m) reaction zone thickness used to define
Damköhler number
lt (m) turbulent length scale used to define
Damköhler number
p (N/m2) pressure
r (m) cylindrical coordinate position variable
R (−) residual or relative equation error
SΦ source term in conservation equations for
general property
SL(m/s) laminar flame speed
Z
tflow (s) characteristic time for flow to adjust to
Y X
imposed shear
trxn (s) characteristic time for chemical species to
FIGURE 9.15
react with each other
Smaller combustion product envelopes indicate improved mixing u (m/s) axial gas velocity
between oxidizer and fuel. v (m/s) radial gas velocity
4.00e+03
3.80e+03
3.60e+03
3.40e+03
3.20e+03
3.00e+03
2.80e+03
2.60e+03
2.40e+03
2.20e+03
2.00e+03
1.80e+03
1.60e+03
1.40e+03
1.20e+03
1.00e+03
8.00e+02
6.00e+02
4.00e+02
2.00e+02
–7.91e–04
Contours of static temperature (f)
FIGURE 9.18
Temperature distribution in flame for a tangential fired boiler.
μe (kg/m s) eddy viscosity

ν (m2/s) eddy diffusivity used in k−ε turbulence
FIGURE 9.16
Velocity contours of multiburner boiler.
model
ρ (kg/m3) density
1.40e+03
τ (kg/m s2) viscous stress tensor
1.36e+03
1.31e+03
ΓΦ (−) general transport coefficient for transport
1.27e+03
1.22e+03
property
1.18e+03
1.14e+03 Φ (−) conserved scalar
1.09e+03
1.05e+03 Φ (−) general transport property
1.00e+03
9.60e+02 Ω (−) represents a solid angle in radiation trans-
9.16e+02
8.72e+02 port equation
8.28e+02
7.84e+02 Ψ (−) dimensionless position variable, differ-
7.40e+02
6.96e+02
6.52e+02
ence equation truncation Ψ (–) truncation
6.08e+02
5.64e+02
error from difference equation
5.20e+02
4.76e+02
4.32e+02
3.88e+02
Y
Overlines
3.44e+02 Z X
3.00e+02
− time-averaged value
Contours of static temperature (k) Aug 16, 2011 ∼ favre or mass weighted averaged value
ANSYS FLUENT 13.0 (3d, pbns, spe, rke)
→ vector quantity
FIGURE 9.17
Predicted temperature distribution for stabilizer wing on a duct Superscripts
burner element.
′ fluctuating portion of instantaneous value
v′ (−) turbulence intensity used to define o initial value
Damköhler number p center point in difference scheme
w (m/s) tangential gas velocity
x (m) Cartesian coordinate position variable
y (m) Cartesian coordinate position variable Subscripts
z (m) Cartesian coordinate position variable i, j, k indices representing coordinate direc-
δij (−) Kronecker delta tions in 3-space
ε (m2/s2) kinetic energy dissipation rate Φ general transport property
μ (kg/m s) viscosity E,W,N,S,T,B
east, west, north, South, top, bottom-
μb (kg/m s) bulk viscosity relative directions in grid
(a) (b)
FIGURE 9.19
(a) Path lines colored with velocity before improved turning vanes added. (b) Path lines colored with velocity showing improved flow distribu-
tion with turning vanes.
(a) (b)

FIGURE 9.20
(a) Windbox velocity distribution to burners before improved baffle system. (b) Improved velocity to burners with improved baffle location.
20. Roberson, J.A. and Crowe, C.T. (1975). Engineering Fluid

Mechanics, Houghton Mifflin, Boston, MA.
References 21. Kim, S.E. and Choudhury, D. (1995). A near wall treat-
1. Bird, R.B., Stewart, W.E., and Lightfoot, E.N. (1960). ment using wall functions sensitized to pressure gradi-
Transport Phenomena, John Wiley & Sons, New York. ent, ASME FED Vol. 217, Separated and Complex Flows,
2. Wilcox, D.C. (1993). Turbulence modeling for CFD, DCW ASME.
Industries, Inc., La Canada, CA. 22. Oran, E.S. and Boris, J.P. (2001). Numerical Simulation of
3. Rodi, W. (1993). Turbulence models and their application Reactive Flow, 2nd edn., Cambridge University Press,
in hydraulics—A state of the art review, Association for Cambridge, U.K.
Hydraulic Research, 3rd edn., Delft, the Netherlands. 23. Bray, K.N.C. (1996). The challenge of turbulent combus-
4. Baldwin, B.S. and Barth, T.J. (1990). A one-equation tion, 26th International Symposium on Combustion, Vol. 1,
turbulence transport model for high Reynolds number pp. 1–26, the Combustion Institute, Pittsburgh, PA.
wall-bounded flows, NASA TM-102847. 24. Warnatz, J., Mass, U., and Dibble, R.W. (1995). Combustion:
5. Spalart, P.R. and Allmaras, S.R. (1992). A one-equation Physical and Chemical Fundamentals, Modeling and
turbulence model for aerodynamic flows, AIAA Paper Simulation, Experiments, Pollutant Formation, Springer-
92–439, Reno, NV. Verlag, Berlin, Germany.
6. Libby, P.A. (1996). Introduction to Turbulence, Taylor & 25. Williams, F.A. (1985). Combustion Theory, Addison-
Francis Group, New York. Wesley, New York.
7. Launder, B.E. and Spalding, D.B. (1972). Mathematical 26. Law, C.K. (1984). Dynamics of stretched flames, chemi-
Models of Turbulence, Academic Press, London, U.K. cal and physical processes in combustion, Fall Technical
8. Speziale, C.G. (1987). On non-linear k-l and k-ε models of Meeting, the Eastern States Section, Clearwater Beach,
turbulence, J. Fluid Mech., 178, 459–478. FL, pp. C1–C11.
9. Yakhot, V. et al. (1992). Development of turbulence mod- 27. Peters, N. (1991). Length scales in laminar and turbulent
els for shear flows by a double expansion technique, flames, Prog. Astronaut. Aeronaut., 35, 155–183.
Phys. Fluids A, 4(7), 1510–1520. 28. Borghi, R. (1988). Turbulent combustion modelling, Prog.
10. Shih, T.-H., Liou, W.W., Shabbir, A., and Zhu, J. (1995). Energy Combust. Sci., 14, 245–292.
A new k-e eddy-viscosity model for high Reynolds num- 29. Kee, R.J. et al. (1989) Chemkin II: A FORTRAN chemi-
ber turbulent flows—Model development and valida- cal kinetics package for the analysis of gas-phase chemi-
tion, Computers Fluids, 24(3), 227–238. cal kinetics, Sandia Report SAND 89-8009B-UC706,
11. Versteeg, H.K. and Malalasekera, W. (1995). An Introduction Reprinted 1992.
to Computational Fluid Dynamics. The Finite Volume Method, 30. Kee, R.J. et al. SENKIN: A FORTRAN program for pre-
Addison Wesley Longman Limited, England, U.K. dicting homogeneous gas phase chemical kinetics with
12. Chou, P.Y. (1945). On the velocity correlations and the sensitivity analysis, Sandia-Report SAND 87-8248UC401,
solution of the equations of turbulent fluctuation, Quart. Reprinted 1991.
Appl. Math., 3, 38–54. 31. Glarborg, P. et al. PSR: A FORTRAN program for mod-
13. Davidov, B.I. (1961). On the statistical dynamics of an elling well-stirred reactors, Sandia-Report SAND
incompressible fluid, Doklady AN. SSSR, 136, 47–50. 86-8209UC4, Reprinted 1992.
14. Harlow, F.H. and Nakayama, P.I. (1968). Transport of 32. Rogg, B. (1993). RUN-1DL: The Cambridge Universal
turbulence energy decay rate, Los Alamos National Laminar Flamelet Computer Code, in: Reduced Kinetic
Laboratory, University of California Report LA-3854. Mechanisms for Application in Combustion Systems,
15. Jones, W.P. and Launder, B.E. (1972). The prediction of Appendix C, Peters, N. and Rogg, B., Eds., Springer-
laminarization with a two-equation model of turbulence, Verlag, Berlin-Heidelberg, Germany.
Int. J. Heat Mass Transf., 15, 301–314. 33. Rogg, B. (1994). RUN-1DL: The Universal Laminar Flame
16. Nallasamy, M. (May, 1985). A critical evaluation of vari- and Flamelet Computer Code, User Manual.
ous turbulence models as applied to internal fluid flows, 34. Rogg, B. (1989). Numerical modelling and computation
NASA TP-2474, Marshall Space Flight Center, Redstone of reactive stagnation-point-flows, in: G.M. Carlomagno
Arsenal, AL. and C.A. Brebbia, eds., Computers and Experiments in
17. Harlow, F.H. and Nakayama, P.I. (January, 1968). Fluid Flow, Carlomagno, G.M. and Brebbia, C.A. eds.,
Transport of turbulence energy decay rate, Rpt. # pp. 75–85, Springer Verlag, Berlin, Germany.
LA-3854, Los Alamos Scientific Laboratory, Los Alamos, 35. Pope, S.B. (1990). Computations of turbulent combustion:
NM. progress and challenges, 23rd International Symposium on
18. Lilleheie, N.I., Ertesvag, I., Bjorge, T., Byggstoyl, S., and Combustion, the Combustion Institute, Pittsburgh, PA.
Magnussen, B.F. (July 11, 1989). Modeling and chemical 36. Westbrook, C.K. and Dryer, F.L. (1984). Chemical kinetic
reactions: Review of turbulence and combustion mod- modeling of hydrocarbon combustion, Prog. Energy
els, NEI-DK-286, DE90 760249, SINTEF/The Norwegian Combust. Sci., 10, 1–57.
Institute of Technology, Division of Thermodynamics. 37. Spalding, D.B. (1971). Mixing and chemical reaction
19. Jones, W.P. and Whitelaw, J.H. (1982). Calculation meth- in steady confined turbulent flames, 13th International
ods for reacting turbulent flows: A review, Combust. Symposium on Combustion, pp. 649–657, Pittsburgh,
Flame, 48, 1–26. PA.
38. Magnussen, B.F. and Hjertager, B.H. (1977). On math- 54. Fiveland, W.A. (1988). Three-dimensional radiative heat-
ematical modeling of turbulent combustion with spe- transfer solutions by the discrete-ordinates method,
cial emphasis on soot formation and combustion, 16th J. Thermophys. Heat Trans., 2(4), 309–316.
International Symposium on Combustion, pp. 719–727, 55. Fiveland, W.A. and Jamaluddin, A.S. (1991). Three-
Combustion Institute, Pittsburgh, PA. dimensional spectral radiative heat transfer solutions by
39. Magnussen, B.F., Hjertager, B.H. Olsen, J.G., and Bhaduri, the discrete ordinates method, J. Thermophys. Heat Trans.,
D. (1978). Effects of turbulent structure and local concen- 5(3), 335–339.
trations on soot formation and combustion in C2H2 diffu- 56. Fiveland, W.A. and Jessee, J.P. (1994). Mathematical
sion flames, 17th International Symposium on Combustion, modeling of pulverized coal combustion in axisymmet-
pp. 1383–1393, Combustion Institute, Pittsburgh, PA. ric geometries, Proceedings of the ASME/EPRI Joint Power
40. Lorra, M. (1999). Investigation of NOx reduction in tur- Generation Conference, Phoenix, AZ.
bulent exhaust gas flows with reburning methods using 57. Fiveland, W.A. and Jessee, J.P. (1995). Comparison of dis-
experimental and mathematical modeling regarding crete ordinates formulations for radiative heat transfer in
detailed reaction mechanisms using the laminar flamelet multidimensional geometries, J. Thermophys. Heat Trans.,
theory, PhD thesis, Ruhr-Universitaet Bochum, Bochum, 9(1), 47–54.
Germany (in German). 58. Sazhin, S.S., Sazhina, E.M., Faltsi-Saravelou, O., and
41. Lorra, M. and Kremer, H. (1999). Mathematical model- Wild, P. (1996). The P-1 model for thermal radia-
ling and experimental evaluation of NOx reduction at tion transfer: Advantages and limitations, Fuel, 75(3),
glass melting furnaces with reburning, Proceedings of the 289–294.
5th International Conference of Combustion Technologies for a 59. Mazumder, S. and Kersch, A. (2000). A fast Monte-
Clean Environment, Lisbon, Portugal. Carlo scheme for thermal radiation in semiconductor
42. Lorra, M., Schnepper, C., and Stephen, S. (2002). processing applications, Numer. Heat Trans., Part B, 37,
Investigation of a duct burner design using CFD capa- 185–199.
bilities in conjunction with full-scale experiments, 60. Thurgood, C.P., Pollard, A., and Becker, H.A. (1995). The
Proceedings of the 6th European Conference on Industrial TN quadrature set for the discrete ordinates method,
Furnaces and Boilers, Estoril, Portugal. ASME J. Heat Trans., 117(4), 1068–1070.
43. Peters, N. (1984). Laminar flamelet models in non-premixed 61. Bayazitoglu, Y. and Higenyi, J. (1979). The higher-order
turbulent combustion, Prog. Energy Combust. Sci., 10, 319. differential equations of radiative transfer: P3 approxi-
44. Bilger, R.W. (1988). The structure of turbulent non- mation, AIAA J., 17, 424–431.
premixed flames, 22nd Symposium (International) on 62. Modest, M.F. (2003). Radiative Heat Transfer, Academic
Combustion, the Combustion Institute, Pittsburgh, PA. Press, an imprint of Elsevier Science, San Diego, CA.
45. Bish, E.S. and Dahm, W.J.A. (1995). Strained dissipation 63. Sun, X. and Smith, P.J. (2009). A parametric case study
and reaction layer analyses of nonequilibrium chemistry in radiative heat transfer using the reverse Monte-Carlo
in turbulent reaction flows, Combust. Flame, 100, 3. ray-tracing with full-spectrum k-distribution method,
46. Bilger, R.W. (1993). Conditional moment closure for tur- J. Heat Transfer, 132(2), 024501(5p).
bulent reacting flow, Phys. Fluids A, 5, 436. 64. Zeeb, C.N. (October 2002). Performance and accu-
47. Mazumder, S. and Modest, M.F. (1998). Turbulence- racy enhancements of radiative heat transfer model-
radiation interactions in nonreactive flow of combustion ing via Monte Carlo, PhD dissertation, Colorado State
gases, J. Heat Transfer, 121(3), 726–729. University, Fort Collins, CO.
48. Barlow, R.S. (1999). Nitric oxide formation in dilute hydro- 65. Siegel, R. and Howell, Jr. (1992). Thermal Radiation Heat
gen jet flames: Isolation of the effects of radiation and tur- Transfer, Hemisphere, Washington, DC.
bulence-chemistry submodels, Combust. Flame, 117, 4. 66. Khan, I.M. and Greeves, G. (1974). A Method for cal-
49. Modest, M.F. (1993). Radiative Heat Transfer, McGraw- culating the formation and combustion of soot in die-
Hill, New York. sel engines, in: Heat Transfer in Flames, Afgan, N.H.
50. Chandrasekhar, S. (1950). Radiative Transfer, Oxford and Beer, J.M., eds. Chapter 25, Scripta, Washington,
University Press, London, U.K. DC.
51. Chandrasekhar, S. (1960). Radiative Transfer, Dover 67. Tesner, P.A., Snegiriova, T.D., and Knorre, V.G. (1971).
Publications, New York. Kinetics of dispersed carbon formation, Combust. Flame,
52. Lathrop, K.D. (1966). Use of discrete-ordinate methods 17, 253–260.
for solution of photon transport problems, Nucl. Sci. 68. Anderson, J.D. Jr. (1995). Computational Fluid Dynamics: The
Eng., 24, 381–388, 1966. Basics with Applications, McGraw-Hill, New York, p. 125.
53. Fiveland, W.A. (1982). A discrete ordinates method for 69. Baukal, C.E., Gerstein, V.Y., and Li, X., Computational Fluid
predicting radiative heat transfer in axisymmetric enclo- Dynamics in Industrial Combustion, CRC Press, Boca Raton,
sure. ASME Paper No. 82-HT-20. FL, 2001.
10
Pollutant Emissions
Charles E. Baukal, Jr., I.-Ping Chung, Stephen B. Londerville, James G. Seebold, and Richard T. Waibel
CONTENTS
10.1 Introduction.................................................................................................................................................................... 212
10.1.1 Emissions in the Hydrocarbon and Petrochemical Industries....................................................................214
10.1.2 Conversions........................................................................................................................................................ 215
10.2 Nitrogen Oxides............................................................................................................................................................. 217
10.2.1 Formation Mechanisms.................................................................................................................................... 218
10.2.1.1 Thermal NOx........................................................................................................................................ 218
10.2.1.2 Prompt NOx.......................................................................................................................................... 220
10.2.1.3 Fuel NOx............................................................................................................................................... 220
10.2.2 Important Factors Affecting NOx.................................................................................................................... 221
10.2.2.1 Air–Fuel Ratio (Stoichiometry)......................................................................................................... 221
10.2.2.2 Gas Temperature................................................................................................................................. 222
10.2.2.3 Air and Fuel Preheat Temperature................................................................................................... 222
10.2.2.4 Fuel Composition................................................................................................................................ 223
10.2.2.5 Air–Fuel Mixing.................................................................................................................................. 226
10.2.3 Abatement Strategies......................................................................................................................................... 227
10.2.3.1 Air Preheat Reduction........................................................................................................................ 227
10.2.3.2 Low Excess Air (LEA)......................................................................................................................... 228
10.2.3.3 Staging.................................................................................................................................................. 228
10.2.3.4 Gas Recirculation................................................................................................................................ 229
10.2.3.5 Ultralean Premix................................................................................................................................. 230
10.2.3.6 Water Injection..................................................................................................................................... 232
10.2.3.7 Reburning............................................................................................................................................ 232
10.2.3.8 Burner Out-of-Service (BOOS).......................................................................................................... 232
10.2.3.9 Low NOx Burners................................................................................................................................ 233
10.3 Combustibles.................................................................................................................................................................. 233
10.3.1 CO and Unburned Fuel..................................................................................................................................... 233
10.3.2 Volatile Organic Compounds........................................................................................................................... 237
10.4 Particulates..................................................................................................................................................................... 239
10.4.1 Combustion-Generated Particulates............................................................................................................... 239
10.4.2 Parameters Controlling Combustion-Generated Particulates.................................................................... 239
10.4.3 Measuring Methods.......................................................................................................................................... 239
10.4.3.1 Bacharach Method.............................................................................................................................. 240
10.4.3.2 Opacity Method.................................................................................................................................. 240
10.4.3.3 Method 5 or ISO-9096......................................................................................................................... 240
10.5 Carbon Dioxide.............................................................................................................................................................. 245
10.5.1 CO2 Generation................................................................................................................................................... 245
10.5.2 CO2 Capture........................................................................................................................................................ 246
10.5.3 CO2 Transport..................................................................................................................................................... 246
10.5.4 CO2 Storage......................................................................................................................................................... 246
10.5.5 CO2 Usage........................................................................................................................................................... 246
10.6 SOx.................................................................................................................................................................................... 247
10.7 Hazardous Air Pollutants............................................................................................................................................. 247
10.7.1 Experimental Setup........................................................................................................................................... 247
211
10.7.1.1 Experimental Facility......................................................................................................................... 247

10.7.1.2 Full-Scale Burner Tests....................................................................................................................... 248
10.7.2 Experimental Results........................................................................................................................................ 249
10.7.2.1 No Systematic Variation..................................................................................................................... 249
10.7.2.2 High-Velocity Jet Mixing Produces Low PICs................................................................................ 252
10.7.2.3 Turndown versus Mixing Rate......................................................................................................... 253
10.7.2.4 BERL-Field Connection...................................................................................................................... 254
10.7.2.5 Refinery Fuel Gas, Natural Gas Equivalency.................................................................................. 255
10.7.2.6 No Effect of Burner Type................................................................................................................... 255
10.7.2.7 Detection Limits.................................................................................................................................. 255
10.7.2.8 Results of the Final Full-Scale Trials................................................................................................ 257
10.7.2.9 Summary.............................................................................................................................................. 257
10.7.3 Process Heater, Petroleum Refinery Emissions Factors............................................................................... 261
10.8 Dioxins and Furans....................................................................................................................................................... 262
References................................................................................................................................................................................. 262
a broad cross section of organizations to improve existing

techniques and to develop new techniques for minimizing
10.1 Introduction
pollution. Table 10.1 shows the air quality standards for
The purpose of this chapter is to alert the interested reader ambient air concentrations of six criteria pollutants identi-
about the potential effects on pollutant emissions of the fied by the U.S. Environmental Protection Agency. Figure
combustion processes in the petrochemical and hydrocar- 10.1 shows that 127 million people in the United States are
bon industries. There continues to be increasing interest living in counties where the ambient air quality does not
in reducing pollutant emissions of all types from all com- meet at least one health-based standard, where the pri-
bustion processes. One prognosticator predicts that this mary pollutants are ozone and particulate matter.2
will continue well into the future.1 These pollutants have Figure 10.2 shows that since the Clean Air Act
deleterious effects on both the environment and on the Amendment was passed in the United States in 1970,
health of humans and animals. Efforts are underway from there has been significant growth in many categories that
TABLE 10.1
U.S. National Ambient Air Quality Standards (NAAQS) as of October 2011
Pollutant Primary/Secondary Averaging Time Level Form
Carbon monoxide Primary 8 h 9 ppm Not to be exceeded more than once
1 h 35 ppm per year
Lead Primary and Rolling 3 month 0.15 μg/m3 Not to be exceeded
secondary average
Nitrogen dioxide Primary 1 h 100 ppb 98th percentile, averaged over 3
years
Primary and Annual 53 ppb Annual mean
secondary
Ozone Primary and 8 h 0.075 ppm Annual fourth-highest daily
secondary maximum 8 h concentration,
averaged over 3 years
Particle PM2.5 Primary and Annual 15 μg/m3 Annual mean, averaged over 3 years
pollution secondary 24 h 35 μg/m3 98th percentile, averaged over 3 years
PM10 Primary and 24 h 150 μg/m3 Not to be exceeded more than once
secondary per year on average over 3 years
Sulfur dioxide Primary 1 h 75 ppb 99th percentile of 1 h daily
maximum concentrations,
averaged over 3 years
Secondary 3 h 0.5 ppm Not to be exceeded more than once
per year
Source: U.S. Environmental Protection Agency.
Pollutant Emissions 213
Ozone (8 h) 119.5
PM2.5 (annual and/or 24 h) 36.9
PM10 14.9
Lead 4.8
NO2 0.0
CO 0.0
SO2 (annual and/or 24 h) 0.2
One or more NAAQS 126.8
0 20 40 60 80 100 120 140 160 180

Millions of people
FIGURE 10.1
Number of people (in millions) living in counties with air quality concentrations above the level of the primary (health-based) National
Ambient Air Quality Standards (NAAQS) in 2008. (From U.S. Environmental Protection Agency, Our nation’s air: Status and trends through
2008, report EPA-454/R-09-002, February 2010.)
220% drive energy consumption such as gross domestic product,

200% vehicle miles traveled, and population. Energy consump-
210% tion directly influences many air pollutants—the more
180%
fuel consumed the more emissions that are generated.
160% 170% However, due to increased emphasis on reducing pollu-
140% tion, aggregate emissions have actually dropped signifi-
120% cantly despite the growth in energy consumption. Much
100% of this decline is due to advancements in technologies such
80% as burners and engines, where less pollution is generated
60%
compared to previous generations of equipment.
44% 51%
Figure 10.3 shows the distribution of air pollution emis-
40%
sions by pollutant type and by source category. The cat-
20% 30% egory of interest here is “Industrial and Other Processes,”
0% which includes hydrocarbon and petrochemical facilities.
–20% While there are other pollutants potentially produced in
–40% those industries, this chapter is only concerned with the
–60% –71%
air pollutants resulting from the combustion processes. For
typical gaseous fuels, the two major pollutants are nitrogen
–80%
dioxide and carbon monoxide. Figure 10.1 shows these are
70 80 90 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10
not currently ambient air concerns compared to the other
Gross domestic product Energy consumption four criteria pollutants (ozone, particulate matter, lead,
Vehicle miles travelled CO2 emissions and sulfur dioxide). There are numerous factors that affect
Population Aggregate emissions the pollutant emissions generated from the combustion of
fuels. The U.S. Department of Energy has classified emis-
FIGURE 10.2 sion factors by fuel type for petroleum refining, as shown
Comparison of growth measures (gross domestic product, vehicle in Table 10.2.3 An EPA report identified the following
miles traveled, population, and energy consumption) and emis- heater design parameters that affect NOx emissions from
sions (CO2 and aggregate emissions) from 1970 to 2010 in the United
States. (From U.S. Environmental Protection Agency, Air quality
process heaters: fuel type, burner type, combustion air pre-
trends, Research Triangle Park, NC, http://www.epa.gov/airtrends/ heat, firebox temperature, and draft type.4 The important
agtrends.html#comparision, accessed on December 10, 2012.) factors that influence pollution are considered here.
Direct PM2.5
Direct PM10
NH3
Air pollutants SO2

NOx
VOC
CO
Lead
0 20 40 60 80 100
Percent of emissions
Source category
Stationary fuel Industrial and Highway Non-road

combustion other processes vehicles mobile
FIGURE 10.3
Distribution of air pollution emissions by pollutant type and source category. (From U.S. Environmental Protection Agency, Our nation’s air:
Status and trends through 2008, report EPA-454/R-09-002, February 2010.)
TABLE 10.2
Combustion Emission Factors (lb/106 Btu) by Fuel Type
Fuel Type SOx NOx CO Particulates VOCs
Distillate fuel 0.160 0.140 0.0361 0.010 0.002
Residual fuel 1.700 0.370 0.0334 0.080 0.009
Other oils 1.700 0.370 0.0334 0.080 0.009
Natural gas 0.000 0.140 0.0351 0.003 0.006
Refinery gas 0.000 0.140 0.0340 0.003 0.006
LPG 0.000 0.208 0.0351 0.007 0.006
Propane 0.000 0.208 0.0351 0.003 0.006
Steam coal 2.500 0.950 0.3044 0.720 0.005
Petroleum coke 2.500 0.950 0.3044 0.720 0.005
Electricity 1.450 0.550 0.1760 0.400 0.004
Source: U.S. Dept. of Energy Office of Industrial Technology,
Petroleum—Industry of the future: Energy and environmental
profile of the U.S. petroleum refining industry, U.S. DOE,
Washington, DC, December 1998, Table 1-11, p. 16.
10.1.1 Emissions in the Hydrocarbon fuel oil combustion, natural gas combustion, and lique-
and Petrochemical Industries fied petroleum gas combustion, respectively. Chapter
5 of AP-42 concerns the petroleum industry, where the
The Western States Petroleum Association (WSPA) and
reader is referred to Sections 1.3 and 1.4 for boilers and
the American Petroleum Institute (API) worked with
process heaters using fuel oil and natural gas, respec-
the California Air Resources Board (CARB) to develop
tively. Chapter 6 of AP-42 is on the organic chemical
air toxics emission factors for the petroleum industry.5
process industry. Reis has written a general book on
Source data were provided in 18 groups. Some of those
environmental issues in petroleum engineering, includ-
groups of relevance here include both refinery-gas-fired
ing drilling and production operations.7 Baukal has
and fuel-oil-fired boilers and heaters fired on natural gas,
written a general purpose book on pollution emissions
refinery gas, oil, and a combination of natural gas and
and control in a wide range of industries including the
refinery oil. The U.S. Environmental Protection Agency
hydrocarbon and petrochemical industries.8 American
has compiled an extensive list of emission factors for a
Petroleum Institute Recommended Practice 556 gives
wide range of industrial processes.6 Chapter 1 of AP-42
some guidelines for instrumentation used to make pol-
concerns external combustion sources and focuses on
lutant emission measurements in gas-fired heaters.9
the fuel type. Sections 1.3 through 1.5 of AP-42 concern
10.1.2 Conversions Example 10.2
It is often necessary to convert pollutant measurements Given: Measured NOx = 20 ppmvd, measured fur-
(e.g., NOx and CO) into a standard basis for both regu- nace temp. = 1800°F.
latory and comparison purposes. The concentration of Find: NOx at a reference temperature of 2000°F.
Solution: ppmmeas = 20, Tmeas = 1800°F, assume
each pollutant is measured at the stack in dry parts per
Tref = 400°F
million by volume (ppmvd). One conversion that is often
necessary is from the measured O2 level in the exhaust
gases to a standard basis O2 level. The method for con-  2000 − 400 
ppm corr = (20)  = 22.9 ppmvd
verting measurements to a standard basis is given by10  1800 − 400 

20.9 − O 2ref
ppm corr = ppm meas (10.1) There are two things to note in the aforementioned
20.9 − O 2meas
example. The first is that the basis temperature was
where chosen as 400°F (200°C), which is an empirically deter-
ppmmeas is the measured pollutant concentration in mined value that applies to many burners commonly
flue gases (ppmvd) used in the hydrocarbon and petrochemical industries.
ppmcorr is the pollutant concentration corrected to a However, this equation should be used with care for
reference O2 basis (ppmvd) more unique burner designs and when there is a very
O2meas is the measured O2 concentration in flue gases large difference between the measured and the refer-
(vol.%, dry basis) ence furnace temperatures. The second thing to notice
O2ref is the reference O2 basis (vol.%, dry basis) is that the NOx increases when the reference tempera-
ture is higher than the measured temperature and vice
versa. As will be shown later, NOx generally increases
Example 10.1
with the furnace temperature.
Given:
Measured CO = 20  ppmvd, measured These two corrections can also be combined into a
O2 = 2% on a dry basis. single correction when both the measured O2 level and
Find: CO at 3% O2 on a dry basis. furnace temperature are different from the reference O2
Solution: ppmmeas = 20, O2meas = 2, O2ref = 3 level and furnace temperature:
 20.9 − 3 
ppmcorr = ( 20)  = 18.9 ppmvd  20.9 − O 2ref   Tref − Tbasis 
 20.9 − 2  ppm corr = ppm meas    (10.3)
 20.9 − O 2meas   Tmeas − Tbasis 

This example shows that CO values will be lower when
the basis O2 is higher than the measured O2, because where the variables are defined in Equations 10.1 and 10.2.
higher O2 levels mean more air dilution and, therefore,
lower CO concentrations. The reverse is true when the
basis O2 is lower than the measured O2 level. Example 10.3
Another correction that may be required is to con- Given: Measured NOx = 20 ppmvd, measured
vert the measured pollutants from a measured furnace O2 = 2% on a dry basis, measured furnace
temperature to a different reference temperature. This temp. = 1800°F.
may be required when a burner is tested at one furnace Find: NOx at 3% O2 on a dry basis at a reference
temperature and needs to be modified to find out the temperature of 2000°F.
equivalent at another furnace temperature. The correc- Solution: ppmmeas = 20, O2meas = 2, O2ref = 3, Tmeas =
tion for temperature is 1800°F, assume Tref = 400°F
 Tref − Tbasis   20.9 − 3   2000 − 400 

ppm corr = ppm meas  ppm corr = ( 20)  = 21.6
 Tmeas − Tbasis 
(10.2)
 20.9 − 2   1800 − 400 

where
ppmmeas is the measured pollutant concentration in In this case, the increase in NOx due to the temperature
flue gases (ppmvd) correction is greater than the reduction in NOx due to
ppmcorr is the pollutant concentration corrected to a the higher O2 reference.
reference temp. basis (ppmvd) Some very common simple conversion calculations
Tref is the reference furnace temperature (°F) for gaseous emissions that are frequently used are
Tmeas is the measured furnace temperature (°F) introduced next. Common examples of gaseous emis-
Tbasis is the basis furnace temperature (°F) sions are NO2, CO, and SO2. As ppmvd varies with the
amount of O2 in the stack, it is sometimes necessary to (ppmvd). Using the general chemical equation for fuel
establish a reference percentage of O2 such as 100 ppm oxidation, the following can be computed:
NOx at 3% O2.
In order to establish the percentage excess air, the dry
 nppmvd  lbmole Chem lbmole  FG  lbmole Fuel
O2 needs to be known versus excess air. This can be   ×   ×
computed from the general HC reaction as follows: 106  lbmole FG lbmole F MWFlb Fuel
MWchem lbm lbm Fuel 106 Btu nlb

 X × × × = 6 m
CH X + ε  1 +  ( O 2 + 3.76 N 2 )
lbmoleChem HHV million 10 Btu
 4
X  X  X
→ CO 2 + H 2O + ( ε − 1)  1 +  O 2 + ε  1 +  3.76 N 2 where n ppmvd = ppmvd of chemical n
2  4  4
 FG   X  X
(10.4)   = 1 + ( ε − 1)  1 +  + ε  1 +  (10.9)
F dry , vol  4   4
From the products of the general form of hydrocar-
bons reactions
% Excess air
Substituting ε = + 1 and simplifying,
 X 100
(ε − 1)  1 +  100
4
O 2 % dry = (10.5)
 X  X n ppmdMWn   X    4.76EA 
1 + ( ε − 1)  1 +  + ε  1 +  3.76  1 +  1 +    + 3.76 
 4  4 HHV(12 + X )  4  100 

where lbm n
= (10.10)
106 Btu
 H
  = “X ” (10.6)
C mole The aforementioned equation is a general relation-

ship between a product species in dry ppm by volume
% Excess air and (lb 10 Btu) for any chemical compound, with a fuel
6
Substituting ε = + 1 and simplifying, a
100 HHV = Btu lbm and EA = % excess air and the molecular
unique relationship between measured oxygen and weight of chemical n. It is applicable for all HC fuels,
excess air is obtained. A general relation for EA and O2 both gas and liquids that are mostly composed of HCs.
dry for any HC fuel is given by It is also common to use units of g/Nm 3 . The conver-
sion is as follows:
1
O 2 %†dry = (10.7)  lbm n   453.6 g   HHV†Btu 
   106 Btu   lb   
  1 lb
1 m m
 + 3.76 + 0.0476
  1 + X   EA  lbm Fuel   lbm FG  gn
  
4  lb FG   N m 3  = N m 3
m

Back solving for the EA, the EA as a function of O2 dry Example 10.4
is given by Given: Fuel = CH4, HHV = 23,875 Btu/lbm
Required: NO2 emission = 0.1 lbm/106 Btu
  Find: ppm at 20% EA on a dry basis
 1  Solution: MWNO2 = 46, x = 4, EA = 20%
+ 3.76
 X
 1 + 
4 x ppm ( 46 )   4   4.76 × 20 
EA% = (10.8)  1 + 1 +   + 3.76 
 1  23 , 875 × (12 + 4 )   4   100 
 − 0.0476 
O
 2 %† dry  lbm
= 0.1
10 6 Btu

It is often necessary to calculate mass emission rate
( )
per Btu produced lb 10 Btu from parts per million
6

x = 80.0 ppm
It is important to note here that the mass emis- Example 10.7

sion rates for NOx are always expressed as NO2 and Given: Firing #6 oil, FBN = 0.3%, Excess air = 25%,
the ppmvd measured in the stack can be all NOx in API gravity = 20
ppm since the compounds only contain one atom of Find: Fuel NO2 contribution given 40% conversion
nitrogen. lbm
in and ppmvd
10 6 Btu
Solution: HHV = 18,740 Btu/lbm, For API g ravity =
Example 10.5
20; %C = 88.01, %H = 10.99
Given: Fuel = CH4, HHV = 23,875 Btu/lbm
Required: CO emission = 0.2 lbm/106 Btu  H 10.99†lbm H 12†lbm C mole H†
  = × × = 1.50
Find: p
pm at 20% EA on a dry basis C mole 88.10 †lbm C mole C 1†lbm H
Solution: MWCO = 28, X = 4, EA = 20%
 0.3 †lbm N   1†lbmole N   1 lbmole NO 2 
 100 lb Fuel   14 †lb N   1 †mole N 
 m  m  
xppm ( 28 )   4   4.76 × 20 
1 + 1 +   + 3.76 
23, 875 (12 + 4 )   4   100   46 lbm NO 2   1 †lbm Fuel   10 6 Btu 
 lb   
 mole NO 2   18, 470 Btu   million Btu 
lbm
= 0.2
10 6 Btu lbm NO 2 lb
= 0.53 †( 40% ) = 0.21 6 m
10 6 Btu 10 Btu

x = 262 ppm

xppm ( 46 )   4.76 ( 25) 
1 + (1.375)  + 3.76 
18,740 (12 + 1.5)   100  
These two examples show the ppm values necessary
to achieve certain mass flow rates for NO2 and CO. It lbm
is often necessary to measure the emission and back = 0.21
10 6 Btu
calculate the mass emission rate as shown herewith in
Example 10.6. x = 147 ppm
It should be noted that the conversion of elemental

Example 10.6 fuel-bound nitrogen compounds to NOx varies widely
Given: Fuel = CH4, HHV = 23,875 Btu/lbm from about 4% to 100% and generally are a function of
x = 140 ppm NOx at 4% O2 dry fuel type and concentration. Low concentrations in gas-
Find: E
A% and lbm/106 Btu eous fuels are near 100%. High concentrations in solid
Solution: MWNO2 = 46, X = 4, O2% dry = 4% and liquid fuels range from 20% to 80%. Staging tech-
niques can reduce this conversion in solid and liquid
  fuels up to 80% lower conversion.
 1 
 + 3.76
4
 1 + 
EA% = 4 = 21%
1  10.2 Nitrogen Oxides
 4 − 0. 0476 

NOx refers to oxides of nitrogen.8,11,12 These gener-
ally include nitrogen monoxide, also known as nitric
140 ppm ( 46 )   4    4.76 × 21  oxide (NO), and nitrogen dioxide (NO2). They may also
 1 +  1 +    + 3.76 
23, 875 (12 + 4 )   4    100  include nitrous oxide (N2O), also known as laughing
gas, as well as other less common combinations of nitro-
lbm
= 0.177 gen and oxygen such as nitrogen tetroxide (N2O4).
10 6 Btu
In most high-temperature heating applications, the
majority of the NOx exiting the exhaust stack is in the
In liquids, solids, and sometimes gaseous fuels, ele- form of nitric oxide (NO).13 NO is a colorless gas that rap-
mental nitrogen compounds exist that convert partially idly combines with O2 in the atmosphere to form NO2
to NOx. The following example shows a general method (see Figure 10.4). NO is poisonous to humans and can
to compute this contribution. It is important to note that cause irritation of the eyes and throat, tightness of the
these computed NOx emissions are in addition to ther- chest, nausea, headache, and gradual loss of strength.
mal and prompt NOx and are a function of f uel-bound Prolonged exposure to NO can cause violent cough-
nitrogen only. ing, difficulty in breathing, cyanosis, and could be fatal.
NO + O2 NO2
H2O
NO + O2 NO2
HNO2, HNO3
FIGURE 10.4
Schematic of NO exiting a stack and combining with O2 to form NO2.
(From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC FIGURE 10.5
Press, Boca Raton, FL, 2001.) Schematic of acid rain. (From Baukal, C.E. (ed.), The John Zink
Science magazine named nitric oxide as its 1992 Molecule

contact with moisture, acid rain is produced. Acid rain is
of the Year.14 The reason is that NO is absolutely essen-
destructive to many things it contacts, including plants,
tial in human physiology. A growing body of research
trees, and man-made structures like buildings, bridges,
indicates its importance in many things such as aiding
and statues (see Figure 10.616,17). Another problem with
digestion, regulating blood pressure, and acting as a
NO2 is its contribution to smog. When sunlight contacts a
messenger in the nervous system. It is also a promising
mixture of NO2 and unburned hydrocarbons in the atmo-
drug in the treatment of persistent pulmonary hyper-
sphere, photochemical smog is produced (see Figure 10.7).
tension, which is a life-threatening lung condition that
Many combustion processes operate at elevated tem-
affects about 4000 babies each year.
peratures and high excess air levels. The combustion
In most high-temperature combustion processes, nitro-
products may have long residence times in the combus-
gen dioxide (NO2) is a significant fraction of the total NOx
tion chamber. These conditions produce high thermal
exiting the stack, although it is generally not as much as NO.
efficiencies and product throughput rates. Unfortunately,
NO2 is a reddish-brown gas that has a suffocating odor. It
these conditions also favor the formation of NOx. Only
is highly toxic and hazardous because of its ability to cause
about 5% of typical NOx in an industrial region of the
delayed chemical pneumonitis and pulmonary edema.
United States comes from industrial sources, compared
NO2 vapors are a strong irritant to the pulmonary tract.
to 44% from highway and off-road vehicles18 as seen in
Inhalation may also cause irritation of the eyes and throat,
Figure 10.3. Figure 10.8 shows that NOx emissions for
tightness of the chest, headache, nausea, and gradual loss
petroleum refineries are comparable to other industries.19
of strength. Severe symptoms may be delayed and include
cyanosis, increased difficulty in breathing, irregular res-
piration, lassitude, and possible death due to pulmonary 10.2.1 Formation Mechanisms
edema. Chronic or repeated exposure to NO2 could cause There are three generally accepted mechanisms for NOx
a permanent decrease in pulmonary function. Intermittent production: thermal, prompt, and fuel. These are each
low-level NO2 exposure may also induce kidney, liver, discussed briefly in this section.
spleen, red blood cell, and immune system alterations.15
Additionally, to the poisoning effect that NOx has on 10.2.1.1 Thermal NOx
humans, there are also other problems associated with
these chemicals. In the lower atmosphere, NO reacts with Thermal NOx is formed by the high temperature reaction
oxygen to form ozone (O3), as well as NO2. Ozone is also of nitrogen with oxygen, by the well-known Zeldovich
a health hazard, which can cause respiratory problems in mechanism.20 It is given by the simplified reaction
humans. NO2 is extremely reactive and is a strong oxidiz- N 2 + O 2 → NO , NO 2 (10.11)

ing agent. It reacts violently on contact with alcohols, hydro-
carbons, organic materials, and fuels. NO2 decomposes on The reaction of the so-called extended Zeldovich mech-
contact with water to produce nitrous acid (HNO2) and anism is
nitric acid (HNO3) which are highly corrosive (see Figure
10.5). When NO2 forms in the atmosphere and comes in N 2 + O ↔ NO + N (10.12)
1908 1969
1944 2002
FIGURE 10.6
Acid rain deterioration examples.
N + O 2 ↔ NO + O (10.13)
N + OH ↔ NO + H (10.14)
Sun
Thermal NOx increases exponentially with temperature.
Above about 2000°F (1100°C), it is generally the predom-
inant mechanism in combustion processes, making it
important in most high-temperature heating applica-
NOx + O2 + Reactive HCs
tions. This means that this mechanism becomes even
more important when the combustion air is preheated
or enriched with oxygen,21 both of which increase the
flame temperature.
Smog
Results from laboratory experiments provided the
information necessary to develop an empirical relation-
ship for high-temperature (i.e., thermal) NO formation.
This empirical relationship reveals several important
parameters that affect NO formation22:
FIGURE 10.7
Schematic of smog formation. (From Baukal, C.E. (ed.), The John Zink
[ NO ] = K1t e K /T [ N 2 ][ O 2 ]1/2 (10.15)
2
3000
2500
NOx emissions (short tons)
2000
1500
1000
500
0
1970 1975 1980 1985 1990 1995 2000
Oil fuel industrial combustion Petroleum refineries and related ind.

Gaseous fuel industrial combustion Wood, pulp and paper and pub. prod.
Nonferrous metals processing Mineral products
Ferrous metals processing Incineration
FIGURE 10.8
NOx emissions in the United States between 1970 and 1999 based on the process. (From Elkins J. et al., National Air Quality and Emissions
Trends Report, 1999, U.S. Environmental Protection Agency, Washington, DC, Report EPA 454/R-01-004, 2001.)
where be influenced significantly is the N2 concentration. The

[NO] is the volume concentration of NO (ppm) NOx reduction techniques utilizing combustion modi-
K1 = 5.2 × 1017 fication (see Section 10.2.3) are based on reducing the
K 2 = −72,300 peak flame temperature, the residence time in the flame
t is the residence time (s), valid for 0–5 s zone, and the excess O2 concentration.
T is temperature (K), valid for the range 1800°C–2000°C
[N2] is the volume concentration of N2
10.2.1.2 Prompt NOx
[O2] is the volume concentration of O2
Prompt NOx is formed by the relatively fast (“prompt”)
For example, if t = 2 s, T = 1975°C, [N2] = 0.7 (70% N2 by reaction between nitrogen, oxygen, and hydrocarbon
volume), and [O2] = 0.30 (30% O2 by volume), then [NO] radicals.23 It is given by the following overall reaction:
can be calculated:
CH 4 + O 2 + N 2 → NO , NO 2 , CO 2 , H 2O , trace species
[ NO ] = 5.2 × 1017 (†2 s)†e −72300/2248K[0.70][0.30]1/2 = 4295 ppm (10.16)
This empirical equation is only valid over a fairly nar- In reality, this very complicated process consists of hun-
row range of conditions, but it shows the impact of vari- dreds of reactions. The hydrocarbon radicals are inter-
ous parameters. NO is linearly dependent on residence mediate species formed during the combustion process.
time, but only up to 5 s; thereafter NO is approximately Prompt NOx is generally an important mechanism in
constant. NO is exponentially dependent on tempera- lower-temperature combustion processes.
ture, linearly dependent on N2 concentration, and
dependent on the square root of O2 concentration. In
most industrial combustion processes, the concentration 10.2.1.3 Fuel NOx
of N2 is generally large and approximately fixed, while Fuel NOx is formed by the oxidation of nitrogen con-
the concentration of O2 is relatively small and somewhat tained in the fuel24 and is given by the following overall
variable. In general, increasing the residence time, the reaction:
temperature, the N2 concentration, and/or the O2 con-
centration in the high-temperature zone all increase
R x N + O 2 → NO , NO 2 , CO 2 , H 2O , trace species (10.17)
NOx. Of these variables, the only one that cannot usually
Chemistry of gaseous pollutant TABLE 10.3

formation and destruction
H NOx Emission Factors by Fuel Type
O
CN N NO
OH Fuel Type Emission Factor (lb/MMBtu)
O2 OH Distillate fuel 0.140
H OH N2
Residual fuel 0.370
O OH O2 Other oils 0.370
O H OH H
Fuel in HCN NCO NH HNO ,O Natural gas 0.140
,H
Refinery gas 0.140
O
O LPG 0.208
NO
OH H2 OH H, OH, O O Coal 0.950
Petroleum coke 0.950
HNCO NH2 O
H Electricity 0.550
Fuel-nitrogen mechanism Source: Pellegrino, J. et al., Energy and environ-
mental profile of the U.S. petroleum refin-
FIGURE 10.9 ing industry, report prepared by
Schematic of fuel NOx formation pathways. (From Baukal, C.E. (ed.), Energetics, Inc. for the U.S. Department of
The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) Energy, November 2007.
An example of a “fuel” (RxN) is ammonia (NH3) where TABLE 10.4

nitrogen is chemically bound to hydrogen atoms. In
Uncontrolled NOx Emission Factors for Typical Process Heaters
reality, there are many intermediate reactions for this
formation mechanism as indicated in Figure 10.9. Fuel Uncontrolled Emission
Factor, lb/106 Btu
NOx is not a concern for high-quality gaseous fuels like
natural gas or propane, which normally have no organi- Thermal
cally bound nitrogen. Although, sometimes natural Model Heater Type NOx Fuel NOx Total NOxa
gas may contain some nitrogen in the form of N2 that ND, natural gas-firedb 0.098 N/A 0.098
contributes to NOx formation; this, however, would not MD, natural gas-firedb 0.197 N/A 0.197
be an example of fuel NOx because the nitrogen is not ND, distillate oil-fired 0.140 0.060 0.200
organically bound. ND, residual oil-fired 0.140 0.280 0.420
Sometimes fuel NOx may be important when oil (e.g., MD, distillate oil-fired 0.260 0.060 0.320
residual fuel oil), coal, or waste fuels are used, which ND, residual oil-fired 0.260 0.280 0.540
may contain significant amounts of organically bound ND, pyrolysis, natural 0.104 N/A 0.104
nitrogen. Table 10.3 shows that heavy oil (residual fuel) gas-fired
produces much more NOx per unit energy (lb/MMBtu) ND, pyrolysis, high- 0.140d N/A 0.140
hydrogen-fuel gas-firedc
than light oil (distillate fuel), generally because heavy
oils contain more organically bound nitrogen than Source: Sanderford, E.B., Alternative control techniques
document—NOx emissions from process heaters, U.S.
lighter oils.25 The challenge with fuel NOx is that it is
Environment Protection Agency report EPA-453/R-93-015,
dependent not only on both the operating conditions February, 1993, Table 2.1, pp. 2–3.
and amount of nitrogen in the fuel, but also on the com- N/A, not applicable; ND, natural draft; MD, mechanical draft.
position and specific characteristics of the fuel.26 This a Total NO = Thermal NO + Fuel NO .
x x x
b Heaters firing refinery fuel gas with up to 50 mole% hydrogen can
makes it challenging to estimate emissions without
have up to 20% higher NOx emissions than similar heaters firing
making measurements.
natural gas.
Table 10.4 shows typical thermal and fuel NOx emis- c High-hydrogen-fuel gas is fuel gas with 50 mole% or greater
sions for process heaters.27 The amount of fuel-bound hydrogen content.

nitrogen converted to NOx ranges from 15% to 100%.27 d Calculated assuming approximately 50 mole% hydrogen.
The conversion efficiency is generally higher the lower

the nitrogen content in the fuel. combustion system (e.g., furnace air infiltration). Some
of these factors are considered next.
10.2.2 Important Factors Affecting NOx
There are many factors that impact NOx formation.
10.2.2.1 Air–Fuel Ratio (Stoichiometry)
These include the oxidizer and fuel compositions and
temperatures, the ratio of the fuel to the oxidizer, the Figure 10.10 shows the predicted (theoretical) NO as
burner and heater designs, the furnace and flue gas a function of the air-to-fuel ratio (stoichiometry) for an
temperatures, and the operational parameters of the air/CH4 flame under adiabatic equilibrium conditions.
Equivalence ratio
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
5000 5000
Fuel lean Fuel rich
4000 4000
H2
CH4
C3H8
NO (ppmvw)
NO (ppmvw)
3000 3000
2000 2000
1000 1000
0 0
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Equivalence ratio
FIGURE 10.10
Adiabatic equilibrium NO as a function of equivalence ratio for air/fuel flames. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook,
Notice that the NO concentration increases at fuel- 10.2.2.2 Gas Temperature

lean conditions and decreases at fuel-rich conditions.
Figure 10.11 shows the predicted NOx as a function of gas
Figure 3.5 shows a plot of the adiabatic equilibrium flame
temperature for stoichiometric air/fuel flames under adia-
temperature for an air/CH4 flame as a function of the
batic equilibrium conditions. This plot shows the impor-
flame equivalence ratio. Notice that the flame tempera-
tance of the gas temperature in thermal NOx formation.
ture for the air/CH4 flame is very dependent on the stoi-
Notice that the NOx rises rapidly at temperatures above
chiometry. This figure helps explain why, for example,
about 2000°F (1100°C) for all three fuel cases shown. This
NOx is reduced dramatically under fuel-rich conditions.
demonstrates the exponential increase in thermal NOx as
One reason is the dramatic reduction in the flame tem-
a function of temperature. Many combustion modification
perature. Another reason is related to the chemistry. In a
strategies used for reducing NOx emissions involve low-
reducing atmosphere, CO is formed preferentially to NO;
ering the flame temperature because it has such a large
this is exploited in some of the NOx reduction techniques.
impact on NOx. For example, one strategy is to inject water
An example of this technique is methane reburn (see
(usually in the form of steam) into the flame in order to
Section 10.2.3.7).28 The exhaust gases from the combus-
reduce NOx by cooling the flame to a lower temperature
tion process flow through a zone that is at reducing con-
where NOx formation is less favorable (see Section 10.2.3.5).
ditions. NOx is reduced back to N2. Any CO that may have
However, water injection reduces the thermal efficiency of
formed in the reduction zone and any other unburned
the heater because it is an added heat load.
fuels are then combusted downstream of the reduction
zone. However, they are combusted at temperatures well
below those found in the main combustion process; these 10.2.2.3 Air and Fuel Preheat Temperature
lower temperatures are not favorable to NOx formation. Figure 10.12 shows how NOx increases with combustion
Another thing to notice in Figure 10.10 is that NOx also air preheat temperature. Air preheating is commonly used
decreases under very-fuel-lean (oxidizing or excess O2) to help increase the overall thermal efficiency of the heat-
conditions. This is because of the reduced flame temper- ing process by recovering energy from the flue gases and
ature caused by all the diluent excess air (see Figure 3.5). transferring it to the incoming combustion air. However,
One technology that takes advantage of this principle it can dramatically increase NOx emissions because of the
is called ultralean premix (see Section 10.2.3.5). With strong temperature dependence of NO formation. Figure
this technology, part of the fuel burns under very-fuel- 3.6 shows how the adiabatic flame temperature increases
lean conditions with the balance of the fuel being added approximately linearly with air preheating. The increase in
somewhere else in the system (e.g., downstream of the NO emissions mimics the increase in flame temperature.
ultralean zone). Overall, the system operates at slightly Figure 10.13 shows how NOx increases with the
fuel-lean conditions (low excess O2 levels). fuel preheat temperature. Fuel preheating is another

1000 1200 1400 1600 1800 2000 2200
2800 2800
2400 2400
2000 H2 2000
CH4
NO (ppmvw)
NO (ppmvw)
1600 C3H8 1600
1200 1200
800 800
400 400
0 0
1000 2000 2400 2800 3200 3600 4000
FIGURE 10.11
Adiabatic equilibrium NO as a function of gas temperature for stoichiometric air/fuel flames. (From Baukal, C.E. (ed.), The John Zink Combustion

0 100 200 300 400 500 600 700 800 900 1000
9000 9000
8000 H2 8000
CH4
7000 C3H8 7000
NO (ppmvw)
NO (ppmvw)
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
FIGURE 10.12
Adiabatic equilibrium NO as a function of air preheat temperature for stoichiometric air/fuel flames. (From Baukal, C.E. (ed.), The John Zink
method used to improve the overall thermal efficiency 10.2.2.4 Fuel Composition
of a heating process. Figure 3.7 shows how the adia-
Figure 10.14 shows how the fuel composition affects
batic flame temperature increases due to fuel preheat-
NO for a blend of CH4 and H 2. First, it is important to
ing. The increase in NOx emissions follows the same
note that NO increases as the H 2 content in the blend
pattern as the increase in flame temperature. Fuel pre-
increases. This is similar to the effect on the adiabatic
heating is usually less important for NOx formation as
flame temperature as shown in Figure 3.8. The second
compared to air preheating for a given preheat tem-
observation is that the effect is not linear between
perature, because the mass flow rate of air is typically
pure CH4 and pure H 2. That is, NOx increases more
much greater than the mass flow rate of fuel.

0 50 100 150 200 250 300 350 400 450 500
3400 3400
3200 H2 3200
CH4
3000 C3H8 3000
2800 2800
NO (ppmvw)
NO (ppmvw)
2600 2600
2400 2400
2200 2200
2000 2000
1800 1800
0 100 200 300 400 500 600 700 800 900 1000
FIGURE 10.13
Adiabatic equilibrium NO as a function of fuel preheat temperature for a stoichiometric air/CH4 flame. (From Baukal, C.E. (ed.), The John Zink
2700 2700
2600 2600
2500 2500
2400 2400
NO (ppmvw)
NO (ppmvw)
2300 2300
2200 2200
2100 2100
2000 2000
1900 1900
0 10 20 30 40 50 60 70 80 90 100
H2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 10.14
Adiabatic equilibrium NO as a function of fuel composition (CH4/H2) for a stoichiometric air/fuel flame. (From Baukal, C.E. (ed.), The John Zink
rapidly as the H2 content increases. The third obser- increases. At 100% N2, the “fuel” produces no NO (nor
vation is that there is a significant difference between does it produce any heat!). The additional quantity of
the two extremes as the NOx ranges from a little less N2 in the fuel does not increase NOx because of the
than 2000 ppmvw to a little more than 2600 ppmvw. increased availability of N2 to make NOx since there is
Figure 10.15 shows how the fuel composition affects already plenty of N2 available from the combustion air.
NO for a blend of CH4 and N2. NO (ppmvw) drops Another aspect of NOx determined by fuel composi-
off rapidly as the N2 concentration in the fuel blend tion is the organically bound nitrogen content in liquid
2000 2000
1800 1800
1600 1600
1400 1400
NO (ppmvw)
1200
NO (ppmvw)
1200
1000 1000
800 800
600 600
400 400
200 200
0 0
0 10 20 30 40 50 60 70 80 90 100
N2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 10.15
Adiabatic equilibrium NO as a function of fuel composition (CH4/N2) for a stoichiometric air/fuel flame. (From Baukal, C.E. (ed.), The John Zink
Quinoline (new) Quinaldine (new)

1.0 2-6-Dimethylquinoline Pyrrole
2-4-Dimethylpyrrole 2-5-Dimethylpyrrole
1,2,5-Trimethylpyrrole Indole
1-Methylindole 2-Methylindole
0.9 3-Methylindole
Conversion rate rN
0.8
0.7
0.0
100 300 600 900
Fuel-N content ξN in mg/kg
FIGURE 10.16
Conversion ratio of fuel-bound nitrogen to NO2 of various nitrogen-containing fuels as a function of fuel–nitrogen content. (From Medeiros, A.
et al., Clean Air, 8(4), 359, 2007.)
fuels. The higher the nitrogen content, the more NOx where
generated, although it is not a linear correlation. Figure ENO2 is the NO2 emissions value in mg of NO2/kWh
10.16 shows that the experimentally determined conver- ξN is the fuel-bound nitrogen content
sion ratio rN of fuel-bound nitrogen to NO2 in a variety MN and MNO2 are the molar masses of nitrogen and
of compounds decreases with fuel–nitrogen content nitrogen dioxide
with rN defined as Hu is the net calorific value of the sample30
Figure 10.17 shows some experimental data gathered

ENO2 MN
rN = Hu (10.18) from John Zink burners firing oils with different levels
ξ N MNO2 of fuel-bound nitrogen. The data again show that the
45
40
35
Conversion rate (%) 30

Burner A
25 Burner B
20
15
10
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Fuel-N, by weight %
FIGURE 10.17
Conversion rate of fuel-bound nitrogen to NOx for two different oil-fired burners.
conversion efficiency declines as the amount of fuel- particularly the actual burner design; therefore, the
bound nitrogen increases, although the actual shape of curves are only meant to be representative, which is
the curve is dependent on the burner design. why no specific NOx levels are given.
The NOx for premix flames has the same general shape
as that predicted by adiabatic equilibrium calculations,
10.2.2.5 Air–Fuel Mixing
as shown in Figure 10.10. However, the magnitude of NOx
The mixing of the air and fuel, which is related to from actual premixed flames is generally much less than
the burner design, directly impacts NOx emissions.30 predicted by adiabatic calculations because actual flames
Figure 10.18 shows a diagram of the relative effects of radiate heat and, therefore, are at lower temperatures as
both the air/fuel mixture ratio and the general type compared to adiabatic conditions. Notice that the peak
of burner design (premix vs. diffusion). Note that the NOx occurs approximately at the air/fuel ratio where
actual shapes of these curves depend on several factors, most industrial combustion systems are operated.
Assuming all other variables are the same, diffusion
flames produce less NOx than premixed flames because
of the delayed mixing. In a diffusion flame, the fuel and
Stoichiometric
combustion air mix over some distance after exiting the burner,
which depends on the specific design. Heat is being con-
Diffusion
tinually released as the fuel and air continue to mix, so
Premix flame
Relative NOx emissions
flame
diffusion flames typically do not reach the same peak
temperatures as premix flames do. As shown in Figure
10.11, NOx is very temperature dependent, so the lower
peak temperatures for diffusion flames generate less
NOx compared to premix flames.
The shape of the curve for the diffusion flame is
Flammability also different than for the premix flame because of
limit the delayed mixing. The peak NOx is shifted to a more
fuel-lean mixture for diffusion flames. While overall
the flame is more fuel lean, the actual air/fuel mix-
Fuel rich Fuel lean ture ratio varies considerably depending on location.
combustion combustion Higher overall levels of excess air result in zones in the
Stoichiometry flame that are closer to stoichiometric conditions while
FIGURE 10.18 the air and fuel are still mixing. Then, there are other
Relative NOx versus air/fuel ratio for premix and diffusion flames. zones in the flame that still have high excess air levels
where the fuel still has not completely mixed with the 10.2.3.1 Air Preheat Reduction
air. Those zones serve to absorb heat from the more
One combustion modification technique is reducing
stoichiometric zones, which reduces the peak flame
the combustion air preheat temperature. As shown in
temperatures, which also reduces NOx compared to
Figure 10.12, reducing the level of air preheat can sig-
premixed flames. At high enough excess air levels,
nificantly reduce NOx emissions. Reducing air pre-
NOx levels decline as well for diffusion flames because
heat significantly reduces NOx for processes that use
of the flame quenching.
heat recuperation. However, air preheat reduction also
reduces the overall system efficiency as shown in Figure
10.2.3 Abatement Strategies
3.12. The loss of efficiency can be somewhat mitigated if
A general strategy for reducing NOx is known as com- the heater is equipped with a convection section. This is
bustion modification. Combustion modification reduces an easy strategy to implement and may be cost-effective
NOx formation by changing the combustion process. if the lost efficiency is more than offset by the cost of
Combustion modification methods are usually less capi- alternative NOx reduction techniques.
tal intensive than most posttreatment methods. In many It is possible to achieve low NOx using high levels
cases, there is a limit to how much NOx reduction can be of air preheat using a specially designed combustion
achieved using these methods. system that utilizes advanced mixing combined with
There are numerous methods that have been used to intense internal flue gas recirculation and strict O2
modify the combustion process for low NOx. A popu- control in the flame zone.31,32 This technology reduces
lar method is low NOx burner design where specially energy consumption because of a heat recovery system
designed burners generate less NOx than previous used to preheat the combustion air, which is built into
burner technologies. Low NOx burners may incorporate the HiTAC (high-temperature air combustion) system
a number of techniques for minimizing NOx including (see Figure 10.19).32 Reducing energy consumption
flue gas recirculation, staging, pulse combustion, and also indirectly reduces all pollution emissions because
advanced mixing. Many of the more common combus- less fuel is consumed, which means less pollution is
tion modification techniques are discussed next. produced.
Pilot burner B
Pilot burner A
Spark plug A
Air Spark plug B Air
Fuel Fuel
Open Close
Flow and combustion
Heat transfer
Fuel Slab Fuel

Open Close
Burner A
Burner B
4-way switching valve
Insulation
Air Exhaust gas

Ceramic honeycomb
FIGURE 10.19
Schematic of HiTAC furnace. (Adapted from Tsuji, H. et al., High Temperature Air Combustion: From Energy Conservation to Pollution Reduction,
10.2.3.2 Low Excess Air (LEA) is carried out of the stack with the exhaust products.35
However, there is a practical limit to the minimum target
As shown in Figure 10.10, excess air causes NOx emissions
excess O2. Since the mixing of the fuel and air in a diffu-
to increase at excess air levels near stoichiometric condi-
sion flame burner is not perfect, some excess air is neces-
tions (0% excess air). Notice, however, that at higher excess
sary to ensure both complete combustion of the fuel and
air levels, NOx actually decreases. There are two primary
minimization of CO emissions. The limit on reducing the
reasons for this increase in NOx near low levels of excess
excess air is CO emissions. If the excess O2 is reduced too
air and then subsequent decrease at higher levels. The
much, then CO emissions will increase. CO is not only a
first concerns chemistry and the priority of reacting with
pollutant, but it is an indication that the fuel is not being
available oxygen. Hydrogen is the most reactive element
fully combusted, resulting in lower system efficiencies.
and reacts first with oxygen to form H2O. Next, carbon
reacts with oxygen. If there is enough oxygen available
10.2.3.3 Staging
and sufficient temperature and mixing, then the carbon
and oxygen react to form CO2. The least reactive element Staged combustion is an effective technique for lowering
in the H-C-N mixture to react with any remaining oxy- NOx. Staging means that some of the fuel or oxidizer or
gen is nitrogen. If there is not enough oxygen (fuel-rich both is added downstream of the main combustion zone.
or substoichiometric conditions) available to fully react The fuel, oxidizer, or both may be staged into the flame.
with carbon, then significant amounts of CO are typically For example, there may be primary and secondary fuel
produced and very little NOx is generated. Some excess inlets where a portion of the fuel is injected into the main
oxygen is required to react with the nitrogen to form NOx. flame zone and the balance of the fuel is injected down-
The second reason for the increase in NOx near low lev- stream of that main flame zone. In fuel staging, some of
els of excess air (near stoichiometric conditions) is that the fuel is directed into the primary combustion zone
the flame temperature (see Figure 3.5) is high under those while the balance is directed into secondary and even
conditions. The combination of available oxygen and high tertiary zones in some cases. This makes the primary
temperatures leads to high NOx. Most combustion sys- zone fuel lean, which is less conducive to NOx formation
tems are designed to operate near stoichiometric condi- when compared to stoichiometric conditions. The excess
tions, with low levels of excess air to maximize thermal O2 from the primary zone is then used to combust the
efficiency. It is also desirable to operate with low levels of fuel added in the secondary and sometimes even tertiary
excess air to minimize NOx emissions as well. The lower zones. While the overall stoichiometry may be the same
limit on target excess air levels is usually CO generation as in a conventional burner, the peak flame temperature
where CO can spike up dramatically (sometimes called is much lower in the staged fuel case because the com-
“CO breakthrough”) if there is insufficient O2, mixing, bustion process is staged over some distance while heat is
and temperature. Since both NOx and CO typically need simultaneously being released from the flame. The lower
to be minimized, a system designed to minimize both peak temperatures in the staged fuel flame help to reduce
will often be operated at the lowest possible excess O2 the NOx emissions. Then, fuel staging is effective for two
level without producing high levels of CO. reasons: (1) the peak flame temperatures are reduced,
Excess air generally comes from two sources: the com- which reduces NOx, and (2) the fuel-rich chemistry in the
bustion air supplied to the burner and air infiltration primary flame zone also reduces NOx. Waibel et al. have
into the heater. Excess air produced by either source is shown that fuel staging is one of the most cost-effective
detrimental to NOx emissions. Excess air increases NOx methods for reducing NOx in process heaters.36 Figure
formation by providing additional N2 and O2 that can 10.20 shows an example of a staged fuel burner.
combine in a high-temperature reaction zone to form In air staging, some of the combustion air is directed
NO. In many cases, NOx can be reduced by simply into the primary combustion zone while the balance is
reducing the excess air through the burners. directed into secondary and even tertiary zones in some
Air infiltration, sometimes referred to as “tramp air,” cases. This makes the primary zone fuel rich, which is
into a combustion system affects the excess air in the less conducive to NOx formation when compared to
combustor and can affect NOx emissions. The quantity stoichiometric conditions. The unburned combustibles
and location of the leakage are important. Small leaks far from the primary zone are then combusted in second-
from the burners are not nearly as detrimental in terms ary and tertiary zones. While the overall stoichiometry
of NOx emissions as large leaks near the flames. By reduc- may be the same as in a conventional burner, the peak
ing air infiltration (leakage) into the furnace, NOx can be flame temperatures are much lower in the staged air
reduced because excess O2 generally increases NOx at lev- case because the combustion process is staged over
els common in industrial combustion systems.33,34 some distance while heat is simultaneously being
Reducing excess air is also beneficial in helping increase released from the flame. The lower temperatures in
thermal efficiency. Any unnecessary air absorbs heat that the staged air flame help to reduce the NOx emissions.
Air
Staged fuel
(a) Primary fuel (b)
FIGURE 10.20
Example of a staged fuel burner: (a) staged fuel burner schematic and (b) example of a staged fuel burner.
Combustion air
Gas gun
Fuel gas
Windbox
Burner block
Primary air
Tertiary air ports
Refractory quarl
Secondary air
FIGURE 10.21
Example of a staged air burner (Hamworthy DFR burner).
Figure 10.21 is an example of a staged air burner com- are to avoid hot spots and make the flame temperature
monly used in down-fired hydrogen reformer furnaces. uniform, to increase the heat release from the flame that
lowers the flame temperature, and to control the chemis-
try in the flame zone to minimize NOx formation.
10.2.3.4 Gas Recirculation
There are two common ways to recirculate combustion
Gas recirculation is a process that causes the products of exhaust products through a flame—furnace gas recircu-
combustion to be recirculated back into the flame. It may lation (FuGR) and flue gas recirculation (FlGR). In FlGR,
seem at first counterproductive to recirculate hot flue exhaust gases are recirculated from the exhaust stack or
gases back into the flame, since NOx increases with tem- flue back through the burner. This requires some type
perature. However, although the combustion products of fan or fuel eductor system to circulate the gases exter-
are hot, they are considerably cooler than the flame itself. nal to the furnace and back through the burner. The
The cooler furnace or flue gases act as a diluent, reducing burner must be designed to handle both the added flow
the flame temperature, which in turn reduces NOx (see and elevated temperature of the recirculated gases that
Figure 10.11). Advanced mixing techniques use carefully are often partially or fully blended with the combustion
designed burner aerodynamics to control the mixing air. Garg estimates NOx reductions of up to 50% using
of the fuel and the oxidizer. The goals of this technique flue gas recirculation.37
In furnace gas recirculation, the combustion prod-

ucts are recirculated inside the furnace back toward
the burner and are inspirated into the flame to moder-
ate its temperature (see Figure 10.22). Figure 10.23 is an
example of a burner that uses furnace gas recirculation
to reduce NOx emissions. The Halo™ burner uses a spe- Burner
cially shaped Coanda surface on the top of the tile to
control the air–fuel mixing and to enhance furnace gas
entrainment into the flame.38 The unique design allows
for more furnace gas entrainment compared to other
burner designs, which cannot entrain as much fur- Remote stage fuel tip
nace gas because of stability issues where the flame can
become so diluted that the mixture is no longer flam-
mable. The difference with the Halo burner design is
that the furnace gases are not as uniformly mixed with
the air and fuel, so there is a region that is still flam-
mable, despite higher levels of furnace gas entrainment. FIGURE 10.24
Another method utilizing the furnace gas recirculation Remote stage fuel tip.
technique is called remote stage fuel injection.39 A remote
staged furnace burner configuration includes placement
of secondary fuel gas nozzles remote from burners (see
Figure 10.24). This configuration brings about an increased
mixing of secondary fuel with furnace flue gases as illus-
trated in Figure 10.25. As a result, the temperature of the
burning fuel gas is lowered reducing NOx emission and
providing flameless combustion, as shown in Figure 10.26.
There are some important advantages of FuGR as com-
pared to FlGR. When using the FlGR technique, an addi-
tional hot fan may be needed to extract flue gases from
the stack and deliver it to the burners. In some cases, if
the flue gas temperatures are low enough, it may be pos-
sible to use the combustion air fan to pull both combus-
FIGURE 10.22
tion air and flue gases and send them to the burners. This
Schematic of furnace gas recirculation. (From Baukal, C.E. (ed.), The is commonly done in boiler burner applications where
John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) the flue gas temperature is generally fairly low. However,
even if an additional fan is not needed, additional power
is required to move the flue gases from the exhaust stack
to the burners. Another potential disadvantage of using
FlGR is that additional ductwork may need to be insu-
lated. Finally, the physical size of the burner usually needs
to be larger to accommodate the added flow of the recir-
culated flue gases. The burner components may also need
to be modified to handle the higher flue gas temperatures.
10.2.3.5 Ultralean Premix

A relatively new development in industrial burner tech-
nology for reducing NOx is called lean premix.40 Lean pre-
mix involves running the primary part of the flame with
as much air as possible and as little fuel as possible—that
is, to make the air/fuel mixture as fuel lean as possible.
This is not conducive to NOx emissions as previously
FIGURE 10.23
shown in Figure 10.10. The main challenge with operat-
Example of a burner incorporating furnace gas recirculation (John ing the primary zone near the lean flammability limit is
Zink Halo™ burner). to maintain flame stability. Ren et al. discuss the use of
4. Staged fuel and flue gas mixture

mixes with air-fuel premix and burns
3. Stage fuel entrains furnace flue gas

prior to mixing with premix air-fuel
stream resulting in lower NOx emissions
2. Stage fuel eliminated
1. Stage fuel injected below radiant wall burners
FIGURE 10.25
Illustration showing how the remote stage method provides lower NOx emissions.
With remote staging, NOx = 16 ppmvd Without remote staging, NOx = 24 ppmvd
(a) (b)
FIGURE 10.26
Radiant wall burners firing (a) with and (b) without remote staging.
lean premix combined with methane reforming as an stoichiometric levels of air (referred to as fuel lean) pro-
option to either preheating the fuel or adding expensive duce less NOx, as shown in Figure 10.10. The primary rea-
components like hydrogen or carbon monoxide to the son this technique lowers the NOx emissions is because
fuel to increase stability.41 Both of those techniques add it lowers the flame temperatures as shown in Figure 3.5;
cost and complexity to the combustion system. As pre- as discussed earlier, lower flame temperatures can dra-
viously shown, the addition of hydrogen also increases matically reduce NOx. This principle may seem to con-
flame temperatures and usually NOx emissions. tradict the discussion in Section 10.2.3.2 where low excess
Lean premix technology utilizes the principle air levels are recommended to reduce NOx. Notice how
that air/fuel mixtures with significantly higher than NOx emissions peak in Figure 10.10 at slightly fuel-lean
conditions and then steadily decline as the air–fuel mix- combustion process flow through this fuel-rich reduc-
ture becomes leaner. The reason for this peak is the coming zone, NOx is reduced back to N2 and O2. Any CO and
bination of available oxygen and high flame temperatures other unburned fuels in the exhaust gases are then com-
as previously discussed in Section 10.2.2. Although there busted downstream of the reburn zone at temperatures
is even more available oxygen at very lean conditions, the well below those found in the main combustion process.
flame temperature is much lower, which produces less These lower temperature reactions are less favorable to
NOx. The air–fuel ratio for ultralean premix is typically NOx formation so the net effect is that NOx is reduced.
limited to the lower flammability limit of the mixture. If This is a type of nitrogen reformer.43
the mixture is too lean, the flame can become unstable or There are some challenges with this technique. One is to
even go out because there is not enough fuel to sustain get proper injection of the reburning gas into the exhaust
combustion. products. Another challenge is that the reburn zone must
be capable of sustaining combustion. It is important that
the gas be injected in a lower-temperature region of the
10.2.3.6 Water Injection
process in order to minimize NOx formation. For exam-
Many of the combustion modification methods attempt to ple, the gas may need to be injected in a previously unin-
reduce the temperature of the flame to lower NOx emis- sulated portion of ductwork that may have to be replaced
sions. In some cases, this may result in a reduction in com- with higher-temperature materials and insulated. A third
bustion efficiency.42 Another NOx reduction technique is challenge is trying to take advantage of some of the
to inject water into the flame. However, the water absorbs energy produced during the reburning. In some cases, a
heat from the flame and carries some of that energy out heat recovery system may need to be added in order to
with the exhaust gases, preventing the transfer of much of increase the overall thermal system efficiency.
that energy to the load, so the efficiency is usually reduced. Dagaut et al. have shown that the reburning fuel has
Another form of water injection is to inject steam. some effect on NOx reduction.43 For the fuels considered,
There are several advantages to using steam as com- the ranking from lowest to highest NOx reduction was as
pared to liquid water. One advantage is that steam is follows: methane, natural gas, ethane, ethylene, and acety-
much hotter than liquid water and includes the latent lene. Zamansky et al. showed experimentally that adding
heat of vaporization needed to change the liquid water nitrogen agents like ammonia, urea, and sodium salts into
to a vapor. When liquid water is injected into a com- the flue gases with a delay time of 0.1–0.5 s after injection
bustion process, it can put a large heat load on the proof the reburning fuel can enhance the reburning process.44
cess because liquid water can absorb a large amount of At that point in the process, the oxygen content in the flue
energy before vaporizing due to its high latent heat of gases has been depleted to the point that the injection of
vaporization. Steam puts a much smaller load on the nitrogen agents does not promote NO generation. The
process because it absorbs less energy than liquid water; enhanced NO reduction can be explained by the presence
therefore, it does not reduce the thermal efficiency as of additional active radicals that promote the reduction.
much as liquid water. Another reason for using steam Miller et al. experimentally demonstrated that fuel-
instead of liquid water is that steam is already in vapor lean reburning can reduce NOx emissions in exhaust
form and mixes readily with the combustion gases. gas streams.45 The application of reburning in field
Liquid water must be injected through nozzles to form applications has often been restricted to overall fuel-
a fine mist to disperse it uniformly into the combus- rich or reducing conditions in the reburning zone.
tion gases. Therefore, it is often easier to blend steam This research shows it is technically possible to utilize
into the combustion products compared to liquid water. reburning for NOx reduction of up to 50% in fuel-lean
The advantage of water injection is that the water flow (up to 6% O2 in the flue gases) environments. Watts
rate can be easily adjusted to get variable levels of NOx et al. have written a brief review of the use of reburn
reduction. Fine tuning is not generally as easy with technology to reduce NOx emissions in coal-fired boil-
some of the other NOx reduction methods. ers.46 Braun-Unkhoff et al. compared experimental data
on reburning technology applied on a large scale with
numerical predictions.47 There is a good correlation
10.2.3.7 Reburning
between the measurements and predictions. Wendt has
Reburning is a technique similar to fuel staging, but uses a shown that nonideal mixing can enhance the effective-
slightly different strategy. An example is methane reburn. ness of reburning for diffusion flames.48
The methane reburn technique injects some methane in
the exhaust gas, usually well after the primary combus-
10.2.3.8 Burner Out-of-Service (BOOS)
tion zone, where the gases are at a lower temperature.
As previously shown in Figure 10.10, fuel-rich conditions This is a technique primarily used in boilers where the
are not favorable to NOx. As the exhaust gases from the fuel is turned off to the upper burners, while maintaining
the airflow.49 The fuel removed from the upper burners Today’s lowest NOx burners can be in the single digits
is then redirected to the lower burners, while maintain- for NOx, depending on the operating conditions such
ing the same airflow to the lower burners. Therefore, the as fuel composition and combustor temperature. Figure
overall fuel and airflow to the boiler remains the same, 10.27b shows a similar graph, but for radiant wall burn-
but is redistributed; this makes the lower burners fuel ers. Radiant wall burners are commonly used in crack-
rich, which is less conducive to NOx formation due to ing furnaces. These furnaces are generally hotter than
the lower flame temperatures and fuel-rich chemistry. process heaters, which means the NOx levels are typi-
The upper burners, running on air only, provide the rest cally higher due to the higher temperatures compared
of the air needed to fully combust the fuel. Rather than to round flame burners, all other things being the same.
air staging in individual burners, the BOOS technique These graphs are only representative. One could expect
stages air over the entire boiler. This technique is rela- that new technologies will add additional and even
tively inexpensive to implement. Ensuring proper heat lower curves in the future since the downward trend of
distribution is important to prevent overheating the lower NOx is evident.
tubes or derating the firing capacity. An example of a low-NOx burner is the COOLstar™
burner (see Figure 10.28).51 This design solves some of
the previous issues associated with low-NOx burners.
10.2.3.9 Low NOx Burners
It is easier to retrofit in existing heaters because of its
Garg discusses the use of low NOx burners to achieve compact size, which reduces some of the burner-to-
emissions reductions compared to standard gas burners.50 burner interaction issues with previous designs. It pro-
Table 10.5 shows typical NOx reductions using vari- duces shorter flames and has a higher turndown ratio as
ous low NOx burner techniques. An EPA study found compared to many other low-NOx burners. One of the
that ultralow NOx burners were the most cost effective unique aspects of this patented burner52 is the folded
means to reduce NOx.27 flame pattern that produces a shorter flame length com-
Low NOx burners incorporate some of the techniques pared to other low-NOx burners. This is produced pri-
discussed in this section such as air and fuel staging, fur- marily by the combination of the burner tile shape and
nace gas recirculation, and ultralean premix. “Low NOx” the fuel injection pattern. A variety of NOx reduction
is a relative term because there is no agreed-upon defini- techniques have been used in the burner design includ-
tion of exactly what “low” means, such as a specific value ing fuel gas staging, internal flue gas recirculation, and
of NOx emissions. A low NOx burner for one manufac- advanced mixing concepts. The burner was developed
turer might be an ultralow NOx burner for another. through a combination of computational fluid dynamics
Figure 10.27a shows a graph of the history of low-NOx (CFD) modeling53 (see Chapter 9) and experimentation.
burner development for round flame process burners Figure 10.29 shows how CFD was used to identify prob-
since the passage of the Clean Air Act Amendment in lems with the initial design and to show the improved
the United States in 1970. Going from top to bottom, design. Subsequent field testing validated the design.
the curves are for increasing more modern technolo-
gies with the oldest at the top and most modern at the
bottom. The graph shows two things: NOx emissions
have significantly decreased with each new genera-
tion of technology and NOx emissions increase with
10.3 Combustibles
excess air level for the range given (see Section 10.2.3.2).
This section has been broken into two types of combus-
TABLE 10.5 tibles. The first involves the incomplete combustion of
the fuel, which usually produces carbon monoxide and,
NOx Reductions for Different Low-NOx Burner Types
in some limited cases, not all of the hydrocarbon fuel
Burner Type Typical NOx Reductions is consumed and passes through the combustor unre-
Staged-air burner 25%–35% acted. The second type of combustible is volatile organic
Staged-fuel burner 40%–50% compounds (VOCs), which are generally only important
Low-excess-air burner 20%–25% in a limited number of processes, typically involving
Burner with external flue gas 50%–60% contaminated or otherwise hazardous waste streams.
recirculation (FGR)
Burner with internal flue gas 40%–50%
recirculation 10.3.1 CO and Unburned Fuel
Air or fuel-gas staging with internal FGR 55%–75% CO is generally produced in trace quantities in many com-
Air or fuel-gas staging with external FGR 60%–80% bustion processes as a product of incomplete combustion.
Source: Garg, A., Chem. Eng. Prog., 90(1), 46, 1994. CO is a flammable gas, which is nonirritating, colorless,
234
140
Conventional
120
100
80 Staged air
60
Staged fuel >100 ppm
40 NOx
NOx, ppm (corr. 3% oxygen)

Internal FGR
20
60–100 ppm
Ultra-lean premix
0
0 1 2 3 4 5
Excess oxygen, (%) 40–60 ppm
25–40 ppm
(b) 1930–1970 1970–2000 2002 Today

(a)
FIGURE 10.27
History of low NO burner development for (a) round flame burners and (b) radiant wall burners, firing on gaseous fuels.
(a) (b)
FIGURE 10.28
COOLstar burner: (a) drawing, (b) photograph.
(a) (b)
FIGURE 10.29
Computational fluid dynamic modeling of the COOLstar burner: (a) initial design and (b) revised design. (From Chung, I.-P. et al., Hydrocarbon
Eng., 10 (8), 77–80, 2005.)
odorless, tasteless, and normally noncorrosive. CO is system is operated slightly fuel lean (excess O2) to ensure
highly toxic and acts as a chemical asphyxiant by combin- complete combustion and to minimize CO emissions.
ing with hemoglobin in the blood that normally transports Figure 10.30 shows the calculated CO as a function of
oxygen inside the body. The affinity of CO for hemoglobin the equivalence ratio (ratio of one is stoichiometric, >1
is approximately 300 times more than the affinity of oxy- is fuel rich, and <1 is fuel lean). Because these are adia-
gen for hemoglobin.54 CO preferentially combines with batic calculations with very high flame temperatures,
hemoglobin to the exclusion of oxygen so that the body the dissociation in the flame produces high quantities
becomes starved for oxygen, which can eventually lead to of CO even under fuel-lean conditions. This is graphi-
asphyxiation. Therefore, CO is a regulated pollutant with cally shown in Figure 10.31, where much more CO is
specific emissions guidelines depending on the applica- produced at higher gas temperatures, all other variables
tion and the geographical location. remaining the same.
CO is generally produced by the incomplete combus- Figures 10.32 and 10.33 show the effects on CO pro-
tion of a carbon-containing fuel. Normally, a combustion duction of air and fuel preheating, respectively. In both
Equivalence ratio
0.5 0.6 0.7 0.8 0.9 1.0 1.1
10,000 10,000
9,000 9,000
8,000 8,000
CH4
7,000 C3H8 7,000
6,000 6,000
CO (ppmvw)
CO (ppmvw)
5,000 5,000
4,000 4,000
3,000 3,000
2,000 Fuel Fuel 2,000

lean rich
1,000 1,000
0 0
0.5 0.6 0.7 0.8 0.9 1.0 1.1
Equivalence ratio
FIGURE 10.30
Adiabatic equilibrium CO as a function of equivalence ratio for air/fuel flames. (From Baukal, C.E., Industrial Combustion Pollution and Control,
Marcel Dekker, New York, 2004.)

1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600
9000 9000
8000 8000
7000 CH4 7000

C3H8
6000 6000
CO (ppmvw)
CO (ppmvw)
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000
0 0
1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600
FIGURE 10.31
Adiabatic equilibrium CO as a function of gas temperature for stoichiometric air/fuel flames. (From Baukal, C.E., Industrial Combustion
Pollution and Control, Marcel Dekker, New York, 2004.)
cases, the higher flame temperatures produced by pre- no CO is generated. Similarly, Figure 10.35 shows the
heating cause more CO formation as the preheat tem- effect of fuel composition for CH4/N2 blends. Higher
perature increases. Figure 10.34 shows the effect of fuel concentrations of N2 both reduce the flame temperature
composition for H2/CH4 blends. As expected, higher and the concentration of carbon available to make CO,
concentrations of H2 produce less CO, and for pure H2, which both reduce CO generation.

0 200 400 600 800 1000 1200 1400 1600 1800 2000
45,000 45,000
40,000 40,000
CH4
C3H8
35,000 35,000
30,000 30,000
CO (ppmvw)
CO (ppmvw)
25,000 25,000
20,000 20,000
15,000 15,000
10,000 10,000
5,000 5,000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
FIGURE 10.32
Adiabatic equilibrium CO as a function of air preheat temperature for stoichiometric air/fuel flames. (From Baukal, C.E., Industrial Combustion

0 100 200 300 400 500 600 700 800 900 1000
15,000 15,000
CH4
14,000 14,000
C3H8
13,000 13,000
CO (ppmvw)
12,000 12,000 CO (ppmvw)
11,000 11,000
10,000 10,000
9,000 9,000
8,000 8,000
0 100 200 300 400 500 600 700 800 900 1000
FIGURE 10.33
Adiabatic equilibrium CO as a function of fuel preheat temperature for a stoichiometric air/CH4 flame. (From Baukal, C.E., Industrial Combustion
10.3.2 Volatile Organic Compounds pollutants because they can cause photochemical smog
and depletion of the ozone layer if they are released
VOCs are generally low-molecular-weight aliphatic and
into the atmosphere. They are not normally produced
aromatic hydrocarbons like alcohols, ketones, esters, and
in the combustion process, but they may be contained in
aldehydes.55 Typical VOCs include benzene, acetone, acet-
the material that is being heated, such as in the case of a
aldehyde, chloroform, toluene, methanol, and formalde-
contaminated hazardous waste in a waste incinerator.
hyde. These compounds are considered to be regulated
9000 9000
8000 8000
7000 7000
6000 6000
CO (ppmvw)
CO (ppmvw)
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000
0 0
0 10 20 30 40 50 60 70 80 90 100
H2 in fuel (vol.%)
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
FIGURE 10.34
Adiabatic equilibrium CO as a function of fuel composition (CH4/H2) for a stoichiometric air/fuel flame. (From Baukal, C.E., Industrial
Combustion Pollution and Control, Marcel Dekker, New York, 2004.)
9000 9000
8000 8000
7000 7000
6000 6000
CO (ppmvw)
CO (ppmvw)
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000
0 0
100 90 80 70 60 50 40 30 20 10 0
CH4 in fuel (vol.%)
0 10 20 30 40 50 60 70 80 90 100
N2 in fuel (vol.%)
FIGURE 10.35
Adiabatic equilibrium CO as a function of fuel composition (CH4/N2) for a stoichiometric air/fuel flame. (From Baukal, C.E., Industrial
Combustion Pollution and Control, Marcel Dekker, New York, 2004.)
In that case, the objective of the heating process is usually them difficult to incinerate; they are then emitted with the
to volatilize the VOCs out of the waste and combust them exhaust products. The quenching could be caused by con-
before they can be emitted to the atmosphere. tact with cold gases or possibly by impingement on a cool
There are two strategies for removing VOCs from the surface, such as a boiler tube.
waste gases of a combustion process.55 One is to separate Soot particles can cling to the exhaust ductwork, clog-
and recover them using techniques like carbon adsorp- ging the ductwork and other pollution treatment equip-
tion or condensation. The other method involves oxidizing ment in the system. This creates equipment maintenance
the VOCs to CO2 and H2O. This process includes tech- problems. In addition, most industrial heating processes
niques like thermal oxidation, catalytic oxidation, and have a regulated limit for particulate emission.
bio-
oxidation. One common way to ensure complete
destruction of VOCs in waste incinerators is to add an
10.4.2 Parameters Controlling Combustion-
afterburner or secondary combustion chamber, which
Generated Particulates
may or may not have a catalyst, after the main or primary
combustion chamber.56 Particulate emissions depends on many variables. Some
of them are controlled by the burner or oil gun design. In
addition to the oil gun or burner designs, oil gun opera-
tion conditions must be considered. Atomization quality
is the main driving force affecting particulate genera-
10.4 Particulates
tion—the better the atomization, the fewer the particulate
The impact of particulate emissions on human health emissions. This is discussed in detail in Chapter 5.
has been studied tremendously. Some examples are The other parameters that are not related to the oil gun or
listed in the references.57–61 This section concentrates the burner are (1) oil quality and (2) furnace environment.
on particulate generation and measuring methods in Heavy oil usually contains some solid particles. Any
industries for particulate pollutant emissions. particulates that are originally contained in the oil will
Usually, gas firing and light oil firing do not gener- leave the furnace with the exhaust products. The chemi-
ate particulate emissions that impact the environment. cal composition of the particulates may be changed, but
Heavy oils like No. 6 causes concern on particulate the overall mass is approximately constant. The original
emissions. Heavy oil flames tend to generate particu- particulate loading can be quite significant. This should
lates due to high carbon contents and the difficulty in be noted when considering the particulate emissions.
fully oxidizing droplets prior to exiting the stack. As previously discussed, the furnace environment also
influences particulate emissions. The two most impor-
tant parameters are the furnace temperature and the resi-
10.4.1 Combustion-Generated Particulates
dence time. Particulates will be generated when the soot
Combustion-generated particles are called soot. Under particle is quenched. Therefore, the hotter the furnace
certain conditions, even gaseous fuels may produce is, the lower the particulate emissions are. The residence
soot. Fuels that have a higher carbon-to-hydrogen time is the time that a particle spends in the furnace. The
mass ratio tend to produce more soot than fuels with longer the residence time, the more chance the soot parti-
a lower ratio. For example, propane (C3H8), which has cles would be burnt out; consequently, the less the partic-
a C:H mass ratio of about 4.5, is more likely to produce ulate emissions will be. Therefore, particulate emissions
soot than methane (CH4), which has a C:H mass ratio might be different for two different furnaces even though
of about 3.0. For clean burning fuels like natural gas, it the oil guns or oil burners are the same with similar oper-
is more difficult to produce sooty flames compared to ation conditions.
other fuels, such as oil and coal, which have little hydro-
gen and a high concentration of carbon.
10.4.3 Measuring Methods
Soot can be produced by operating in a very fuel rich
mode or by the incomplete combustion of the fuel due to Many methods are used in industry for particulate mea-
poor mixing. Flames containing more soot are more lumi- surements, for example, Bacharach method,62 Opacity
nous and tend to radiate heat more efficiently than flames method,63 Method 5,64 and ISO-9096.65 The Bacharach and
containing less soot, which tend to be transparent. Soot Opacity methods are qualitative methods. The Method 5
particles generally consist of high-molecular-weight poly- and ISO-9096 methods are quantitative methods. Method
cyclic hydrocarbons. Ideally, soot is expected to be gen- 5 is recommended by the U.S. EPA and usually used for
erated at the beginning of the flame and to be destroyed industrial compliance. In European countries, ISO-9096
before exiting the flame. In this way, no particulates would is well accepted. The procedure and instrument required
be emitted to the environment. However, in reality, soot for Method 5 are very similar to those in ISO-9096. ISO-
particles are sometimes quenched or “frozen,” making 9096 is a concise version of the U.S. EPA Method 5.
1800s, Ringelmann developed the Ringelmann smoke

charts that were introduced in the United States in 1904
and adopted as a standard in 1910. Six charts were cre-
ated ranging from all white to all black corresponding to
charts 0–5.67 Comparing stack opacity to these charts was
used for many years as the standard measure of opacity.
The EPA has adopted Method 968 as a standard to
determine opacity, “Visual determination of the Opacity
of Emissions from Stationary Sources.” This method
uses trained and certified observers to determine opac-
ity by visual observation; not significantly different
than that of Ringelmann charts of 100 years ago.
Today it is common for local air quality districts to use
FIGURE 10.36 both EPA Method 9 and/or opacity measuring instru-
Bacharach smoke tester included a hand pump, filter papers, and spot ments. These instruments must be certified to standard
scale sheet. (Courtesy of Bacharach, Inc., New Kensington, PA.)
specification.69 Opacity in stack plumes is caused by
the presence of particles in the plume that scatter light.
10.4.3.1 Bacharach Method These particles are generally generated from solid and
The Bacharach method is specified in U.S. Standard liquid fuels and are a combination of ash, carbon loss,
ASTM D1256-56 and in German Standard DIN51-402. and sulfur related compounds. The formation of these
The Bacharach smoke tester is a hand pump to suck the particles is covered in other chapters under atomization
exhaust gases from the stack as shown in Figure 10.36. (Chapter 5) and solid fuels (Chapter 6).
Filter paper is installed inside the suction tube, which For the same mass particulate emission rates, the opac-
is connected to the hand pump. After 10 full strokes of ity can change significantly according to the size of the
the hand pump, the filter paper is filled with particu- particles. Generally, smaller particle size distributions will
lates. Then, the filter paper is removed and compared result in high light scatter and higher opacity. For instance,
with the Bacharach smoke spot scale, which consists of 1 μm particles produce eight times greater opacity than
10 spots from 0 to 9. Scale 0 is the lightest color with 10 μm particles. However, this is not continuous and peaks
the least particulate emissions and scale 9 is the darkest at about 0.8 μm. Below 0.8 μm, this trend reverses and
color with the worst particulate emissions. further reduction in particles results in lower opacity. In
To achieve accurate sampling, it is suggested to hold actual systems, most particulates are greater than 1 μm.70
each stroke for several seconds at the end of stroke. Also,
pull the stroke as steadily as possible so that each stroke is 10.4.3.3 Method 5 or ISO-9096
about 3–4 s. Every 10th sample, the sampling tube should
be cleaned and the pump should be purged with several Method 5 is a quantity measurement and requires a spe-
quick strokes by pumping air without the filter paper. cial instrument, that is, a particulate sampling train as
illustrated in Figure 10.37. When performing Method 5,
10.4.3.2 Opacity Method Methods 1, 2, 3, and 4 will be automatically included:
Method 1: sample and velocity traverses for stationary
In the simplest definition, opacity can be defined as the sources; Method 2: determination of stack gas velocity
amount of light blocked by a medium, measured as a per- and volumetric flow rate (Type S pitot tube); Method 3:
centage from 0 to 100. An opacity of zero means that no gas analysis for carbon dioxide, oxygen, excess air, and
light is blocked and all passes through the medium. An dry molecular weight; Method 4: determination of mois-
opacity of 100 means that 100% of the light is blocked. A ture content in stack gas; and Method 5: determination
more rigorous definition often used in opacity meters is66 of particulate emissions from stationary sources.
In the sampling train, the flue gas is sucked into the
Opacity = 1 − e( − Db ) (10.19)
sampling probe by a motor-driven suction pump. The
where flue gas passes through a filter paper, an impinger train,
D is the duct size, length of light beam control valves, and a dry gas meter. The particulates
b is the light extinction coefficient of the medium pro- are collected in the filter paper and some of them may
portional to the volumetric portion of the particles deposit on the sampling line, which requires collection
in the flow also. The moisture is collected in the impingers.
The reason that Method 5 procedures are so compli-
The earliest attempt to measure opacity was devel- cated is the requirement for isokinetic sampling. An
oped by Maximilian Ringelmann in Paris. In the late isokinetic condition is where the velocity entering the
Temperature sensor
Impinger train optional, may be replaced

Probe by an equivalent condenser
Temperature
Temperature
Type S pitot sensor
sensor
tube
Temperature
sensor Heat traced
Gooseneck Impingers
glass-lined
nozzle
probe
Check
valve
Glass filter
holder
Ice
Type S pitot Heated area water bath Vacuum
tube Stack line
wall Manometer Temperature
Water Empty Silica gel
sensors
Vacuum
gauge
Orifice
By-pass Main
valve valve
Dry gas
meter
Air tight
pump
FIGURE 10.37
Particulate sampling train.
sampling probe is the same as the velocity of the com- in sampling streamlines represented by the dashed lines.
bustion gas in the stack. Isokinetic sampling is the only Small particles follow the streamlines and escape from
way to get true particulate concentration from particu- the nozzle. The large particles go straight into the nozzle.
late-suspended fluid flow.71–75 When performing iso- Therefore, the nozzle collects more large particles and less
kinetic sampling, a sample is taken without distorting small particles than the environmental stream. The mea-
the streamlines. If the sample rate is too low or too high sured particulate concentration, therefore, is higher than
with respect to the environmental gas velocity, errors will the true value. For isokinetic variation larger than 100%,
result in the particulate concentration measurements. as shown in Figure 10.38c, the situation becomes reversed.
A simple schematic, as illustrated in Figure 10.38, dem- The nozzle collects more small particles and less large par-
onstrates the deviation of sampling concentrations at dif- ticles than the environmental gas. The particulate concen-
ferent isokinetic variations. Isokinetic variation is the ratio tration is thus lower than the true value.
of sampling nozzle suction velocity to the environmental To achieve isokinetic sampling, the stack gas veloc-
gas velocity. At 100% isokinetic variation as shown in ity needs to be determined first. To measure stack
Figure 10.38a, the nozzle sampling velocity is the same as velocity accurately, a complicated instrument as indi-
the environmental gas velocity. A true particulate concen- cated in Figure 10.37 is required. Before performing
tration is obtained. If isokinetic variation is less than 100%, any measurements, those instruments have to be care-
as indicated in Figure 10.38b, the nozzle sampling veloc- fully calibrated. This involves tedious preparation
ity is lower than the environmental gas velocity, resulting and calibration procedures. The whole procedure for
to obtain a representative particulate concentration of the

exhaust gas, the sampling probe must be moved to vari-
ous points across the stack. Method 1 provides a guide-
line for selecting the representative measuring points
Flow Nozzle across the stack.
Every stack has its own structure and the location of
the sampling opening is different for different furnaces.
Method 1 suggests that the number of traverse points in
the stack should follow the guideline, as shown in Figure
10.39. As for the location of traverse points, it is determined
by Table 10.6. When any of the traverse points fall within
1 in. of the stack walls, relocate them away from the stack
(a) walls to (1) a distance of 1 in. (2.5 cm); or (2) a distance equal
to the inside diameter of the nozzle, whichever is larger.
At each sampling point, the isokinetic condition must
be maintained. Method 2 describes the techniques for
determining the stack gas velocity and volumetric flow
rate. The measurements in Method 2 are necessary for
Nozzle the determination of the nozzle diameter and the suc-
Flow tion rate of an isokinetic sample probe.
Methods 3 and 4 describe how to measure the molecu
lar weight and the moisture content of the stack gas, which
are required for determining the pressure drop in a pitot
tube. The pitot tube is an instrument used to measure the
gas velocity based on Bernoulli’s principle. Bernoulli’s
principle states that the static pressure in a fluid decreases
(b) as the velocity of the fluid increases. Bernoulli’s equation
can be found in any fluid dynamics textbook.76
Due to the particulate contents in the stack gas, a
special pitot tube, a Type S as shown in Figure 10.40, is
usually used. The Type S pitot tube has several advan-
tages over the standard type pitot tube. It has large sens-
Nozzle ing openings that minimize the chance of plugging in
Flow heavy particulate concentrations, and provides a higher
Δp reading than a standard pitot tube, which is benefi-
cial in measuring low gas velocities.
The Type S pitot tube is not a designated standard
and, therefore, it should be calibrated against a standard
pitot tube, whose calibration factor is known. This is
the first calibration procedure that is required. During
(c) calibration, the Cp value is obtained as indicated in
Equation 10.20. The equation is used to calculate the flue
FIGURE 10.38
Sampling at different isokinetic variations. (a) 100% isokinetic sam-
gas velocity from pitot tube pressure drop reading (Δp):
pling, (b) less than 100% isokinetic sampling, and (c) more than 100%
Ts ∆ p
isokinetic sampling.
Vs = K pCp (10.20)
Ps Ms
Method 5 includes three major parts: (1) preparation, where
(2) on-site measurements, and (3) particulate recovery. Vs is the flue gas velocity (ft/s)
Kp is the unit factor = 85.49 ft/s ((lb/lb mole) (in Hg)/
10.4.3.3.1 Preparation (oR)(in H2O))1/2
Preparation includes (1) determination of sampling points; Cp is the pitot tube coefficient
(2) instrument calibration, and (3) filter paper preparation. Ts is the stack gas temperature (oR)
When combustion gas exits out of the stack, the velocity Δp = velocity head of the stack gas (in H2O)
distribution inside the stack is not uniform and its bulk Ps is the absolute stack gas pressure (in Hg)
velocity varies with the heat loads in the furnace. In order Ms is the stack gas molecular weight (lb/lb mole)
Duct diameters upstream from flow disturbance* (Distance A)
0.5 1.0 1.5 2.0 2.5

50
Disturbance
40
Minimum number of traverse points
A
Measurement
site
30 B
24 or 25a
Disturbance
20
20
16
Stack diameter > 0.61 m (24 in.)
12
10
8 or 9a
Stack diameter = 0.30–0.61 m (12–24 in.)

0
2 3 4 5 6 7 8 9 10
Duct diameters downstream from flow disturbance* (Distance B)
FIGURE 10.39
Minimum number of traverse points for particulate traverses. aHigher number is for rectangular stacks or ducts. *From point of any type of
disturbance (bend, expansion, contraction, etc.).
TABLE 10.6
Location of Traverse Points in Circular Stacks
Number of Traverse Points on a Diameter
Traverse Point on
a Diameter 2 4 6 8 10 12 14 16 18 20 22 24
1 14.6 6.7 4.4 3.2 2.6 2.1 1.8 1.6 1.4 1.3 1.1 1.1
2 85.4 25.0 14.6 10.5 8.2 6.7 5.7 4.9 4.4 3.9 3.5 3.2
3 75.0 29.6 19.4 14.6 11.8 9.9 8.5 7.5 6.7 6.0 5.5
4 93.3 70.4 32.3 22.6 17.7 14.6 12.5 10.9 9.7 8.7 7.9
5 85.4 67.7 34.2 25.0 20.1 16.9 14.6 12.9 11.6 10.5
6 95.6 80.6 65.8 35.6 26.9 22.0 18.8 16.5 14.6 13.2
7 89.5 77.4 64.4 36.6 28.3 23.6 20.4 18.0 16.1
8 96.8 85.4 75.0 63.4 37.5 29.6 25.0 21.8 19.4
9 91.8 82.3 73.1 62.5 38.2 30.6 26.2 23.0
10 97.4 88.2 79.9 71.7 61.8 38.8 31.5 27.2
11 93.3 85.4 78.0 70.4 61.2 39.3 32.3
12 97.9 90.1 83.1 76.4 69.4 60.7 39.8
13 94.3 87.5 81.2 75.0 68.5 60.2
14 98.2 91.5 85.4 79.6 73.8 67.7
15 95.1 89.1 83.5 78.2 72.8
16 98.4 92.5 87.1 82.0 77.0
17 95.6 90.3 85.4 80.6
18 98.6 93.3 88.4 83.9
19 96.1 91.3 86.8
20 98.7 94.0 89.5
21 96.5 92.1
22 98.9 94.5
23 96.8
24 99.9
Source: Baukal, C.E., Industrial Combustion Pollution and Control, Marcel Dekker, New York, 2004.
Dt Type S pitot tube
X ≥ 1.90 cm (¾ in.) for D = 1.3 cm (½ in.)
Sample nozzle Dn
(a)
Sample nozzle
Sampling Static pressure opening
probe plane
Dt Impact pressure
opening plane
Nozzle
Type S entry
pitot tube plane
(b)
FIGURE 10.40
Type S pitot tube and manometer assembly. (a) Bottom view; showing minimum pitot tube–nozzle separation. (b) Side view; to prevent pitot
tube from interfering with gas flow streamlines approaching the nozzle. The impact pressure opening plane of the pitot tube shall be even
with or above the nozzle entry plane.
The other calibration involved is the orifice meter. To flow rate have to be measured so that an isokinetic condi-
measure the suction pump volumetric flow rate, Qm, an tion can be set. Before measuring, the sampling line needs
orifice meter is installed to indicate the flow rate, based to be leak-checked by blowing through one leg opening of
on the following equation: the Type S pitot tube until 7.6 cm (3 in.) water column pres-
sure shows on the manometer. Then, the opening is closed
Tm ∆ H off. The pressure has to be stable at least for 15 s.
Qm = K m (10.21)
Pm Mm Based on the number of traverse points that were
where obtained from the aforementioned preparation procedure,
Km is the orifice meter calibration factor the pitot tube is moved across all the traverse points and
Tm is the gas temperature the Δp readings on the manometer are recorded. The stack
ΔH is the pressure differential across orifice (in H2O) gas velocity is calculated based on Equation 10.20 and the
Mm is the molecular weight of gas passing through stack gas volume flow rate is corrected to the standard con-
meter (lb/lb mole) dition, which is 68°F (20°C) and 29.92 in. Hg (.03989 bar).
Since the suction pump in the sampling train has a lim-
The Km value is unknown and obtained by calibrating ited range of volumetric variation, a correct probe nozzle
through different suction volume flow rates, Qm. diameter has to be selected to attain isokinetic velocity. The
After the instruments have been calibrated, the filter probe nozzle diameter, Dn, can be estimated as follows:
paper should be prepared. The filter paper should be
desiccated for at least 24 h or be dried in an oven for 0.0358 Qm Pm Ts Ms
2–3 h. The dried filter paper needs to be weighed to a Dn = (10.22)
TmC p (1 − Bws ) Ps ∆ pave
precision of the nearest 0.1 mg.
where
10.4.3.3.2 On-Site Measurements Dn is the nozzle diameter (in.)
The furnace should maintain a steady heat load. Before Qm is the volumetric flow rate of suction pump (cfm)
sampling particulates, the stack gas velocity and volume Pm is the absolute meter pressure (in Hg)
Tm is the dry gas meter temperature (°R) then evaporated to dryness at ambient temperature and
Bws is the moisture content of the stack gas pressure, and desiccated for 24 h before weighing.
Ms is the gas molecular weight The silicone gel and water impingers are weighed before
Ps is the gas pressure and after sampling. The weight difference is used for calcu-
Δpave is the average velocity head of the stack gas across lating stack gas moisture content. The dry gas volume flow
the stack (in H2O) rate is required when evaluating the isokinetic variation.
As described earlier, if the experiment is 100% iso-
After the nozzle size is determined, the meter pump kinetic variation, the measurement is the true particulate
rate reading by ΔH is varied with the pitot tube reading concentration. If isokinetic variation is less than 100%, the
of Δp. There is a linear relationship between these two measurement is larger than the true isokinetic value. If
readings: isokinetic variation is larger than 100%, the measurement
is less than the true isokinetic value. Usually, when the
∆ H = K∆ p (10.23)
isokinetic variation is within ±20%, the particulate mea-
The factor K is expressed as follows: surements deviation is around 10%. If isokinetic varia-
tion is off too much, the measurement is not accurate.
Performing Method 5 measurements requires correct
2 Mm Tm Ps
K = 846.72Dn4 ∆ H @Cp2 (1 − Bws ) (10.24) instrument calibrations, rigorous leakage checking, and
Ms Ts Pm accurate calculations. It is not an easy task. It is highly
recommended to have trained professional personnel to
where perform the test for accurate results.
0.9244
∆H @ = (10.25)
K m2
The value of Km is established from Equation 10.21 when

10.5 Carbon Dioxide
calibrating the orifice meter.
The sampling instrument usually has several differ- CO2 is a colorless, odorless, inert gas that does not support
ent nozzle sizes. The nozzle that is closest to the estima- life since it can displace oxygen and act as an asphyxiant.
tion from Equation 10.22 is chosen. The preweighted filter CO2 is found naturally in the atmosphere at concentrations
paper is then installed into the filter holder. Approximately, averaging 0.03% or 300 ppmv. Concentrations of 3%–6% can
100 mL (6.1 in3) water is filled into the first two impingers cause headaches, dyspnea, and perspiration. Concentrations
and 200–300 g (0.44–0.66 lb) silica gel is put into the fourth of 6%–10% can cause headaches, tremors, visual distur-
impinger. The third impinger is left empty. The water bance, and unconsciousness. Concentrations above 10% can
impingers and silica gel impinger are all weighted. cause unconsciousness eventually leading to death.
The instrument needs to be checked for leakage and Some studies indicate that CO2 is a greenhouse gas
proper function. The ice is then placed into the con- that contributes to global warming. A greenhouse gas is
denser box of the sampling train. The probe and filter a gas in the atmosphere that absorbs and emits radiation
have to be maintained within a fairly narrow operating within the thermal infrared range. This process is the
temperature, that is, 248°F ± 25°F (120°C ± 14°C). fundamental cause of the greenhouse effect.
The probing time for each traverse point has to be at
least 2.5 min. If the number of traverse points is low,
10.5.1 CO2 Generation
the probing time at each traverse point should be lon-
ger. The total sample volume should be at least 30 dscf There are a number of different anthropogenic sources
(0.85 dscm), where d represents dry gas, or total sam- of CO2 emissions. Predominately, the emissions are from
pling time should be at least 60 min. the combustion of fossil fuels. CO2 is produced when a
fuel containing carbon is combusted. The International
10.4.3.3.3 Particulate Recovery Energy Agency (IEA)77 has estimated the CO2 emissions
The procedure for recovering particulates is important. from various sources worldwide in the year 2000. They
The particulates should be recovered not only from the reported that emissions of CO2 from fossil fuels globally
filter paper, but also from the filter holder and fittings. totaled about 23.5 × 109 ton CO2 per year. Of this, close to
All the particulates and filter paper are put in a con- 60% were attributed to large stationary emissions sources,
tainer to be oven dried and weighed. such as power generation, cement production, refineries,
Other particulates that are on the probe nozzle, lines, iron and steel industry, petrochemical industry, and oil
and fittings are washed with acetone. The washed ace- and gas processing. Therefore, reduction of CO2 emis-
tone is collected in the other container. The container is sions is mainly targeted at those stationary sources.
10.5.2 CO2 Capture in Texas for enhanced oil recovery (EOR). For acid gas
applications, the CO2 waste stream could be reinjected to
Many schemes have been suggested for reduction of
the gas wells. Deep injection of liquid wastes, and acid
CO2 emissions including energy efficiency improve-
gas disposal (mixtures of CO2 and H2S) have been con-
ment, fuel switching, nuclear power, solar power, wind
ducted in Canada and the United States since 1990.
power, renewable sources of energy, and CO2 capture
In some locations, transport of CO2 by ship may
and storage (CCS). In this section, only CCS is discussed
be more economically attractive, particularly when
since this technology is mainly applied to large indus-
CO2 has to be moved over a long distance or over-
trial stationary CO2 production sources.
seas. According to one source,79 marine transporta-
Broadly, there are three types of technologies for
tion is more cost-effective than pipelines for distances
CCS: post-combustion, pre-combustion, and oxy/fuel
greater than 1000 km (620 mi) and amounts smaller
combustion.
than a few million tons per annum. Currently, CO2
In the post-combustion capture, CO2 is removed after
transported by ship takes place on a small scale
the combustion of a carbon-containing fuel. Instead of
mainly because of limited demand. CO2 transported
being discharged directly to the atmosphere, the flue gas
by ship is held at a pressure of 0.7 MPa (101 psig).
is passed through equipment that separates most of the
Road and rail tankers also are technically feasible
CO2. The separation is achieved by passing the flue gas in
options. These systems transport CO2 at temperatures
close contact with a liquid absorbent or solid sorbent that
of −20°C (–4°F) and pressures of 2 MPa (290 psig).
is capable of capturing CO2. Other schemes such as sepa-
However, they are uneconomical compared to pipe-
ration with membranes or by cryogenic distillation are
lines and ships, except on a very small scale.
also used. These technologies are well developed, adapted
mainly from the acid gas removal process. The acid gas
removal process is to remove hydrogen sulfide (H2S), car- 10.5.4 CO2 Storage
bon dioxide (CO2), or similar contaminants in raw natural
gas or other gas mixtures in natural gas industries. Most large-scale CO2 storage is underground. Four
For precombustion technology, the fossil fuel is partially options exist for geological storage: (1) use of depleted
oxidized to generate syngas (CO and H2). The syngas is oil and gas reservoirs; (2) use in enhanced coal bed
then reacted with steam in a catalytic reactor to convert methane (ECBM) recovery; (3) injection in deep saline
CO into CO2 and generate more H2. After that, the CO2 formations; and (4) use in enhanced oil and gas recovery
is separated, usually by a physical or chemical absorp- (EOR). In each case, CO2 is injected in dense form.
tion process, resulting in a hydrogen-rich fuel. Compared Geological storage of CO2 is ongoing in three indus-
to post-combustion technology, the resulting CO2 in this trial-scale projects: the Sleipner project in the North Sea,
technology can be captured in a relatively pure stream. the Weyburn project in Canada, and the In Salah project
The other way to obtain pure CO2 stream is oxy/fuel in Algeria. As for EOR, it is mostly in Texas, USA, where
combustion.78 The fuel is burned with pure oxygen instead EOR commenced in the early 1970s. Much of the CO2
of air, resulting in a flue gas that is mainly CO2 and H2O. If injected for EOR is produced with the oil, from which it
fuel is combusted using pure oxygen, the flame tempera- is separated and then reinjected.
ture will be extremely high. The CO2 and/or H2O-rich flue
gas can be recirculated to the combustion chamber to lower
the flame temperature. The advantage of oxy/fuel combus- 10.5.5 CO2 Usage
tion is little NOx emissions, if any, due to no nitrogen in the
Except for large quantities of CO2 used for enhanced
combustion process (assuming little if any N2 in the fuel and
oil recovery (EOR), a large portion of CO2 recovered
air leakage into the combustor). The detailed discussion on
is used to make chemicals. Industrial usage for CO2
CO2 capture technologies can be found in Ref. [79].
includes chemical production—urea, refrigerants,
inert agents in food packaging, carbonation of bever-
ages, industrial gas welding systems, fire extinguish-
10.5.3 CO2 Transport
ers, water treatment, calcium carbonate precipitation
After CO2 is collected, it needs to be transported to a stor- (paper industry), and many other small-scale appli-
age area. Two methods are currently in use for transport cations. Currently, the total industrial use is approxi-
of CO2: pipeline and motorized transport (ship, road, mately 120 MtCO2 per year worldwide, excluding use
and rail tankers). Pipelines are more common. The CO2 is for EOR. This amount is small compared to the emis-
typically compressed to above 8 MPa (1160 psig) at ambi- sions from major sources generated by human activi-
ent temperature to avoid two-phase flow. In the United ties, that is, 13,500 MtCO2 per year.79 Therefore, global
States, over 30 × 106 ton CO2 per year are sent through CO2 reduction, such as alternative energy usage, is
more than 2500 km (1500 miles) of pipeline to the sites necessary.
collaboration.86 This collaboration produced fundamen-

tal knowledge and phenomenological understanding in
10.6 SO x
two important areas, one basic and one applied:
Sulfur oxides, usually referred to as SOx, include SO,
S2O, SnO, SO2, SO3, and SO4 of which SO2 and SO3 are • A flame is an extraordinarily effective reactor.
of particular importance in combustion processes.80 SO2 From a basic standpoint, the program eluci-
tends to be preferred at higher temperatures while SO3 dated why and how the hot, rich diffusive zones
is more preferred at lower temperatures.81 Since most are prolific manufacturers of a myriad of reac-
combustion processes occur at high temperatures, SO2 tion intermediates from the light, partially oxy-
is the more predominant form of SOx emitted from sys- genated species through the heavy polycyclic
tems containing sulfur. For combustion at low excess air aromatic hydrocarbons.
levels, more SO2 is produced and less SO3 is produced.82 • Emissions of products of incomplete combustion
SO2 is a colorless gas with a pungent odor that is used (PIC) from typical petroleum industry burners
in a variety of chemical processes. SO2 can be very cor- are extremely low. From the applied standpoint,
rosive in the presence of water. SO2 is considered a pol- the program elucidated why and how the highly
lutant because of the choking effect it can cause on the reactive diffusive jets of typical petroleum indus-
human respiratory system. It is also damaging to green try burners are extremely effective in destroy-
plants, which are more sensitive to SO2 than people and ing the myriad of reaction intermediates that are
animals. When SO2 is released into the atmosphere, it manufactured in the hot, rich zones.
can produce acid rain by combining with water to pro-
duce sulfuric acid (H2SO4). Sulfuric acid is very corrosive This section discusses that study on hazardous air
and can cause considerable damage to the environment. pollutants (HAPs) in the refinery, chemical, and petro-
It is often assumed that any sulfur in a combustor will chemical industries. Some relevant references on HAPs
be converted to SO2 that will then be carried out with the are given in.87–112
exhaust gases.83 The sulfur may come from the fuel or
from the raw materials used in the production process.
Fuels like heavy oil and coal generally contain signifi- 10.7.1 Experimental Setup
cant amounts of sulfur while gaseous fuels like natural 10.7.1.1 Experimental Facility
gas tend to contain little if any sulfur. The two strate-
gies for minimizing or eliminating SOx are (1) removing The experimental facility in which full-scale petroleum
the sulfur from the incoming fuel or raw materials, and industry burners were tested is illustrated in Figure 10.41.
(2) removing the SOx from the exhaust stream using a The Burner Engineering Research Laboratory (BERL)
variety of dry and wet scrubbing techniques.84 One dry located at the Combustion Research Facility, Sandia
scrubbing technique is limestone injection. After use,
the combined limestone and sulfur can be used in gyp-
sum board. New membrane separation technologies are Exhaust Convection section
another reduction technique being developed. simulator (CSS)
Furnace exit
Exhaust
Furnace Sample 800°F

10.7 Hazardous Air Pollutants
In the United States, the 1990 Amendments to the
Clean Air Act promulgated new regulations by the
U.S. Environmental Protection Agency on emissions of Air and
hazardous air pollutants from the process heaters and fuel system
industrial boilers used in the petroleum, petrochemical,
and chemical sectors.85 Unfortunately, the science upon
which to base those regulations was extremely limited
and the paucity of field data then extant was severely
flawed. To address those deficiencies, a 4-year $7-mil-
lion fundamental study on the origin and fate of trace
emissions from gaseous hydrocarbon external combus- FIGURE 10.41
tion was initiated by a government–university–industry BERL experimental facility. (Courtesy of Sandia National Labs.)
National Laboratories (Livermore, California), provided a CSS, was added to BERL. Instead of an actual convec-
controlled, stable, and well-characterized operating con- tion section, the project team decided to construct a CSS.
ditions for full-scale burners. The BERL also featured The CSS closely matched the temperature–time profiles of
laser diagnostic and air emissions testing capabilities petroleum refinery process heaters over a range of mean
that allowed complete characterization of industrial cooling rates. Mixing was accomplished by adding a series
burner flames and emissions. of steps and meters to the walls of the CSS, which created
The experimental rig accommodated up to a 900 kW (3 × turbulence in the gas flow. Catalytic effects between the
106 Btu/h) full-scale burner. It consisted of two distinct sec- gas stream and the CSS were prevented by lining all parts
tions: the furnace section and the convection section simu- of the CSS exposed to the gas stream with quartz. The CSS
lator (CSS) that is illustrated in Figure 10.42. The walls of is illustrated in Figure 10.42. Sample locations were chosen
the furnace section were water-cooled for radiative extrac- to ensure that the sample obtained represented the flue gas
tion of heat from the combustion gases. From the furnace exhaust stream and desired sample conditions. Flue gas
section, the combustion gases flowed to the quartz-lined slipstreams were pulled from BERL both before and after
CSS, which was designed to duplicate the exit temperature the CSS. This combined approach enabled the flexibility to
profile seen in full-scale process heaters. study independently the effects of both flame character-
While BERL was optimized for the study of flame struc- istic variability and exhaust temperature/residence time
ture, petroleum refinery process heaters are optimized for variability on products of incomplete combustion (PICs)
efficiency and economy. Because of this, a petroleum refin- formation and destruction.
ery process heater employs a convection section for addi-
tional heat extraction before rejecting the combustion gases
10.7.1.2 Full-Scale Burner Tests
to atmosphere, a feature that was originally not needed
in BERL. Since previous kinetics studies indicated that a A conventional diffusion flame burner (CDFB) is
significant amount of reaction chemistry may occur in the illustrated in Figure 10.43. This CDFB has all of the fuel
convection section, a fact that was verified in this study, injection tips in the combustion air stream inside the
Flow control
and metering
Aluminum
Test section
jacket
Zircar
Sample insulation
flow
Coflow Inconel 601

tubing
Sampling ports
Quartz
step and liner
Coflow dilution air
Thermocouple
Coflow passage
FIGURE 10.42
Convection section simulator (CSS). (Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete combustion (PIC) from petroleum,
petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on Combustion By-Products and Their Health
Effects, Ischia, Italy, June 17–20, 2007.)
10.7.2 Experimental Results

10.7.2.1 No Systematic Variation
In this program, it was found that the nature of the gaseous
hydrocarbon fuel mixture did not make much difference,
neither in the total PIC emissions nor in the individual spe-
cies levels. This observation includes natural gas, which is
itself another hydrocarbon mixture; that is, there is no rea-
son to distinguish “refinery fuel gas” from “natural gas.”
Almost stochastic in nature, the individual species levels
Air Air seemed to be dependent less upon physics and chemis-
try and more upon the vagaries of the sophisticated sam-
pling methods and precise analytical techniques that are
required to detect them at all in the minute concentrations
in which they appear in the combustion products.
In Figures 10.45 through 10.48, these facts are illus-
trated by the mass emission of total hydrocarbons.
Except for an operationally unrealistic superaerated
(450% stoichiometric air) case, there was no significant
Gas effect of heating value, combustion zone stoichiometry,
propylene or ethylene spikes, or hydrogen content.
The heating value variation was achieved at the con-
Pilot gas stant base case 16% hydrogen content by increasing the
proportion of propane in the hydrogen/natural gas/
FIGURE 10.43 propane mixture. The field-operational typical ±15%
Conventional diffusion flame burner (CDFB). theoretical air variation around the base case 125% was
extended substoichiometrically in the combustion zone
burner tile. The near stoichiometric, turbulent diffusion to 50% or one-half of the air theoretically required for
mixing of the fuel and air results in high flame tempera- complete combustion with overfire air added to simu-
tures and high NOx levels. late a leaky furnace while still maintaining the base case
A low-NOx diffusion flame burner (LDFB) is illus- 125% theoretical air in the stack. Additionally, in other
trated in Figure 10.44. This LDFB has both primary and
staged fuel injection tips located in the furnace on the 100
periphery of the burner tile. Both the primary fuel and
staged fuel entrain furnace gases prior to mixing with
air. The inert furnace gases dilute the fuel stream mini- 10–1
mizing prompt NOx and reduce peak flame tempera-
tures minimizing thermal NOx.
10–2
lb/MMBtu
10–3
10–4
Furnace gas
recirculation Primary fuel 10–5
800 1000 1500 2000
LHV (Btu/scf)
Staged fuel Staged fuel
FIGURE 10.45
Fuel inlet Air inlet Conventional diffusion flame burner total hydrocarbon emissions
versus heating value of HC fuel mixture. (Adapted from Seebold,
J.G. and Waibel, R.T., Products of incomplete combustion (PIC)
from petroleum, petrochemical & chemical sector process heat-
FIGURE 10.44 ers and industrial boilers, 10th International Congress on Combustion
Low-NOx diffusion flame burner (LDFB). By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
100 100
10–1 10–1
10–2
lb/MMBtu
10–2
lb/MMBtu
10–3
10–3
10–4
10–4
10–5
0 100 200 300 400 500
10–5
Combustion zone stoichiometric air % 0 15 33
FIGURE 10.46 Hydrogen content %
Conventional diffusion flame burner total hydrocarbon emissions
versus combustion zone stoichiometry. (Adapted from Seebold, FIGURE 10.48
J.G. and Waibel, R.T., Products of incomplete combustion (PIC) Conventional diffusion flame burner total hydrocarbon e missions
from petroleum, petrochemical & chemical sector process heat- versus hydrogen content of HC fuel mixture. (Adapted from Seebold,
ers and industrial boilers, 10th International Congress on Combustion J.G. and Waibel, R.T., Products of incomplete combustion (PIC)
By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.) from petroleum, petrochemical & chemical sector process heat-
ers and industrial boilers, 10th International Congress on Combustion
By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
100
natural gas, and propane while “A4” was 1050 Btu/
10–1 scf natural gas) remained good throughout all of the
CDFB trials in test sequences A, B, and C. While test
sequence A spanned a broad range of fuel compositions
10–2 and operating conditions around the normal operation
lb/MMBtu
base cases Al and A4, there was no systematic variation

10–3 in emissions; all emissions remained exceedingly low
and the small differences were well within the typical
bounds of experimental variability.
10–4
Polycyclic aromatic hydrocarbon (“PAH”) emissions as
high as those reported in some field tests could not be
10–5 reproduced even in the sequence B “failure mode” tests.
10% 1060 Btu/scf 18% In the sequence C “super-failure mode” trials, stack emis-
propylene 13% hydrogen ethylene sions up to 2 × 10−6 lb-PAH/MMBtu (e.g., B13′ in Figures
natural gas, propane
10.49 and 10.50) were measured, as high as any in the “real
FIGURE 10.47 world” field database. These high emissions are often
Conventional diffusion flame burner total hydrocarbon emissions attributed, but without much definition and no detailed
versus propylene and ethylene spikes. (Adapted from Seebold, understanding, to so-called “gross mixing failures.”
J.G. and Waibel, R.T., Products of incomplete combustion (PIC)
from petroleum, petrochemical & chemical sector process heat-
As illustrated in Figures 10.49 and 10.50, to generate
ers and industrial boilers, 10th International Congress on Combustion high stack emissions from gaseous hydrocarbon mix-
By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.) tures in external combustion, severe fuel-air mixing
failures were required, egregious hypothetical extreme
tests to be discussed later, the combustion air was limited combustion conditions that would hardly be tolerated
to less than 50% of theoretical right through and out the nor permitted to persist in any well-run plant.
stack. The theoretical air variation was extended super- To sustain these conclusions, it was important to con-
stoichiometrically in the combustion zone to 450% or 4½ firm that the convection section simulator (“CSS”) did not
times the air theoretically required for complete combus- sequester PAHs nor otherwise produce misleading results.
tion simply by increasing the air delivery to the burner. The suitability of the CSS was confirmed by direct calibra-
The absence of systematic variability in the trace tion runs, injecting known PAH concentrations in an inert,
emissions of PICs in hydrocarbon gaseous external but otherwise identical-condition flow, and confirming by
combustion is illustrated in Figures 10.49 and 10.50. The direct measurement that what went in did, in fact, come
reproducibility of the reference regulatory base cases out. Furthermore, it may be observed that in the actual test
(“A1” was a 1050 Btu/scf mixture of 16% hydrogen, runs in which there was virtually no air in the CSS, the
75% Stoich air + 5% propylene

spike (no air in convection
section simulator)
75% Stoich air
10–5 (no air in convection
section simulator)
50% Stoich air + leakage

10–6 above combustion zone
to 125% (air in convection
section simulator)
450% stoich air
10–7
lb/MMBtu
10–8
10–9
10–10
Regulatory Base cases A2 A4 A6 A8 A10 A12 A11 B2 B7aP B7c B7cP B9 B7 B7c΄B13΄
A1 A4 A3 A5 A7 A9 A11 A13 A12 + B7a B7b B7bP B8 B10 B7c B12 B13
refinery natural Range of normal operation Sequence B Sequence C
fuel gas gas failure modes super-failure
modes
FIGURE 10.49
Conventional diffusion flame burner total PAH at stack outlet. (Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete combus-
tion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on Combustion
10–4 75% Stoich air + 5%

propylene spike
(no air in convection
section simulator)
75% Stoich air
10–5
50% Stoich air + leakage (no air in convection
above combustion zone to 125% section simulator)
(air in convection
lb/MMBtu
section simulator)
10–6
10–7
10–8
A1 A4 B2 B7c B7c΄ B12 B12΄ B13΄
Total PAH Benzo(a)pyrene
FIGURE 10.50
Conventional diffusion flame burner total PAH and benzo(a)pyrene at furnace outlet compared to stack outlet. (Adapted from Seebold, J.G.
and Waibel, R.T., Products of incomplete combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial
boilers, 10th International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
inlet PAHs came through virtually unreacted (e.g., B12, the individual hydrocarbon species that are necessar-
B12′, and B13′ in Figure 10.50), just as they should have done, ily profusely generated in the rich zone would sub-
while in the presence of excess air, which is the usual case sequently quickly be consumed to near extinction, a
in the field, and just as the program’s kinetic modeling led prediction that was seen borne out time and time again
to the expectation, the high CSS-inlet (furnace-outlet) PAH in the full-scale burner trials carried out at the BERL.
concentrations were consumed (e.g., B7C and B7C′ in Figure Most significant, particularly in light of the results of
10.50) before the flue gas emerged into the atmosphere. the last-completed “Sequence E” super-failure mode tri-
als, are the results of Sandia’s application of a two-stage
Lagrangian jet model to a typical conventional diffusion
10.7.2.2 High-Velocity Jet Mixing Produces Low PICs
flame burner jet. Based upon the observed flame struc-
The strong mixing potential of sonic jets is well known. ture of the conventional burner, the jet model was applied
In the case of choked jets exhausted to atmosphere, for twice: first to the individual jet flames that emerge from
instance, it is not uncommon for an initially toxic com- the burner tips inside the quarl and again for the merged
position rapidly to auto-mix down to harmless concen- jet exiting the quarl; thereafter, when mixing is complete,
trations. In the case of the multiple, small reacting jets of a plug-flow reactor model was utilized to represent the
the conventional diffusion flame burner, surrounded as remaining flow to the furnace exit. To give confidence in
they are (under normal conditions) with an excess sup- the results, it may be observed that the model predicted
ply of oxidant (air), similar behavior would be expected. a final CO level of 2 ppm, consistent with the measured
Early in the program, it was hypothesized that the level below the detection limit of 5 ppm, and a final NOx
hot, rich combustion regions that are necessarily pres- concentration of 106 ppm, compared with the measured
ent in a diffusion flame ought to be prolific generators of value of 118 ppm. As shown in Figure 10.51, the jet model
individual hydrocarbon species in all possible elemental predicted that individual hydrocarbon species should be
molecular combinations. The early stirred-reactor, plug- produced to significant levels within the in-quarl flames,
flow computations carried out by Lawrence Livermore but should be consumed well within the substoichiomet-
National Laboratory supported the hypothesis, while ric regime, both just as was shown.
later the laboratory flame measurements carried out at In Figure 10.51, “Theoretical air (%)” may also be inter-
the UCLA Chemical Engineering Laboratory, as well preted as “x” or length along the axis of the jet. Initially,
as the research furnace experiments carried out at the where the reactions are just beginning, there is nothing
BERL, confirmed it. The Lawrence Livermore calcula- but the original fuel reactants and oxidant in abundance.
tions also suggested that, in the presence of excess air, As the reactants and oxidant begin to mix, the reacting
∑ benzene, ethylbenzene, toluene,

102 xylene, naphthalene, phenanthrene
101
100
ppm
10–1
10–2
10–3
10–4
25 30 35 40 45 50 55 60 65 70 75 80 85
Theoretical air (%)
FIGURE 10.51
Lagrangian jet model predictions. (Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete combustion (PIC) from petroleum,
petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on Combustion By-Products and Their Health
Effects, Ischia, Italy, June 17–20, 2007.)
part of the “reacting jet” begins, too, and the reaction mixing potential might be compromised, as, for exam-
products start to appear. Then, as more and more air is ple, in the case of extreme turndown and its attendant
mixed into the jet, with theoretical air% increasing, but low-velocity mixing or, especially, in the presence of liq-
still well within the substoichiometric regime, the reac- uid droplets that may not even vaporize prior to escap-
tion products peak and then are rapidly consumed even ing the intended gaseous “mixing zone,” might higher
before the mixture reaches stoichiometric. levels of PICs be expected to escape to the atmosphere.
The prediction confirmed that PIC species, manufac-
tured in abundance in the hot, rich, diffusive regime, are
subsequently consumed in the high-mixing-potential jet 10.7.2.3 Turndown versus Mixing Rate
well before it reaches even stoichiometric conditions. This In Figure 10.52, some results are shown of experimenta-
is extremely significant with regard to its implication upon tion with a real-time total PAH meter carried out at sev-
the robustness of practical combustion systems in the eral burner firing rates in the BERL. The expected sharp
field. Moreover, it is perhaps remarkable to note that, in the drop-off with increasing theoretical air can be seen, but
“Sequence E” super-failure mode full-scale trials carried well within the substoichiometric regime, is evidently
out at the BERL, it was not until severely substoichiomet- confirmed and that the drop-off “knee” was sharply
ric conditions (stoichiometric ratio < 0.80) were achieved in defined by this pioneering research effort. Moreover, it
the combustion zone and maintained right through and can be seen that the knee shifts toward higher theoretical
out the stack to the atmosphere, and just as predicted by air as the jet mixing potential declines at reduced burner
the Lagrangian jet model, that high levels of PICs emerged. firing rates (reduced gas pressure reduces the momen-
This helps explain why the conventional diffusion tum flux of the fuel jets), just as would be expected and
flame burner, composed as it is of burner tips out of perhaps signaling the breakdown of the burner’s indi-
which there emerge high-mixing-potential jets sur- vidual reacting jets. Perhaps it is not surprising, there-
rounded by an abundant supply of oxidant, simply has fore, that similar trends have been reported (viz., higher
to be a low-PICs burner. Perhaps only when this high PICs under turndown conditions) in some field tests.
102
101
100
1.33 MM Btu/h
2.00 MM Btu/h
pA
10–1
0.67 MM Btu/h
10–2
10–3
70 75 80 85 90 95 100
Theoretical air (%)
FIGURE 10.52
Conventional diffusion flame burner photoionization current (pA) versus theoretical air (%). (Adapted from Seebold, J.G. and Waibel, R.T.,
Products of incomplete combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th
International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
10.7.2.4 BERL-Field Connection and, in contrast to earlier industry testing, this very high-
quality field campaign was conducted with suitably low
Shown superimposed on the charts in Figures 10.53 and
detection limits.
10.54 is the range of measurements by one company on
The comparison shown in Figures 10.53 and 10.54
five process heaters of markedly different configuration
illustrates what has been suggested elsewhere; that
firing refinery fuel gas, one with air-staged low-NOx burn-
is, that the high-mixing-potential reacting jets make
ers, two with fuel-staged ultralow-NOx burners, one with
petroleum industry burners firing gaseous hydro-
raw gas burners, and one with premix burners. Like the
carbon fuels highly effective low-PICs burners; that
proprietary field measurements made by some CRADA-
detailed differences among burners, process heater
signatory companies, the 1993–1994 source testing results
designs and gaseous hydrocarbon fuel compositions
that were commissioned by the non-CRADA-signatory
are unimportant; and that the connection between the
company were obtained following the general recogni-
full-scale BERL trials and high-quality field measure-
tion of the poor quality of the ca. 1989–1990 Western States
ments is strong indeed. Field measurement campaigns
Petroleum Association (“WSPA”) Pooled Source Testing
with adequately low limits of detection in which “non-
campaign. The light hydrocarbons for which PERF 92-19
detects” do not play any substantial role in determin-
sampled in support of the chemical kinetic modeling
ing the reported “emissions” produce, for gaseous
aspects of the program were not reported in the WSPA
hydrocarbon external combustion, substantially the
tests and probably not sought, there being no reason to
same results as were obtained at full-scale by this pro-
do so, but all of the compounds of regulatory importance
gram in the BERL.
were; that is, “aldehydes” (formaldehyde, acetaldehyde);
In connection with the droplet hypothesis and the pro-
“VOC” (benzene, toluene, ethylbenzene, xylenes); and
duction of higher than expected POM emissions, and also
“POM” (naphthalene through dibenz(a,h)anthracene)
10–1
10–2
10–3
10–4
lb/MMBtu
10–5
10–6
10–7
10–8
10–9
Sequence Sequence Sequence Sequence
A B A B A B A B C
–10
10
Aldehydes Light VOC Heavy VOC PAH

Range and average of emissions at the stack outlet
Sequence A = range of normal operation, B = failure modes, C = super-failure modes
Superimposed: range of measurements on five refinery process heaters firing process gas
FIGURE 10.53
Range and average of emissions at the stack outlet for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and Waibel,
R.T., Products of incomplete combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th
10–1
10–2
10–3
10–4
lb/MMBtu
10–5
10–6
Total
Acetylene
Formalde
Butadiene
Propylene
Butane
10–7
Allene
Propyne
Propane
Acetalde
Ethane
Ethylene
Toluene
Benzene
Xylenes
Ebenz
Pah
10–8
10–9
Range of measurements of HAPs at stack outlet

Normal operating range (sequence A) and failure modes (sequence B)
FIGURE 10.54
Range of measurements of HAPs at the stack outlet for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and Waibel,
of interest in relation to the “BERL-Field Connection” is 10.7.2.6 No Effect of Burner Type

an account of the determination of emissions levels of
The exhaustive exploration of the combustion parameter
hazardous air pollutants from the combustion of a #2 oil,
space with the CDFB demonstrated the robustness of
a #5 oil, a low-sulfur #6 oil and a high-sulfur #6 oil in
hydrocarbon gaseous external combustion in producing
a full-scale firetube industrial boiler.113 The emissions of
ultralow-PIC emissions. This was even when subjected to
aldehydes and VOCs for the various oils compared favor-
extreme mixing failures imposed by, for example, a par-
ably with those obtained during the full-scale gaseous
tially blocked quarl, plugged or reversed burner tips, or
combustion burner trials carried out under the PERF
even air starvation with overfire leakage air. This suggested
92-19 CRADA at the BERL, but the POM emissions were
the ultralow-NOx diffusion flame burner (“LDFB”) would
higher by about an order of magnitude.
prove also to be an ultralow-PICs burner. Indeed, because
of the more intense mixing in the LDFB as compared with
the CDFB owing to internal self-recirculation, it was pre-
10.7.2.5 Refinery Fuel Gas, Natural Gas Equivalency
dicted that the trace PIC emissions might become a little
Elsewhere it has been shown, for a broad range of com- “tracer,” as it were, and they did. The lack of any important
bustion conditions, the total hydrocarbon emissions impact on trace PIC emissions of greatly altered mixing
resulting from the external combustion of natural gas owing to changing the burner type is illustrated in Figure
and a variety of refinery fuel gas mixtures in which 10.56 for the normal operation regulatory base cases. In the
it is easily seen that there is no reason to distinguish hypothetical extreme failure-mode tests, a similar lack of
refinery fuel gas from natural gas, the latter being itself significant impact on PIC emissions was seen.
just another mixture of hydrocarbon gases. Figures
10.55 and 10.56 illustrate the same equivalency on but
10.7.2.7 Detection Limits
a speciated basis, there being only small, statistically
insignificant differences in the individual specie’s “Nondetects” recorded at high limits of detection are
emissions. worse than useless. Thus, at the inception of this program,
10–3
10–4
10–5
lb/MMBtu
Acetaldehyde
10–6
Formaldehyde
Total
Propyne
Butadiene
Propylene
Propane
Other
Ethane
Ethylene
Butane
Acetylne
Allene
10–7
Xylenes
Benzene
Ethbenz
Toluene
Pah
10–8
10–9
Refinery fuel gas

Operating condition A1: regulatory base case
FIGURE 10.55
Emissions for refinery fuel gas (16% H2, propane, natural gas) for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and
Waibel, R.T., Products of incomplete combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boil-
ers, 10th International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
10–3
10–4
10–5
lb/MMBtu
Formaldehyde
10–6
Acetaldehyde
Total
Propane
Propylene
Ethane
Other
Butane
Ethylene
Butadiene
Acetylne
10–7
Propyne
Xylenes
Allene
Toluene
Benzene
Ethbenz
Pah
10–8
10–9
Natural gas
Operating condition A4: regulatory base case
FIGURE 10.56
Range of emissions for natural gas and refinery fuel gas for the conventional diffusion flame burner and the ultralow-NOx diffusion burner.
(Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete combustion (PIC) from petroleum, petrochemical & chemical sector pro-
cess heaters and industrial boilers, 10th International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
in part due to the well-known inadequacies of what few to gain useful data. In this regard, it is worth noting that
field testing programs had been carried out, it was con- the emission concentrations in Figure 10.56 are expressed
cluded that QA/QC was of the utmost importance and, in parts per quadrillion and that to reliably detect the
in particular, that detection limits much lower than those exceedingly low levels of polycyclic organic matter
that had been achieved in the field tests would be needed (“POM”) that were produced under typical operating
101
100
10–1
lb/MMcf
10–2
10–3
10–4
1 2 1 2 1 2 1 2
ALDE VOC POM Total
FIGURE 10.57
Emission factor comparison for low-NOx burner and conventional burner. (Adapted from Seebold, J.G. and Waibel, R.T., Products of incom-
plete combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on
Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
conditions, detection limits less than 100 parts per quadril- required to be achieved. In the quest for a surrogate for
lion were required to be achieved. This program’s thor- PAH emissions, as may be seen in Figure 10.59, it appears
oughly documented and highly accurate finding of no that benzene (shown blue), at least in the severely sub-
effect of burner type, conventional versus low NOx, nev- stoichiometric regime, has some promise. Figures 10.60
ertheless, seems to be corroborated by the field database, and 10.61 simply serve to illustrate the fact that neither
as illustrated in Figure 10.57. carbon monoxide (“CO”) nor total hydrocarbon (“HC”)
have any promise whatsoever as PAH surrogates.
10.7.2.8 Results of the Final Full-Scale Trials Similar to the deeply substoichiometric trials, prior
highly aerated tests emphasized the trace polycy-
In the final sequence of full-scale trials, through egre- clic organic hydrocarbon emissions but, nevertheless,
gious excursions into the super- and substoichiomet- showed high total hydrocarbon emissions, as well. In
ric combustion regimes, some systematic behavior Sequence E, the reason for rerunning the superstoichio-
(vs. stoichiometric ratio) was finally shown. Perhaps, a metric case (stoichiometric ratio = 3.0) and, as always, the
PAH surrogate (benzene) was found. This answered a regulatory base case (SR = 1.25) for comparison, was to
question left over from previous trials about the ultra- add stack exit aldehyde and heavy volatile organic com-
superstoichiometric hydrocarbon PICs whether they pound measurements in order to determine whether or
were reaction products (yes) or simply unroasted fuel not those high total hydrocarbons were reaction products
fragments (no). It was also “discovered” (yet again) that or merely unburned fuel fragments. In Figure 10.62, the
CO and HC are poor surrogates, at least in the case of hydrocarbon emissions (shown in yellow) are largely
hydrocarbon gaseous external combustion. accounted for by reaction products, mainly aldehydes.
As shown in Figure 10.58, at severely substoichio- Finally, Figure 10.63 illustrates again (with the excep-
metric conditions maintained right through and out tion of the heavy VOCs, which include benzene, ethyl
the stack to the atmosphere, total polycyclic aromatic benzene, xylenes, and toluene) that systematic behavior
hydrocarbon (PAH) emissions 4 rings and greater as (vs. stoichiometric ratio) can be forced by excursions
great as 98 parts per million (ppm) at the furnace exit into the severely super- and substoichiometric regimes.
(“in” = inlet to the convection section simulator) and
2.5 ppm at the stack exit (“out”) were observed, with
10.7.2.9 Summary
corresponding benzo(a)pyrene emissions of 130 parts
per billion (ppb) and 40 ppb. Note that the emission lev- Jet-mixed hydrocarbon gaseous diffusion flames, such
els are expressed in parts per trillion and that detection as those produced by the burners that are typically used
limits in some cases below 100 parts per quadrillion were in petroleum industry process heaters and industrial
108
107
106
105 b(a)p
b(a)p b(a)p
104
ppt
103 b(a)p b(a)p
102
101
b(a)p b(a)p
100
10–1
Out Out In Out In Out In Out In
3.0 1.25 0.80 0.67 0.60 0.44
Stoichiometric ratio
FIGURE 10.58
Total PAH emissions 4 rings and greater versus stoichiometric ratio. (Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete
combustion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on
Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
108
107
106
105
104
ppt
Benzene
103
102
PAH
101
100
10–1
Out Out Out Out Out
3.0 1.25 0.80 0.67 0.60
FIGURE 10.59
Benzene (blue) and PAH (red) emissions versus stoichiometric ratio for the conventional diffusion flame burner. (Adapted from Seebold, J.G.
1011
109
CO
107
ppt
105
103
PAH
101
10–1

3.0 1.25 0.80 0.67 0.60 0.44
FIGURE 10.60
CO (green) and PAH (red) emissions versus stoichiometric ratio for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and
1011
109
HC
107
ppt
105
103
PAH
101
10–1
3.0 1.25 0.80 0.67 0.60 0.44
FIGURE 10.61
HC (yellow) and PAH (red) emissions versus stoichiometric ratio for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and
108
HC
107 ALDE
106
105
HC ALDE
4
VOC
10
ppt
103 VOC
102
101
100 PAH
PAH
–1
10
3.0 1.25
FIGURE 10.62
HC, aldehyde, VOC, and PAH emissions versus stoichiometric ratio for the conventional diffusion flame burner. (Adapted from Seebold, J.G.
108
107
106
105
104
ppt
103
102
101
100
10–1
3.0 1.25 0.80 0.67 0.60
FIGURE 10.63
Total heavy VOC emissions versus stoichiometric ratio for the conventional diffusion flame burner. (Adapted from Seebold, J.G. and Waibel,
boilers, produce exceedingly low PIC emissions. The of hydrocarbon gases; that the individual hydrocar-
nature of the hydrocarbon gaseous fuel mixture makes bon species levels as measured depend mainly upon
little difference, neither in the total emissions nor in the vagaries of the sophisticated sampling methods
the individual species concentrations; that there is no and precise analytical techniques that are required
reason to distinguish between process gas and natural to detect them at all in the minute concentrations in
gas, the latter being itself merely just another mixture which they appear in the combustion products; that,
despite their markedly different mixing patterns, both the PERF 92-19 CRADA program were far more sen-
the conventional diffusion flame burner and today’s sitive than those used in most field tests, the agree-
ultralow-NOx burner are ultralow-PIC burners; that all ment between the PERF and field results nevertheless
of this is true for an exceedingly broad range of hydro- demonstrates the excellent representativeness of the
carbon gaseous combustion conditions; and that high- PERF data for jet-mixed gaseous hydrocarbon exter-
quality field measurement campaigns produce results nal combustion.
that are the same as those obtained in this program Given the excellent agreement of all data sources, in
in the BERL. In short, the jet-mixed hydrocarbon gas- the subsequent analysis referred to earlier, it seemed
eous combustion process is extremely robust, produc- reasonable to combine all data to develop robust emis-
ing ultralow PIC emissions even when subjected to sion factors applicable to all gas-fired industrial boil-
extreme mixing failures. ers and process heaters which, in turn, enabled reliable
characterization of PIC emissions from a typical gas-
fired process heater and a typical petroleum refinery.114
10.7.3 Process Heater, Petroleum
The result is shown in Figure 10.64. Calculated organic
Refinery Emissions Factors
PIC mass emissions in tons per year are shown for a
As indicated in Figures 10.53 and 10.54, competent hypothetical process heater or industrial boiler with a
field measurement campaigns (i.e., those carried out fuel heat input value of 100 MMBtu/h (29 MW), which
with adequately low limits of detection in which “non- is in the same order of magnitude as those commonly
detects” do not play any substantial role in determin- found in petroleum industry facilities such as refiner-
ing the reported “emissions”) produce, for jet-mixed ies and natural gas processing plants. The total heat
gaseous hydrocarbon external combustion, substan- input to gas-fired combustion equipment in a typical oil
tially the same results as were obtained at full-scale refinery is on the order of 5000 MMBtu/h (1500 MW).
by this program in the BERL. Following completion Organic PIC mass emissions from a refinery of this size,
of the PERF 92-19 CRADA Project, a more extensive estimated in the same way, are also shown in Figure
comparison with a broad range of field data was car- 10.64. The equivalent gas concentrations are also shown
ried out that confirmed that the landmark PERF data to emphasize that these mass emission rates correspond
and actual field data from the “WSPA-CATEF” data- to exceedingly low concentrations measured in parts
base compared well.85 While the test methods used in per billion.
101 103
102
100
102
Tons per year—5000 MMBtu/h
Tons per year—100 MMbtu/h
101
10–1
Parts per billion
101
process heater
100
refinery
10–2
10–1 100
10–3
10–2 10–1
10–4
10–3
10–5 10–2
BTX
Formaldehyde
Total PAH
BTX
Formaldehyde
Total PAH
FIGURE 10.64
Typical process heater, petroleum refinery emissions factors. (Adapted from Seebold, J.G. and Waibel, R.T., Products of incomplete combus-
tion (PIC) from petroleum, petrochemical & chemical sector process heaters and industrial boilers, 10th International Congress on Combustion
6. U.S. EPA, AP-42, Compilation of Air Pollutant Emission Factors,

5th Edn., Office of Air Quality Planning and Standards, Office
10.8 Dioxins and Furans of Air and Radiation, U.S. Environmental Protection Agency,
This class of pollutants includes the carbon-hydrogen- Research Triangle Park, NC, January 1995, http://www.
oxygen-halogen compounds and has received con- epa.gov/ttnchie1/ap42/, accessed November 26, 2012.
7. J.C. Reis, Environmental Control in Petroleum Engineering,
siderable attention from both the general public
Gulf Publishing, Houston, TX, 1996.
and from regulatory agencies because of the poten-
8. C.E. Baukal, Industrial Combustion Pollution and Control,
tial health hazards associated with them. Dioxins Marcel Dekker, New York, 2004.
generally refer to polychlorinated dibenzo-p-dioxin 9. American Petroleum Institute Recommended Practice
(PCDD) compounds while furans generally refer to 556, Instrumentation, Control, and Protective Systems for Gas
polychlorinated dibenzo-furan (PCDF) compounds. Fired Heaters, 2nd Edn., API, Washington, DC, April 2011.
Some of the potential health risks include toxicity 10. American National Standards Institute/Am. Soc. Mech.
because of the poisoning effect on cell tissues, carci- Eng., Performance Test Code PTC 19.10, Part 10: Flue and
nogenicity because cancerous growth may be stimu- Exhaust Gas Analyses, American Society of Mechanical
lated, mutagenicity because of possible mutations in Engineers, New York, 1981.
cell structure or function, and teratogenicity because 11. C.E. Baukal, Everything you need to know about NOx,
Met. Finish., 103 (11), 18–24, 2005.
of the potential changes to fetal tissue.56 The over 200
12. C.E. Baukal, NOx 101: A primer on controlling this highly
dioxin/furan compounds are regulated in certain
regulated pollutant, Process Heat., 15 (2), 34–37, 2008.
industries, particularly in waste incineration, and also 13. U.S. Environmental Protection Agency, Nitrogen oxide
in certain geographical locations for a wide range of control for stationary combustion sources, U.S. EPA
applications, especially in Europe. report EPA/625/5-86/020, 1986.
In the vast majority of cases, dioxin/furan emissions 14. E. Culotta and D.E. Khoshland, NO news is good news,
result from some contaminant in the load materials Science, 258 (5050), 1862–1865, 1992.
being heated in the combustor. A quick scan of most of 15. M. Sandell, Putting NOx in a box, Pollut. Eng., 30 (3),
the textbooks on combustion shows that these emissions 56–58, 1998.
are essentially ignored because they are not generally 16. Pollution biology: Acid precipitation effects on veg-
produced in the flame, except in certain limited cases. etation, University of Alberta, http://www.biology.
ualberta.ca/taylor.hp/Biology381/topic-07-Small.htm,
This is primarily because there are not usually any halo-
accessed December 18, 2012.
gens in either the fuel or the oxidizer to produce dioxins
17. Environmental Degradation of Marble, European
or furans. An exception is the case when waste materials Chemistry Thematic Network, www.whatischemistry.
are burned as a fuel by direct injection into a flame. One unina.it/en/martmarmo.html accessed July 27, 2012.
example is the destruction of waste solvents that may be 18. M. Moreton and S. Beal, Controlling NOx emissions,
injected into an incinerator through the burner. Pollut. Eng. Int., Winter, 14–16, 1998.
19. J. Elkins, N. Frank, J. Hemby, D. Mintz, J. Szykman, A. Rush,
T. Fitz-Simons, T. Rao, R. Thompson, E. Wildermann, and
G. Lear, National Air quality and emissions trends report,
1999. Washington, DC: U.S. Environmental Protection
Agency, Report EPA 454/R-01-004, 2001.
20. Y.B. Zeldovich, The oxidation of nitrogen in combustion
References
and explosions, Acta Physicochim. (USSR), 21, 557–628,
1. J.A. Stanislaw, Petroleum industry faces tectonic shifts 1946.
changing global energy map, Oil Gas J., 97 (50), 8–14, 1999. 21. C.E. Baukal (ed.), Oxygen-Enhanced Combustion, CRC
2. U.S. Environmental Protection Agency, Our nation’s Press, Boca Raton, FL, 1998.
air: Status and trends through 2008, report EPA- 22. D.J. MacKinnon, Nitric oxide formation at high tempera-
454/R-09-002, February 2010. tures, J. Air Pollut. Control Assoc., 24 (3), 237–239, 1974.
3. U.S. Dept. of Energy Office of Industrial Technology, 23. C.P. Fenimore, Formation of nitric oxide in premixed hydro-
Petroleum—Industry of the future: Energy and environ- carbon flames, Proc. Combust. Inst., 13 (1), 373–380, 1971.
mental profile of the U.S. petroleum refining industry, 24. T. Yamamoto, T. Kuwahara, K. Nakaso, and T. Yamamoto,
U.S. DOE, Washington, DC, December 1998. Kinetic study of fuel NO formation from pyrrole type
4. S.A. Shareef, C.L. Anderson, and L.E. Keller, Fired nitrogen, Fuel, 93, 213–220, 2012.
heaters: Nitrogen oxides emissions and controls, U.S. 25. J. Pellegrino, S. Brueske, T. Carole, and H. Andres, Energy
Environmental Protection Agency, Research Triangle and environmental profile of the U.S. petroleum refining
Park, NC, EPA Contract No. 68-02-4286, June 1988. industry, report prepared by Energetics, Inc. for the U.S.
5. D. Hansell and G. England, Air toxic emission factors for Department of Energy, November 2007.
combustion sources using petroleum based fuels, 3 vol- 26. I. Vermeulen, C. Block, and C. Vandecasteele, Estimation
umes, Energy and Environmental Research Corp., Irvine, of fuel-nitrogen oxide emissions from the element com-
CA, 1998. Available at www.api.org/step/piep.htm position of the solid or waste fuel, Fuel, 94, 75–80, 2012.
27. E.B. Sanderford, Alternative control techniques 45. C.A. Miller, A.D. Touati, J. Becker, and J.O.L. Wendt. NOx
document—NOx emissions from process heaters, abatement by fuel-lean reburning: Laboratory combus-
U.S. Environment Protection Agency report EPA- tor and pilot-scale package boiler results. 27th Symposium
453/R-93-015, February, 1993. (Int’l) on Combustion, The Combustion Institute,
28. U.S. Environmental Protection Agency, Alternative con- Pittsburgh, PA, pp. 3189–3195, 1998.
trol techniques—NOx emissions from utility boilers, U.S. 46. J.U. Watts, A. Mann, J. Harvilla, and D. Engelhardt,
EPA report EPA-453/R-94-023, 1994. NOx control by utilization of reburn technologies in
29. A. Medeiros, R. Edenhofer, K. Lucka, and H. Köhne, the United States. Fifth Int’l Conf. on Technologies and
Steady-state determination of fuel-bond (sic) nitrogen in Combustion for a Clean Environment, Lisbon, Portugal,
diesel and industrial gas oil, Clean Air, 8 (4), 359–371, 2007. Vol. 2, The Combustion Institute—Portuguese Section,
30. J. Singer (ed.), Combustion Fossil Power, 4th Edn., pp. 1017–1021, 1999.
Combustion Engineering, Inc., Windsor, CT, 1991. 47. M. Braun-Unkhoff, P. Frank, S. Koger, and W. Leuckel,
31. H. Tsuji, A.K. Gupta, T. Hasegawa, M. Katsuki, Evaluation of NOx reburning models under large scale con-
K. Kishimoto, and M. Morita, High Temperature Air ditions. Fifth Int’l Conf. on Technologies and Combustion for a
Combustion: From Energy Conservation to Pollution Clean Environment, Lisbon, Portugal, Vol. 2, The Combustion
Reduction, CRC Press, Boca Raton, FL, 2003. Institute—Portuguese Section, pp. 1023–1032, 1999.
32. W. Yang and W. Blasiak, High temperature air combus- 48. J.O.L. Wendt, Pollutant formation in furnaces: NOx and
tion for steam reformers, Hydrocarbon Process., 84 (9), fine particulates. Proeedings of Fifth European Conference
115–122, 2005. on Industrial Furnaces and Boilers, Lisbon, Portugal, Vol. 1,
33. W. Bussman and C. Baukal, Ambient conditions impact pp. 129–156, 2000.
CO and NOx emissions: Part I, Petrol. Technol. Quart., 14 (3), 49. J. Colannino, Low-cost techniques reduce boiler NOx,
93–99, 2009. Chem. Eng., February, 100–106, 1993.
34. W. Bussman and C. Baukal, Ambient conditions impact 50. A. Garg, Specify better low-NOx burners for burnaces,
CO and NOx emissions: Part II, Petrol. Technol. Quart., 14 Chem. Eng. Prog., 90 (1), 46–49, 1994.
(4), 37–41, 2009. 51. I.-P. Chung, D. Meinen, R. Poe, J. Lewallen, C. Baukal,
35. C. Baukal and W. Bussman, Chapter 4: Air infiltra- and C. Schnepper, Solving the low NOx dilemma,
tion effects on industrial combustion efficiency, in Fuel Hydrocarbon Eng., 10 (8), 77–80, 2005.
Efficiency, ed. J.K. Bernard, Nova Science Publishers, Inc., 52. I.-P. Chung, J. Chambers, C. Schnepper, R. Poe, J. Jayakaran,
Hauppauge, NY, pp. 101–134, 2011. R. Waibel, M. Claxton, M. Zimola, and D. Foote, Compact
36. R. Waibel, D. Nickeson, L. Radak, and W. Boyd, Fuel stag- low NOx gas burner apparatus and methods, U.S. Patent
ing for NOx control, in Industrial Combustion Technologies, 6,695,609, issued February 24, 2004.
ed. M.A. Lukasiewicz, American Society of Metals, 53. C. Baukal, V. Gershtein, and X. Li (eds.), Computational
Warren, PA, pp. 345–350, 1986. Fluid Dynamics in Industrial Combustion, CRC Press, Boca
37. A. Garg, Trimming NOx, Chem. Eng., 99 (11), 122–124, 1992. Raton, FL, 2001.
38. R. Poe, A. Patel, C. Baukal, and D. Wright, Advanced 54. K. Ahlberg (ed.), AGA Gas Handbook, AGA AB, Lidingö,
combustion system for cracking furnaces, Petrol. Technol. Sweden, 1985.
Quart., 12 (3), 113–117, 2007. 55. S. Setia, VOC Emissions—Hazards and techniques for
39. W. Bussman , R. Waibel, C. Baukal, R. Ruiz, and I. their control, Chem. Eng. World, XXXI(9), 43–47, 1996.
Chung, Remote staged furnace burner configurations 56. W.R. Niessen, Combustion and Incineration Processes, 2nd
and methods, U.S. Patent, US 7,153,129 B2, December Edn., Marcel Dekker, Inc., New York, 1995.
26, 2006. 57. M. Lippmann and R. Albert, The effect of particle size
40. J.G. Seebold, R.T. Waibel, and T.L. Webster, Control refin- on the regional deposition of inhaled aerosols in human
ery NOx emissions cost-effectively. Hydrocarbon Process., respiratory tract, Am. Ind. Hyg. Assoc. J., 30, 257–275, 1969.
80 (11), 55–59, 2001. 58. J. Heyder, L. Armbruster, J. Gebhart et al., Total deposition
41. J.-Y. Ren, F.-N. Egolfopoulos, and T.T. Tsotsis, NOx of aerosol particles in human respiratory tract for nose
emission control of lean methane-air combustion with and mouth breathing, J. Aerosol Sci., 6, 311–320, 1975.
addition of methane reforming products. Combust. Sci. 59. L. Armbruster and H. Breuer, Investigation into defining
Technol., 174, 181–205, 2002. inhalable dust, Ann. Occup. Hyg, 26, 21–32, 1982.
42. H.L. Shelton, Find the right low-NOx solution, Environ. 60. S.C. Soderholm, Particle size-selective sampling in the
Eng. World, November–December, 24, 1996. work place. Report of ACGIH Technical Committee
43. P. Dagaut, J. Luche, F. Lecomte, and M. Cathonnet, The on Air Sampling Procedures, pp. 27–32. American
kinetics of C1 to C4 hydrocarbons-NO reactions in relation Conference of Governmental Industrial Hygienists,
with reburning. Proceedings of Fifth European Conference Cincinnati, OH, 1985.
on Industrial Furnaces and Boilers, Lisbon, Portugal, Vol. 61. I.P. Chung, R.F. Phalen, and D. Dunn-Rankin, Predicted
II, pp. 499–509, 2000. aerosol aspiration efficiency for infants, children, and
44. V.M. Zamansky, L. Ho, P.M. Maly, and W.R. Seeker, adults, Appl. Occup. Environ. Hyg., 8 (7), 1993.
Reburning promoted by nitrogen- and sodium-containing 62. U.S. Department of Commerce, Bacharach test method—
compounds. 26th Symposium (Int’l) on Combustion, The Installation standards of the National Oil Fuel Institute,
Combustion Institute, Pittsburgh, PA, pp. 2075–2082, ASTM Designation D2156–65, 1965, also in German
1996. Standard DIN 51402.
63. Environmental Protection Agency Method 9— 79. B. Metz, O. Davidson, H. de Coninck, M. Loos, and
Determination of stack plume opacity, 40 CFR (Code L. Meyer, Carbon Dioxide Capture and Storage, Prepared
of Federal Regulation) Chapter 1 Part 60 Appendixes A— by Working Group III of the Intergovernmental Panel on
Standards of Performance for New Stationary Sources Climate Change, Cambridge University Press, Cambirdge,
Test Methods, 1907, revised as of July, 1995. U.K., 2005.
64. Environmental Protection Agency Method 5—Determi 80. E.D. Weil, Sulfur compounds, in Kirk-Othmer Encyclopedia
nation of particulate emissions from stationary sources, of Chemical Technology, 3rd Edn., Vol. 22, John Wiley &
40 CFR (Code of Federal Regulation) Chapter 1 Part 60 Sons, New York, 1983.
Appendixes A—Standards of performance for new sta- 81. C.T. Bowman, Chemistry of gaseous pollutant formation
tionary sources test methods, 1970, revised as July, 1995. and destruction, in Fossil Fuel Combustion, eds. W. Bartok
65. ISO 9096, International Standard ISO 9096: Stationary and A.F. Sarofim, John Wiley & Sons, New York, 1991.
Source Emissions—Determination and mass flow rate of 82. American Petroleum Institute, Burners for Fired Heaters
particulate material in gas-carrying ducts, International in General Refinery Services, API Recommended Practice
Organization for Standardization, 003, American 535, 2nd Edn., Washington, DC, January 2006.
National Standards Institute, Washington, DC. 83. C.R. Bruner, Handbook of Incineration Systems, McGraw-
66. H.C. Van de Hulst, Light Scattering by Small Particles, Hill, New York, 1991.
Dover Publications, New York, 1981. 84. S.R. Turns, An Introduction to Combustion, McGraw-Hill,
67. Circular 8333, Ringelmann Smoke Chart Bureau of New York, 1996.
Mines, Revision IC7718, May 1967. 85. G.C. England, T.P. McGrath, L. Gilmer, J.G. Seebold,
68. 40 CFR Chapter 1, part 60, Appendix A, Method 9, M. Lev-On, and T. Hunt, Hazardous air pollutant emis-
Visual Determination of the Opacity of Emissions from sions from gas-fired combustion sources: Emissions and
Stationary Sources, July 1, 2000, published by the Office the effects of design and fuel type, Chemosphere, 42 (5–7),
of the Federal Register National Archives and Records 745–764, 2001.
Administration, Washington, DC. 86. J.G. Seebold and R.T. Waibel, Products of incomplete
69. CFR 40 Chapter I, Part 60.1, Appendix B, Performance combustion (PIC) from petroleum, petrochemical &
Specifications 1, Specifications and Test Procedures for chemical sector process heaters and industrial boilers,
Opacity Continuous Emission Monitoring Systems in 10th International Congress on Combustion By-Products
Stationary Sources, July 1, 2000, published by the Office and Their Health Effects, Ischia, Italy, June 17–20, 2007.
of the Federal Register National Archives and Records 87. CARB, Stationary Source Test Methods, Volume III:
Administration, Washington, DC. Methods for determining emissions of toxic air contami-
70. S. Hersh, S.E. Kerho, D. Giovanni, and D. Mormile, nants from stationary sources, Monitoring and Laboratory
ASME 87-JPGC-FACT-15, Reductions in carbonaceous Division, California Air Resources Board, 1991.
particulate matter emissions from oil fired utility boilers 88. J.F. Collins, J.P. Brown, S.V. Dawson, and M.A. Marty,
through improvements in atomizer performance, pre- Risk assessment for benzo[a]pyrene, Regul. Toxicol.
sented at ASME/IEEE Power Generation Conference, Miami Pharm., 13, 170–184, 1991.
Beach, Florida, October 4–8, 1987, American Society of 89. G.F. Simes, Preparation aids for the development of cate-
Mechanical Engineers, New York, NY. gory II quality assurance project plans, U. S. Environmental
71. C.N. Davies and M. Subari, Aspiration above wind Protection Agency, EPA/600/8-91/004, 1991.
velocity of aerosols with thin-walled nozzles facing and 90. D.W. Hansell, G.C. England, N. Soelberg, W.R. Seeker,
at right angles to the wind direction, J. Aerosol Sci., 13, M. Lev-On, and S. Folwarkow, Development of air tox-
59–71, 1982. ics emissions factors for petroleum industry combus-
72. I.P. Chung and D. Dunn-Rankin, Numerical simulation tion devices, air and waste management association
of two-dimensional cylindrical blunt sampling in vis- 85th Annual Meeting and Exhibition (Publication No.
cous flow, J. Aerosol Sci., 23, 217–232, 1992. 92-98.02), Kansas City, MO, 1992.
73. I.P. Chung and D. Dunn-Rankin, The effect of bluntness 91. EPRI, Electric utility trace substances synthesis report,
and orientation on two-dimensional samplers in calm Electric Power Research Institute, Palo Alto, CA
air, Aerosol Science and Technology, 19, 371–380, 1993. TR-104614, 1994.
74. I.P. Chung, T. Trinh, and D. Dunn-Rankin, Experimental 92. W.J. Pitz, C.K. Westbrook, A.E. Lutz, R.J. Kee, S.M.
investigation of a two-dimensional cylindrical sampler, Senkan, and J.G. Seebold, Numerical modeling capabili-
J. Aerosol Sci., 25 (5), 935–955, 1994. ties for the simulation of toxic by-products formation in
75. I.P. Chung and D. Dunn-Rankin, Experimental investiga- combustion processes, Combust. Sci. Technol., 101, 383–
tion of air flow around blunt aerosol samplers, J. Aerosol 396, 1994.
Sci., 28 (2), 289–305, 1997. 93. M.J. Castaldi, A. Vincitore, and S.M. Senkan, Micro-
76. F.M. White, Fluid Mechanics, McGraw-Hill Book Co., structures of hydrocarbon flames: Methane, Combust. Sci.
Inc., New York, p. 159, 1979. Technol., 107, 1–19, 1995.
77. IEA, CO2 Emissions from Fuel Combustion, pp. 1971–2001, 94. S.S. Chao and A. Attari, Characterization and measure-
OECD/IEA, Paris, France, 2003. ment of natural gas trace constituents, Final Report,
78. C.E. Baukal, Oxygen-Enhanced Combustion, CRC Press, Contract No. 5089-253-1832, Gas Research Institute,
Boca Raton, FL, 1998. Chicago, IL, 1995.
95. C.F. Edwards and P.J. Goix, Effect of fuel composition 106. J. Huang, I. Onal, and S.M. Senkan, Formation of trace
and excess air on VOC emissions and flame structure byproducts in the premixed flames of CH3Cl/C2H4,
of a small-scale industrial burner, Fourth International Environ. Sci. Technol., 31 (5), 1372–1381, 1997.
Congress on Toxic Combustion By-Products, Berkeley, 107. T.M. McGrath, G.C. England, and D. Jones, The origin
CA, 1995. and fate of toxic combustion byproducts in refinery
96. M.J. Castaldi and S.M. Senkan, Micro-structures of hydro- heaters: Research to enable efficient compliance with
carbon flames: Ethane, Combust. Sci. Technol., 116, 167, 1996. the clean air act, air toxics measurements, Final Report,
97. M.J. Castaldi and S.M. Senkan, PAH formation in pre- Energy and Environmental Research Corporation,
mixed flame of ethane, Combust. Sci. Technol., 116–117, Irvine, CA, 1997.
167–181, 1996. 108. M.J. Castaldi and S. Senkan, Real-time, Ultrasensitive
98. C.F. Edwards and P.J. Goix, Effect of fuel gas composi- monitoring of air toxics by laser photoionization time-of-
tion and excess air on VOC emissions from a small-scale, flight mass spectrometry, J. Air Waste Manage. Assoc.,
industrial style burner, Combust. Sci. Technol., 116–117 48 (1), 77–81, 1998.
(1–6), 375, 1996. 109. A.M. Vincitore and S. Senkan, Polycyclic aromatic hydro-
99. D.W. Hansell, Development of toxic emission factors carbon formation in opposed flow diffusion flames of
from source test data collected under the air toxics hot ethane, Combust. Flame, 114 (1–2), 259–266, 1998.
spots program, Final Report, California Air Resources 110. N.M. Marinov, W.J. Pitz, C.K. Westbrook, A.K. Lutz, A.M.
Board, Sacramento, CA, 1996. Vincitore, and S.M. Senkan, Chemical kinetic modeling
100. D.W. Hansell, Updated air toxics emission factors of a methane opposed flow diffusion flame and com-
for petroleum industry combustion equipment, Final parison to experiments, 27th Symposium (International) on
Report, Western States Petroleum Association and Combustion, The Combustion Institute, Pittsburgh, PA,
American Petroleum Institute, 1996. pp. 605–613, 1998.
101. N.M. Marinov, W.J. Pitz, C.K. Westbrook, M.J. Castaldi and 111. N.M. Marinov, W.J. Pitz, C.K. Westbrook, A.M.
S.M. Senkan, Modeling of aromatic and polycyclic aromatic Vincitore, M.J. Castaldi, and S. Senkan, Aromatic and
hydrocarbon formation in premixed methane and ethane polyaromatic hydrocarbon formation in a laminar pre-
flames, Combust. Sci. Technol., 116–117, 211–287, 1996. mixed n-butane flame, Combust. Flame, 114 (1–2), 192–
102. C. Rossi-Lane, D. Stein, and R. Himes, Gas-fired boiler 213, 1998.
and turbine air toxics summary report, Gas Research 112. N.M. Marinov, W.J. Pitz, and C.K. Westbrook, The forma-
Institute, Chicago, IL GRI-95/0200, 1996. tion of aromatics and PAHs in flames, Plenary Lecture
103. S.M. Senkan and M.J. Castaldi, Formation of polycyclic at the Joint Meeting of the British, German and French
aromatic hydrocarbons (PAH) in hydrocarbon combus- Sections of the Combustion Institute, Nancy, France,
tion: Comparative new results from premixed flames, 18–21, 1999.
Combust. Flame, 107, 141–150, 1996. 113. C.A. Miller and J.V. Ryan, Characterization of air toxics
104. C.M. Gittins, M. J. Castaldi, S. M. Senkan et al., Real-time from an oil-fired firetube boiler, J. AWMA, 46, 742–748, 1996.
quantitative analysis of combustion-generated polycy- 114. Compilation of air pollutant emission factors, EPA
clic aromatic hydrocarbons by resonance-enhanced mul- AP-42, Draft section 1.4—Natural Gas Combustion. US
tiphoton ionization time-of-flight mass spectrometry, Environmental Protection Agency, 1997.
Anal. Chem., 69 (3), 286–293, 1997. 115. U.S. Environmental Protection Agency, Air quality trends,
105. D.W. Hansell, Analysis of HAP emission factors from Research Triangle Park, NC, http://www.epa.gov/airtrends/
petroleum industry combustion devices, Energy and agtrends.html#comparision (accessed on December 10,
Environmental Research Corporation, Irvine, CA, 1997. 2012.)
11
Noise
Wes Bussman, Jay Karan, Carl-Christian Hantschk, and Edwin Schorer
CONTENTS
11.1 Fundamentals of Sound................................................................................................................................................ 268
11.1.1 Introduction........................................................................................................................................................ 268
11.1.2 Basics of Sound................................................................................................................................................... 269
11.1.2.1 Sound Pressure Level and Frequency.............................................................................................. 269
11.1.2.2 Decibel.................................................................................................................................................. 271
11.1.2.3 Sound Power Level............................................................................................................................. 272
11.1.2.4 Threshold of Hearing......................................................................................................................... 272
11.1.2.5 Threshold of Pain................................................................................................................................ 273
11.1.2.6 Correction Scales................................................................................................................................. 273
11.1.3 Measurements.....................................................................................................................................................274
11.1.3.1 Overall Sound Level and How to Add dB Values.......................................................................... 276
11.1.3.2 Atmospheric Attenuation.................................................................................................................. 277
11.2 Industrial Noise Pollution............................................................................................................................................ 278
11.2.1 OSHA Requirements......................................................................................................................................... 280
11.2.2 International Requirements............................................................................................................................. 280
11.2.3 Noise Sources and Environment Interaction................................................................................................. 280
11.3 Mechanisms of Industrial Combustion Equipment Noise...................................................................................... 281
11.3.1 Combustion Roar and Combustion Instability Noise.................................................................................. 281
11.3.1.1 Flare Combustion Roar...................................................................................................................... 281
11.3.1.2 Flare Combustion Instability Noise................................................................................................. 285
11.3.1.3 Burner Combustion Noise................................................................................................................. 286
11.3.1.4 Burner Combustion Instability Noise.............................................................................................. 287
11.3.2 Fan Noise............................................................................................................................................................ 287
11.3.3 Gas Jet Noise....................................................................................................................................................... 287
11.3.3.1 Gas Jet Mixing Noise.......................................................................................................................... 287
11.3.3.2 Shock-Associated Noise..................................................................................................................... 288
11.3.4 Valve and Piping Noise..................................................................................................................................... 289
11.4 Noise Abatement Techniques...................................................................................................................................... 290
11.4.1 Flare Noise Abatement Techniques................................................................................................................ 290
11.4.2 Burner Noise Abatement Techniques............................................................................................................. 292
11.4.3 Valve and Piping Noise Abatement Techniques........................................................................................... 293
11.4.4 Fan Noise Abatement Techniques................................................................................................................... 295
11.5 Analysis of Combustion Equipment Noise................................................................................................................ 295
11.5.1 Multiple Burner Interaction.............................................................................................................................. 295
11.5.2 High-Pressure Flare........................................................................................................................................... 296
11.5.3 Atmospheric Attenuation Example................................................................................................................. 297
Glossary.................................................................................................................................................................................... 297
References................................................................................................................................................................................. 299
Bibliography............................................................................................................................................................................ 299
267
Americans who have some degree of hearing loss: about

one-third of these—more than 9 million—have been
11.1 Fundamentals of Sound
affected, at least in part, by exposure to excessive noise.
11.1.1 Introduction The sense of hearing is a fragile and vital function of the
Silence is golden human body. It resembles the sense of vision, more so than
the other senses, because permanent and complete dam-
—Anonymous age can occur quite easily in an industrial environment.
So noise pollution has been recognized as a safety con-
Noise is referred to as unwanted sound; it is a common cern for a long time and has been appropriately regulated.
by-product of our mechanized civilization and is an Although personnel safety may be the most impor-
insidious danger in industrial environments. Noise pol- tant concern, noise pollution has several other signifi-
lution is usually a local problem and so is not viewed on cant side effects. Sometimes combustion performance
the same scale of importance as the more high-profile must be balanced with noise performance and in some
industrial emissions like NOx, CO, and particulates (see extreme cases, the performance of the equipment must
Chapter 10). Nonetheless, it is an environmental pollut- be compromised somewhat to achieve the noise perfor-
ant of significant impact. mance required. To those in relatively remote or unpop-
Serious concern is merited when a pollutant can result ulated areas, it may seem unthinkable that equipment
in either environmental damage or human discomfort. performance could be compromised for the sake of noise
Considering the impact on people, noise is most often a control. However, given the age and economic drivers of
source of annoyance, but it can also have more detrimen- the petroleum refining and chemical industries, it is now
tal effects, such as causing physical injury. Noise-related common to find plants located in densely populated
injuries range from short-term discomfort to permanent areas (see Figure 11.1). Industry located in close prox-
hearing loss. According to recent statistics, more than 20 imity to residential areas or busy commercial facilities
million Americans are exposed to hazardous sound lev- can create levels of noise that the people in the neighbor-
els on a regular basis. There are approximately 28 million hood find objectionable, resulting in regulations aimed
FIGURE 11.1
Community located close to an industrial plant. (Photograph from Dreamstime.)
Noise 269
at controlling noise emissions. Within the industrial site

itself, the immediate issue with noise is employee safety.
In addition to the concern with physical safety, it has
been found that reduction in noise levels contributes to
improved employee work performance and morale.
Equipment is also affected by noise. In most cases, these
effects lie in the area of vibration control and are beyond
the scope of this chapter. Suffice it to say that noise and
mechanical vibration usually occur together. High levels
of mechanical vibration are accompanied by high lev-
els of noise and vice versa. Vibration leads to mechani-
cal fatigue, which reduces equipment life. The effects of
fatigue are frequently accepted as normal wear and tear
if the equipment life cycle spans a reasonable duration. In
extreme cases, the effects of vibration may be more rap-
idly manifested, such as in the case of premature crack-
ing and spalling of hard refractory linings in furnaces.
This chapter is written as a practical guide, as well
as a reference, for engineers involved in the design,
operation, or maintenance of combustion equipment—
be it burners, furnaces, flares, or thermal oxidizers. In
addition, because this chapter provides a comprehen-
sive coverage of the fundamentals of sound, the cre-
ative engineer will also be able to extend his or her
knowledge to analyze other noise-producing indus- FIGURE 11.2
trial equipment. Tree falling in the forest. (Courtesy of Dreamstime.)
cochlea, that has tiny hair cells on the inside. The hair
11.1.2 Basics of Sound cells sense the vibration conveyed to the cochlea and
convert the vibrations into nerve pulses which are then
What is sound? An interesting question one may have
conveyed to the brain through the auditory nerve.
been asked before is “If a tree falls in the forest and
In reality, most naturally occurring sounds are com-
nobody is around to hear it, does it still make a sound?”
posites of different pressure levels at various frequen-
(see Figure 11.2). Webster’s dictionary defines sound as
cies. On the other hand, a pure tone is a sound at only
“That which is heard.” So, according to this definition,
one frequency. Any pure tone can be uniquely identi-
the tree will not make a sound as it falls. Obviously, an
fied by two of its properties, namely, pressure level and
engineer will find this definition inadequate for techni-
frequency. A tuning fork is an example of a pure tone
cal purposes. The definition provided by the Handbook of
generator. Naturally occurring pure tone generators are
Noise Measurement is more accurate: “Sound is the vibra-
rare. Even musical instruments create notes that have
tion of particles in a gas, liquid or solid.”1
significant pressure levels at two or three multiples, or
Sound is propagated through any medium in waves
harmonics, of the fundamental frequency of the note.
that take the form of pressure peaks (compressions)
and troughs (rarefactions) as illustrated in Figure 11.3.
11.1.2.1 Sound Pressure Level and Frequency
The pressure wave travels through the medium at the
speed of sound in that medium. The auditory system Pressure level defines the loudness of the sound,
in humans and most animals senses the impingement while frequency defines the pitch or tone of the sound.
of these pressure waves on a tissue membrane and con- Pressure level is the amplitude of the compression, or
verts them to electrical impulses that are then sent to the rarefaction, of the pressure wave. The common unit of
brain where they are interpreted as the information we pressure level is decibel, abbreviated to “dB.” Frequency
receive from the sound. is the number of pressure waves that pass by an arbi-
Figure 11.4 shows a cross section of the human ear. trary point of reference, in a given unit of time. As such,
Sound is collected and funneled into the ear canal by the measure of sound frequency can be cycles per sec-
the outer ear. At the end of the ear canal, the sound ond (cps), and as with electricity, the commonly used
impinges on the ear drum. The bones of the middle ear unit is Hertz (Hz); 1 Hz = 1 cps.
convey the ear drum’s vibration to the inner ear. The Another important quantity in the description of
inner ear consists of a fluid-filled membrane called the sound waves is the speed of sound which is typically
Sound wave—A pressure wave moving

at the speed of sound
Peak
+
Pressure
–
Trough
FIGURE 11.3
Pressure peaks and troughs. (Photograph of base from Dreamstime.)
Nerve that sends The speed at which a sound wave propagates depends
electrical signals mainly on the type of fluid and the temperature
Eardrum to the brain
of the fluid. Table 11.1 gives some examples for the
speed of sound in different media and at different
Cochlea temperatures.
The typical range of human hearing extends from
20 Hz to 20 kHz. Young children can hear frequencies
Ear canal slightly higher than 20 kHz but this ability dimin-
Fluid-filled
ishes with age. This trend of reduced high-frequency
membrane sensitivity continues with advancing age. Loss of
hearing in humans in the later stages of life typically
TABLE 11.1
Speed of Sound in Different Media
Hair cells and at Different Temperatures
FIGURE 11.4 Temperature Speed of

Cross section of the human ear. (From Baukal, C.E. (ed.), The John Zink Fluid (°C) Sound, c (m/s)
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) Air, dry −40 307
20 320
designated by the letter c. The speed of sound describes
0 332
the velocity at which a sound wave propagates through
20 344
the ambient air or other fluids (e.g., water). In other
40 356
words, the speed of sound describes the distance a
Hydrogen 20 1316
sound wave propagates per unit of time. The length
Methane 20 448
of a sound wave is another important quantity in the
Ethylene 20 331
description of sound. The wavelength, λ, describes
Water 0 1403
the distance between one wave crest and the next, i.e.,
40 1529
between one point of maximum compression in the
80 1555
sound wave to the next. The frequency f, speed of sound
100 1543
c, and wavelength λ are related as follows:
Steel 20 5180
c = λf (11.1) Soft rubber 20 54

Noise 271
TABLE 11.2 in frequency. Table 11.2 shows the 10 octave bands that
The 10 Octave Bands cover the hearing range and the center frequencies
that can be used to represent each octave band. Each
Full Octave Band Standards
octave band extends over seven fundamental musical
Center notes.
Octave Band (Hz) Frequency (Hz)
22–44 31.5
44–88 63
11.1.2.2 Decibel
88–177 125
177–355 250 The unit of sound level, the decibel, is difficult to visual-
355–710 500 ize and warrants some explanation. While it is possible
710–1420 1,000 to quantify sound in units of either power or pressure,
1,420–2,840 2,000 neither unit is convenient to use because in practice one
2,840–5,680 4,000 has to deal with sounds that extend over a very large
5,680–11,360 8,000 range of power or pressure values. For example, the
11,360–22,720 16,000 sound power of a whisper is 10−9 W, while the sound
Source: Baukal, C.E. (ed.), The John Zink power of a jet plane is 103 W. The range of these two
Combustion Handbook, CRC sound sources spans 1012 W. The decibel, a dimension-
Press, Boca Raton, FL, 2001. less unit, was invented in order to represent these large
ranges conveniently.
manifests itself as diminished sensitivity to frequen- In the 1960s, Bell Laboratories coined the term “deci-
cies from 10 to 20 kHz. Mechanically, this is due to bel.” The “deci” stands for the base 10 log scale on which
the deterioration of the fine hair cells in the basilar the decibel is based and the “bel” was meant to repre-
membrane. sent Bell Labs. See Figure 11.5 for how decibel relates to
It is important to note that the ear is not equally sensi- watts.
tive over the entire range from 20 Hz to 20 kHz. This is In Figure 11.5, the y-axis corresponds to the sound
vital to understanding how noise affects us and how power in watts, and follows a base-10 scale. The x-axis
noise control is implemented. The human ear is much corresponds to the sound power level in dB (PWL).
less sensitive to sound at the extremes of low and high The line provides the relationship between the sound
frequencies and we will discuss this in more detail later power and the PWL; for example, a PWL of 120 dB is
in the chapter. equal to a sound power of 1 W. As an illustration of
The wide range of frequencies in our hearing range the log10 relationship, note that 110 dB is equal to one-
may be conveniently handled by breaking it up into tenth of a watt 0.1 (0.1 W) and 100 dB is equal to a hun-
octave bands. Each octave band represents a doubling dredth of a watt (0.01 W).
1 × 101
1 × 100
1 × 10–1
1 × 10–2
1 × 10–3
1 × 10–4
Power (W)
1 × 10–5
1 × 10–6
1 × 10–7
1 × 10–8
1 × 10–9
1 × 10–10
1 × 10–11
1 × 10–12
0 10 20 30 40 50 60 70 80 90 100 110 120 130
(dB)
FIGURE 11.5
Relationship of decibels to watts. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
The following are useful equations that may be used where

to calculate sound pressure and power levels, in dB, Lp is the sound pressure level (dB)
from the equivalent pressure and power units. LW is the sound power level (dB)
r is the distance from source (ft)
 P  Equation 11.4 is only valid exactly if the following condi-

Lp (dB) = 20 log 10  −5  (11.2)
 2 × 10  tions are fulfilled:
 W  • Noise radiation from the source is uniform and

Lw (dB) = 10 log 10  −12  (11.3) equal in every direction.
 1 × 10 
• The source is small compared to the distance r.
• There is no relevant noise contribution from
where other sources at the point of interest.
Lp is the sound pressure level (dB)
• There is no relevant influence on the noise at the
Lw is the sound power level (dB)
point of interest from sound reflecting surfaces
P is the sound pressure (N/m2)
nearby.
W is the sound power (W)
If the aforementioned conditions are sufficiently ful-

filled, then Equation 11.4 cannot only be used to calcu-
11.1.2.3 Sound Power Level
late the SPL expected at a certain distance r from the
There is a subtle but important difference between the source (knowing the PWL LW), but it can also be used to
terms PWL, and sound pressure level (SPL). PWL is back-calculate the PWL of a source from a measurement
used to indicate the total energy emitting ability of a of the SPL at a known distance r from the source. The
sound source. In other words, sound power is a char- latter procedure is a technique that is standardized in
acteristic of the sound source itself. SPL, on the other DIN EN ISO 3744.2
hand, is used to indicate the intensity of sound received If the aforementioned conditions, however, are not
at any point of interest, from one or more sources. The fulfilled, then the SPL predicted at a certain distance
illustration in Figure 11.6 shows the formula to calcu- from a sound source could differ significantly from
late the SPL to be expected at a distance r from a spheri- the level actually measured, or the PWL of a source
cally radiating source of power level Lw: back-calculated from a measurement of the SPL can
differ significantly from the real PWL of the source. In
(11.4) this context, it is important to keep in mind that the
Lp = Lw − 10 log 10 ( 4πr 2 ) + 10.5
PWL is an intrinsic property of the source which is
independent from the environment while the SPL not
only depends on the distance to the source, but is also
influenced by other sound sources, the environment
(reflecting surfaces, etc.), and the sound propagation
Sound pressure conditions (temperature, atmospheric pressure, wind,
wave at a barriers, etc.).
distance r feet
from the source r
11.1.2.4 Threshold of Hearing

Figure 11.7 reveals a map of the threshold of hearing in
Noise source with a humans. The y-axis represents the SPL in dB and the
The SPL at given SPL x-axis represents frequency. Any SPL that falls below
this point can be (Lw) radiating outward the curve at any given frequency will be inaudible to
calculated as humans. For example, a SPL of 30 dB at 63 Hz will be
inaudible; whereas, a SPL of 70 dB at the same 63 Hz
Lp = Lw – 10 log10 (4πr2) + 10.5 frequency will be audible. Humans are most sensi-
where r is in feet tive to sounds in the so-called “mid-frequencies” from
FIGURE 11.6
1 kHz to about 5 kHz. This is generally the range of
Calculating SPL at a distance r. (From Baukal, C.E. (ed.), The John Zink frequencies of the human voice and many other impor-
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) tant common sounds. Additionally, at a constant level,
Noise 273
130 11.1.2.6 Correction Scales

110 Sound meters are capable of measuring with equal sen-
90 sitivity over the entire audible range. However, because
humans do not hear with equal sensitivity at all fre-
70
quencies, the sound meter’s measurement needs to be
SPL (dB)
Human threshold of hearing

50 modified to quantify what really affects humans. This
can be done using a correction curve. The most common
30
correction is the A-scale correction curve which resem-
10 bles an idealized inverse of the threshold of hearing
–10 curve (refer to Figure 11.9). An A-weighted sound level
correlates reasonably well with hearing-damage risk in
–30 industry and with subjective annoyance for a wide cat-
16
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
20,000
egory of industrial and community noises. After apply-
ing the A-scale correction, the unit of SPL becomes the
Frequency (Hz) dBA scale. Figure 11.10 shows a typical burner noise
FIGURE 11.7 curve as measured by a noise meter (flat scale) and the
Threshold of hearing in humans. (From Baukal, C.E. (ed.), The John result after applying A-scale correction.
sound with a low or very high frequency will not have 80

the same loudness sensation as that in the medium fre-
60
quency range. For example, a 100 Hz tone at a SPL of
50 dB is perceived as the same loudness as a 1000 Hz 40
tone at SPL of 40 dB.3 Threshold of hearing
SPL (dB)
20
11.1.2.5 Threshold of Pain
0
Figure 11.8 shows the threshold of pain superimposed
on the threshold of hearing. Fortunately, the curve rep- –20
resenting the threshold of pain is relatively flat. In gen- A-scale correction factor
eral, a SPL of over 120 dB at any frequency will cause –40
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
pain. An important observation that can be derived
from the two curves is that if a sound is audible at very
low or very high frequencies, persons subject to this Frequency (Hz)
sound are very close to experiencing pain. FIGURE 11.9
A-weighted scale for human hearing threshold. (From Baukal, C.E. (ed.),
130 The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
110
Threshold of pain
90 85
80
70 No weighting
SPL (dB)
Threshold of hearing 75
50 70
SPL (dB)
30 65
60
10 A-weighted
55
–10 50
–30 45
16
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
20,000
40
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
Frequency (Hz)
Frequency (Hz)
FIGURE 11.8
Threshold of hearing and threshold of pain in humans. (From FIGURE 11.10
Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, A-weighted burner noise curve. (From Baukal, C.E. (ed.), The John
Boca Raton, FL, 2001.) Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
+20 signal before it can be analyzed, measured, or dis-

+10
played. Special weighting networks are used to shape
the signal spectrum and apply the various correction
0 scales discussed earlier. The weighted signal then
Relative response (dB)
–10 passes through a second output amplifier into a meter.

The meter and associated electronic circuits detect the
–20 approximate root-mean-square (rms) value of the sig-
A
–30 nal and display it in units of dB.
B
Noise meters range from the simplest—microphone
–40 C
and needle gauge—to sophisticated digital signal pro-
D
–50 cessing (DSP) equipped analyzers. The more sophis-
–60
ticated analyzers are equipped with fast Fourier
transform (FFT) capabilities that aid in accurate nar-
–70 row band analysis. In general, spectrum analyzers
10 2 5 102 2 5 103 2 5 104 2
allow the user to map the SPL at different frequencies,
Frequency (Hz)
or in other words, generate a curve of the sound over
FIGURE 11.11 different frequencies. However, there is a significant
Weighting curves A, B, C, and D. (From Baukal, C.E. (ed.), The John difference between instruments that make one mea-
Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) surement per octave band and those that slice the
octave band up into several intervals and make a mea-
The other, less used correction scales are named, as surement at each interval. Typically, instruments are
might be expected, B, C, and D. Referring to Figure 11.11, capable of carrying out
one can see that the C-scale is essentially flat over
the range of interest and the B-scale lies somewhere 1. Octave band measurements
between the A- and C-scales. Given an understanding 2. One-third octave band measurements
of the influence of low-frequency sounds, one finds that 3. Narrow-band measurements
the B- and C-scales do not apply adequate correction in
the lower frequencies. Finally, the D-scale is different Table 11.3 shows the usual octave and one-third octave
from the others in that it has a pronounced correction in bands. As the name suggests, a one-third octave band
the range of 2–5 kHz. The D-scale was devised for the instrument makes three measurements in each octave
aircraft industry and is rarely used otherwise. as opposed to the single measurement of the octave
band instrument. A narrow-band instrument, on the
other hand, uses DSP to implement FFT analysis, and
11.1.3 Measurements
in the current state of the art, FFT analysis allows the
A simple schematic of a noise meter is shown in Figure analyzed frequency range to be sliced up into a large
11.12. The microphone is designed with a transducer number of smaller intervals, limited in number only by
that transforms pressure variations in air to a cor- the measured time interval’s length and the available
responding electrical signal. Since the electrical sig- computer power.
nal generated by the microphone is relatively small Figure 11.13 provides a comparison of the same sound
in magnitude, a preamplifier is needed to boost the spectrum as analyzed using three different frequency
band intervals: octave band, one-third octave band, and
Weighting networks narrow band. This comparison shows that the addi-
A tional resolution provided by narrower band methods
Microphone Rectifier Fast can be of vital importance. In this example the level at
B
Meter 1 kHz, as recorded by the octave band instrument, is
C 90 dB; on the one-third octave instrument, it is 85 dB,
Amplifier Amplifier Slow
D and on the narrow-band instrument it is 70 dB. The
Flat lower resolution measurements produce higher values
Output due to the spill-over influence of the nearby peak at
1.8 kHz. In addition, in implementing noise control for
Band this source, it is very valuable to know that it is the nar-
filters row peak at 1.8 kHz that is driving the maximum noise.
FIGURE 11.12
This knowledge helps to zero-in on the source.
Block diagram of a sound level meter. (From Baukal, C.E. (ed.), The However, as with many things, there is a cost asso-
John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) ciated with high performance. For most applications,
Noise 275
TABLE 11.3 source in space (sound radiating spherically), whether

Octave and One-Third Octave Bands the source is located close to a flat surface (sound radiat
ing hemispherically), or between two flat surfaces
Octave One-Third Octave
(sound radiating a quarter of a sphere), will make a dif-
Lower Upper Lower Upper ference in how the measurement needs to be performed.
Band Band Band Band
However, a detailed discussion of measurement issues
Band Limit Center Limit Limit Center Limit
is beyond the scope of this chapter and the reader may
12 11 16 22 14.1 16 17.8 use some of the more comprehensive works in the list
13 17.8 20 22.4 of references at the end of this chapter. The American
14 22.4 25 28.2 Petroleum Institute has issued a recommended practice
15 22 31.5 44 28.2 31.5 35.5 for measuring noise from fired process heaters.4
16 35.5 40 44.7
17 44.7 50 56.2 100
18 44 63 88 56.2 63 70.8
19 70.8 80 89.1
90
20 89.1 100 112
21 88 125 177 112 125 141
80
22 141 160 178
23 178 200 224
70
24 177 250 355 224 250 282
25 282 315 355
60
26 355 400 447
27 355 500 710 447 500 562
50
28 562 630 708
(a)
29 708 800 891
30 710 1,000 1,420 891 1,000 1,122 100
31 1,122 1,250 1,413
32 1,413 1,600 1,778 90
33 1,420 2,000 2,840 1,778 2,000 2,239
34 2,239 2,500 2,818 80
35 2,818 3,150 3,548
36 2,840 4,000 5,680 3,548 4,000 4,467 70
37 4,467 5,000 5,623
38 5,623 6,300 7,079
60
39 5,680 8,000 11,360 7,079 8,000 8,913
40 8,913 10,000 11,220
50
41 11,220 12,500 14,130 (b)
42 11,360 16,000 22,720 14,130 16,000 17,780
43 17,780 20,000 22,390 100
Press, Boca Raton, FL, 2001. 90
Note: The advantages of making broad band analyses of sound using
octave or one-third octave band filter sets are that less time is 80
needed to obtain data and the instrumentation required to mea-
sure the data is less expensive. The main disadvantage is the
loss of detailed information about the sound which is available 70
from narrow-band (FFT) analyzers.
60
a one-third octave analysis is usually adequate. The

50
advantages of making broad band analyses using octave 10 100 1k 10 k
or one-third octave band filter sets are that less time is (c) Frequency (Hz)
needed to obtain data and the instrumentation required
to measure the data is less expensive. FIGURE 11.13
Same sound spectrum on three different intervals. (a) Octave band
When making sound measurements, several factors spectrum, (b) one-third octave band spectrum, and (c) narrow-
regarding the nature of the source should be consid- band spectrum. (From Baukal, C.E. (ed.), The John Zink Combustion
ered. For example, whether the source is a true point Handbook, CRC Press, Boca Raton, FL, 2001.)
11.1.3.1 Overall Sound Level and How to Add dB Values TABLE 11.4
As mentioned earlier, most sounds are composites of A-Weighting of the Burner Sound Curve from Figure 11.14
several different levels at different frequencies. This Frequency (Hz) SPL (dB) A-Scale CF (dB) SPL (dBA)
is especially true of industrial noise. A typical burner 31.5 72 −39 33
noise curve is shown in Figure 11.14. As can be seen, 63 75 −26 49
there are significantly higher levels in two frequency 125 79 −16 63
zones, both of which will contribute to the apparent 250 79 −9 70
intensity experienced by a person working in the vicin- 500 72 −3 69
ity of the burner. It is difficult to describe this sound 1,000 69 0 69
without using either a diagram like the one shown or 2,000 68 1 69
a table listing various SPLs occurring in the different 4,000 78 1 79
octave bands. The “overall sound level,” a single num- 8,000 83 −1 82
ber, has been devised to represent such composite 16,000 80 −7 73
sound curves conveniently. If a single number is to be
used to represent the whole curve, then it is not practi-
cal to use the average of the various levels in the octave
bands, since this number would be less than the levels magnitude are added, say 100 dB and 100 dB, the result
at the peaks and the peaks have the most influence on is 103 dB. The formula used to add SPLs is as follows:
the listener. Therefore, one must not confuse the average
with the overall sound level.  n

The overall sound level is calculated by adding the Ltotal (dB) = 10 log 10 

∑
i =1
100.1Li 

individual levels in the various octave bands. In col-
umns 1 and 2 of Table 11.4, the burner sound curve has
been split up into its component levels in each octave
(
= 10 log 10 100.1L1 + 100.1L2 + 100.1L3 + ⋅⋅⋅ ) (11.5)
band. In column 3, the A-weighted correction has
similarly been split up and listed. Column 4 gives the where
A-corrected values for the sound curve by simply sub- Ltotal is the total SPL level (dB)
tracting column 3 from column 2. Now, the values in Li is each individual level (dB)
column 4 must be combined to obtain the A-weighted n is the number of SPLs to be added
overall sound level.
Since the decibel is based on a log10 scale, simple addi- To demonstrate how to use this equation, consider the
tion cannot be used. For example, if two values of equal following example:
85
80
75
70
SPL (dB)
65
60
55
50
45
40
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
Frequency (Hz)
FIGURE 11.14
Typical burner noise curve. (Adapted from Beranek, L.L., Noise and Vibration Control, McGraw Hill Book Co., New York, 1971.)
Noise 277
Example 11.1 TABLE 11.5

Add the following SPLs together: 88, 92, and 86 dB Addition Rules
What is the overall dBA level?
( )
Ltotal (dB) = 10 log 10 10 0.1× 88 + 10 0.1× 92 + 10 0.1× 86 = 94.17 dB Frequency SPL A-scale CF SPL
Hz dB dB dBA
Some simple rules of thumb can be used to perform 31.5 72 –39 33
quick estimates. They are as follows: 63 75 –26 49 49
125 79 –16 63 63
250 79 –9 70 71
1. When adding dB values that are of equal mag- 500 72 –3 69 73
nitude or that differ by 1, the sum is 3 dB added 1,000 69 0 69 75
2,000 68 1 69 76
to the greater number. 4,000 78 1 79 80
2. When the two values are different by 2–3 dB, then 8,000 83 –1 82 84
the sum is 2 dB added to the greater number. 16,000 80 –7 73 85
3. When adding two values that differ by 4–9 dB, Overall sound level = 85 dBA
then the sum is 1 dB added to the greater number.
Caution: Overall SPL does not = average SPL
4. For values that differ by 8 dB or more, the sum
is just the larger number.
CRC Press, Boca Raton, FL, 2001.
5. Always start with the smallest number in the
list and add it to the next larger number. provided are approximations. For exact calculations, the
formulas should be used.
To better understand why these rules work, refer to the Table 11.5 shows the effect of applying the addi-
chart in Figure 11.5. From the chart it can be seen that tion rules to the values generated by breaking up the
1 W is equal to 120 dB. burner noise curve. At the end of the addition list, 1 dB
has been added to compensate for any errors due to
1 W = 120 dB approximation.

As an alternative to using the aforementioned rules
1 W = 120 dB of thumb for level addition, the nomogram in Figure

11.15 can be used to determine the sum of the two levels
2 W = 123 dB
L1 and L2: For a certain difference between the two levels
L1 and L2 in the lower scale of the nomogram, the cor-
On the chart, 2 W registers 123 dB on the line. Similarly, responding level increase ΔL to be added to the higher
the reason that numbers 10 dB or more in difference are one of the two levels can be found in the upper part of
neglected is because the nomogram. For example, add the following pres-
sure levels: 90 dB + 90 dB. The difference between these
1.0 W = 120 dB
two values is zero. Locating zero on the lower scale of
0.1 W = 110 dB the nomogram, one finds a value of 3 directly above it;

therefore, 90 dB + 90 dB = 93 dB.
1.1 W = 120 dB
Since the 110 dB contributes only a 0.1 W, it is neglected 11.1.3.2 Atmospheric Attenuation
in the approximation. The example becomes more vivid When a sound wave travels through still air, it is
when adding two numbers that differ by 20 dB or more. absorbed or attenuated by the atmosphere. Over a
1.00 W = 120 dB ΔL
0.01 W = 100 dB 0 0.5 1 1.5 2 2.5 3

1.01 W = 120 dB
Rule number 5 is especially necessary when adding a

list that contains several numbers that are almost equal 20 10 9 8 7 6 5 4 3 2 1 0
in value and one or more that are 10 dB greater, such as
L1 – L2
in a list that contains six values of 90 dB and one value of
100 dB. If we begin to add from the 100 dB value we will FIGURE 11.15
arrive at a wrong result. It should be noted that the rules Nomogram for noise level addition.
couple of hundred feet, the atmosphere does not sig-

nificantly attenuate the sound; however, over a few
11.2 Industrial Noise Pollution
thousand feet, the sound level can be substantially
reduced. The amount of sound that is attenuated in still Thus far, sound has been discussed. So what is noise?
air largely depends on the atmospheric temperature An all-encompassing definition would be that noise is
and relative humidity. Figure 11.16 depicts the atmo- any undesirable sound. By saying this, the concept is
spheric attenuation for aircraft-to-ground propagation introduced that what is considered to be noise is some-
in SPL per 1000 ft (300 m) distance for center frequencies what subjective, and depends on several temporal and
of 500, 1000, 2000, 4000, and 8000 Hz. Notice that the circumstantial factors.
atmospheric attenuation is larger at higher frequencies For example, it is not unusual for a person to encoun-
than at lower frequencies. For example, suppose that we ter SPLs of 100–110 dB at a sporting event, in a stadium
are 1000 ft (300 m) away from a noise source and that full of cheering fans, and yet not be perturbed by it. On
the atmospheric temperature and relative humidity is the contrary, the barely 45 dB sound of a dripping faucet
80°F (27°C) and 10%, respectively. The plots in Figure may cause considerable annoyance in the quiet of the
11.16 show that the atmospheric attenuation for 500 Hz night. Table 11.6 gives some examples of noise levels.
is approximately 2 dB whereas for 8000 Hz the attenua- Industrial noise pollution is a major concern for soci-
tion is 55 dB. ety as a whole. In a recent survey, the effects of exposure
Atmospheric attenuation, outdoors, can also be to noise in refinery workers were studied extensively.
affected by turbulence, fog, rain, and snow. Typically, A cross section of workers in different divisions/units was
the more turbulence present in the air, the more the chosen. It was found that noise levels averaged 87–88 dBA
attenuation. There appears to be conflicting evidence in the aromatic and paraffin facilities and 89 dBA in alkyl-
as to whether or not fog attenuates sound. It is recom- ation facilities. In comparison, workers in the warehouse,
mended that no excess attenuation be assigned to fog or health clinics, laboratories, and offices were, generally,
light precipitation. found to be exposed to much lower levels.
Temperature (°C)
–10 0 10 20 30 40
40
10
355/710 Hz, fourth octave band, GMF 500 Hz
dB/1000 m
20
0 0
Relative 40
Atmospheric attenuation (dB/1000 ft)
humidity, %
10
710/1400 Hz, fifth octave band, GMF 1000 Hz 10
20 20
dB/1000 m
30
0 50 0
70
20 90
60
1400/2800 Hz, sixth octave band, GMF 2000 Hz
40
dB/1000 m
10
20
0 0
0 10 20 30 40 50 60 70 80 90 100
(a) Temperature (°F)
FIGURE 11.16
Atmospheric attenuation for octave bands with center (geometric-mean) frequencies at (a) 500, 1000, and 2000 Hz. (From Baukal, C.E. (ed.), The
John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
(continued)
Noise 279
Temperature (°C)
–10 0 10 20 30 40
2800/5600 Hz
30 Seventh octave band 100
GMF 4000 Hz
Relative
humidity (%) 80
10
dB/1000 m
20 20
30 60
50
70
90 40
10
20
0 0
0 10 20 30 40 50 60 70 80 90 100
(b) Temperature (°F)
Temperature (°C)
–10 0 10 20 30 40
50
5600/11,200 Hz 160
Eighth octave band
GMF 8000 Hz
140
Relative
40
humidity (%)
120
10
20
100
dB/1000 m
30 30
50
60 80
70
20 90
60
40
10
20
0 0
0 10 20 30 40 50 60 70 80 90 100
(c) Temperature (°F)

Atmospheric attenuation for octave bands with center (geometric-mean) frequencies at (b) 4000 Hz; and (c) 8000 Hz. (From Baukal, C.E. (ed.),
The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Noise can damage hearing, and can cause physical or exposure to noise exceeding 85 dBA can lead to short-
mental stress (increased pulse rate, high blood pressure, term reversible hearing impairment and long-term
nervousness, sleep disorders, lack of concentration, exposure to levels higher than 85 dBA can cause perma-
and irritability). Irreparable damage can be caused by nent hearing loss.
single transient sound events with peak levels exceed- The following is a mathematical model based on
ing 140 dBA (e.g., shots or explosions). Long-duration empirical data (ISO 1999) used to calculate the maximum
TABLE 11.6 TABLE 11.7

Sound Levels of Various Sources OSHA Permissible Noise Exposure
Threshold of hearing 0 dBA Duration per Day (h) SPL (dBA) (Slow Response)
Rustle of leaves 10 dBA 8.0 90
Normal conversation (at 1 M) 30 dBA
6.0 92
Min level in Chicago at night 40 dBA
4.0 95
City street, very busy traffic 70 dBA
3.0 97
Noisiest spot at Niagara Falls 85 dBA
Threshold of pain 120 dBA 2.0 100
Jet engine (at 50 M) 130 dBA 1.5 102
Rocket (at 50 M) 200 dBA 1.0 105
0.5 110
Source: Baukal, C.E. (ed.), The John Zink Combustion 0.25 or less 115
Handbook, CRC Press, Boca Raton, FL, 2001.
Source: Baukal, C.E. (ed.), The John Zink Combustion
permissible continuous noise level at the work place that Handbook, CRC Press, Boca Raton, FL, 2001.
will not lead to permanent hearing loss: Note: Exposure to impulsive or impact noise should
not exceed 140 dBA.
 24 
LA,m (dBA) < 85 + 10 log 10   (11.6) OSHA requires that the employer must provide protec-
 Tn 
tion against the effects of noise exposure when the sound
where Tn is the daily noise exposure time (h). levels exceed those shown in Table 11.7. When the daily
Wearing ear protection devices at continuous noise noise exposure consists of two or more periods of noise
levels greater than 85 dBA can prevent or reduce the exposure at different levels, their combined effect should
danger of permanent hearing damage. There are two be considered rather than the individual effects of each.
levels of protection commonly used by industrial According to OSHA the exposure factor (EF) is defined as5
workers to reduce noise levels: earplugs and ear muffs C1 C2 C3 C
as shown in Figure 11.17. Earplugs can lower the noise EF = + + ++ n (11.7)
T1 T2 T3 Tn
level by 5–45 dB, depending on the type of plug, sound
frequency, and how well the user inserts them into the where
ear. Ear muffs are designed to cover the entire ear and Cn is the total time of exposure at a specific noise level
typically reduce noise levels by 5–50 dB depending on Tn is the total time of exposure permitted at that level
the type of ear muff and the frequency of the sound. and shown in Table 11.7
11.2.1 OSHA Requirements If the EF exceeds 1.0, the employee’s exposure is above
OSHA limits. If OSHA identifies such a situation, a cita-
Title 29 CFR, section 1910.95 of the U.S. Occupational
tion may be issued and a grace period defined in which
Safety and Health Act (OSHA) pertains to the protection
the employer must correct the violation or face penalties
of workers from potentially hazardous noise. Table 11.7
as high as $10,000 per day.
shows OSHA permissible noise exposure levels.
11.2.2 International Requirements
Regulations aimed at protecting individuals from indus-
trial noise pollution have been enforced in almost all
industrialized countries. The noise caused in industries
and the work place is generally taken as a serious issue.
Most countries have adopted 85 dBA as the limit for
the permissible noise. At any work place with sound
levels exceeding 85 dBA, ear protection devices must
be worn and workers exposed to this level should have
their hearing level checked periodically.
11.2.3 Noise Sources and Environment Interaction
FIGURE 11.17 The predominant individual sources of noise in chem-

Typical earplugs and muffs. ical and petrochemical plants are burners (process
Noise 281
furnaces, steam boilers, and flares), fans, compres- air and achieve flammable mixture, tends to create a
sors, blowers, pumps, electric motors, steam turbines, slower combustion process. In either case, when regions
gears, valves, exhausts to open air, conveyors, and in the mixing process form a flammable mixture and
silos, airborne splash noise from cooling towers, coal encounter a source of ignition, combustion takes place.
mills, and loading and unloading of raw and finished The closer an air-fuel mixture is to stoichiometry prior
materials. to encountering an ignition source, the more rapid the
Although noise pollution caused by industrial sec- combustion.
tors is minor compared to that caused by road and rail When combustion occurs near stoichiometric condi-
traffic, industrial noise receives more attention due to tions, more of the energy released is converted into
public representation. ISO 1996 provides information noise. For example, Figure 11.18 shows a test flare
on how to measure and assess environmental noise, operating at the same flow rate of fuel, but with dif-
which can be used to help determine regulations for ferent degrees of mixing. The photograph on the left
noise protection in residential neighborhoods located shows a flame with a high degree of mixing between
near industrial areas. the fuel and ambient air; notice this produces a short,
National or local authorities must enforce noise lim- crisp flame that burns smokelessly. The photograph
its that should not be exceeded in the neighborhood. on the right, however, shows the same flare operat-
The magnitude of limiting values, additional charges for ing at the same fuel flow rate, but with poor mixing
tonality and impulsive noise, and the legalities change between the fuel and ambient air. For this condition,
not only from country to country, but sometimes within notice that the flame is not as crisp (softer) and is sig-
different states and regions in the same country. In gen- nificantly larger. Although the fuel flow rate is the
eral, nighttime noise limits are 10–15 dB lower than that same for each of these cases, the combustion noise
for the daytime. generated is significantly different. The well-mixed
flame (photograph on the left) is about 5 dB higher in
noise emissions than the poorly mixed flame (photo-
graph on the right).
The noise emitted from each small region of rap-
idly combusting mixture adds up to create what is
11.3 Mechanisms of Industrial called combustion roar. Therefore, combustion roar
Combustion Equipment Noise is largely a function of how rapidly the fuel is being
burned. In addition, in the context of combustion
There are four major mechanisms of noise production equipment like burners and flares, usually the larger
in combustion equipment. They can be categorized as the fuel flow rate, the more the turbulence in the
either predominantly high-frequency or low-frequency combustion process. Since turbulence directly influ-
sources. They are as follows: ences the mixing rate, high turbulence processes also
produce more combustion roar. Thus, it is more accu-
1. Low-frequency noise sources rate to state that the level of combustion roar gen-
a. Combustion roar and instability erated by a combustion process is a function of the
b. Fan noise amount of fuel burned and how rapidly one arranges
2. High-frequency noise sources to burn it.
a. Gas jet noise
b. Piping and valve noise
11.3.1.1 Flare Combustion Roar
It has been recognized for a long time that the noise
emitted from a normal operating flare has two mech-
11.3.1 Combustion Roar and Combustion
anisms at work; namely, combustion roar and gas jet
Instability Noise
noise. Combustion roar typically resides in the lower
To better understand combustion roar, the mixing pro- frequency region of the audible frequency spectrum,
cess taking place between the fuel and the oxidant on a while gas jet noise occurs in the higher frequencies, as
very minute scale is considered. It is known that a well- illustrated in Figure 11.19.
blended mixture of fuel and air will burn very rapidly As previously mentioned, the amount of combus-
if the mixture is within the flammability limits for that tion roar emitted from a flare, generally, depends on
fuel. On the other hand, a raw fuel stream that depends how fast the waste gas stream mixes with the ambient
on turbulence and momentum to mix in the ambient air. A waste gas stream that exits a flare tip with a low
5m
5m
FIGURE 11.18
Test flare at John Zink test site in Tulsa, OK. Combustion of identical fuel flow rates with different degrees of mixing. Total noise emissions of
the flame in the photograph on the left are about 5 dB higher than that of the flare flame in the photograph on the right.
velocity flames are referred to as momentum-domi-

nated flames. Increasing the rate at which the waste
gas burns, results in “bigger explosions” of the air-fuel
mixture. These “bigger explosions” create larger dis-
turbances in the atmosphere resulting in higher levels
of combustion roar.
Noise level
Combustion roar emitted from a flare flame is not

Region of Region of
combustion gas jet
highly directional and is considered to be a monopole
roar noise source. That is, it is analogous to a spherical balloon
whose surface is expanding and shrinking at various fre-
quencies and emitting uniform spherical waves.
High levels of turbulence in a flare flame are usually
desirable because it helps reduce radiation and increase
the smokeless capacity of the flare. Unfortunately, high
Noise frequency (Hz) levels of turbulence increase combustion roar. Unlike
the solution for flare radiation reduction, it is not prac-
FIGURE 11.19 tical to increase the height of a flare stack or boom to
Typical noise signature emitted from a flare. (From Baukal, C.E. (ed.), reduce combustion noise because even doubling the
flare stack height would reduce the SPL at the flare base
by only about 6 dB (see Equation 11.4). In addition, com-
velocity and low levels of turbulence will mix slowly bustion roar is a low-frequency sound that can travel a
with the ambient air and burn relatively quietly. These great distance without being substantially attenuated by
types of flames are called buoyancy-dominated flames. the atmosphere. The noise signature of low-frequency
Conversely a waste gas stream that exits a flare tip combustion roar typically consists of a broadband spec-
with a high velocity and high levels of turbulence will trum with a single peak.
burn much faster and create substantially more com- Flare flames emit combustion noise over a broad
bustion noise for the same heat release rate. These high spectrum of frequencies. The maximum (peak) noise
Noise 283
TABLE 11.8
Calculation of the Typical
Combustion Noise Spectrum of a
Stable Burning Flare from the Overall Low pressure flare High pressure flare
Sound Pressure Level (OASPL) (low level of turbulence) (highly turbulent)
Resultant Noise
Frequency (Hz) Spectrum (dB)
31.5 OASPL-5
63 OASPL-4
125 OASPL-9
250 OASPL-15
500 OASPL-20
1000 OASPL-21
2000 OASPL-24
4000 OASPL-28
8000 OASPL-34
level typically occurs at a frequency of about 63 Hz.

If one knows the OASPL emitted from a flare, the
combustion noise spectrum can be estimated by

subtracting the values shown in Table 11.86 from the
OASPL. Notice that at frequencies above about 500 Hz,
the noise contribution from flare combustion becomes
relatively insignificant.
A typical method for estimating the sound power
level in dB (PWL) emitted from a flare flame is to FIGURE 11.20
relate the energy released from the combustion of Photograph of a high-pressure and low-pressure flare burning the
the waste gas stream (chemical energy) to the noise same fuel.
energy liberated by the combustion. The ratio of noise
energy to chemical energy released from the com-
bustion is called the thermoacoustic efficiency (TAE).
For a stable burning flare, the TAE typically varies
between 1 × 10 −9 and 3 × 10 −6. The value of the TAE
largely depends on the turbulent mixing of the waste
gas with ambient air; the faster the mixing, the higher
the TAE.
A flare flame that is highly turbulent, such as the
high-pressure flare in Figure 11.20, can have a TAE on
the order of 1 × 10−6. However, flames with low levels
of turbulence, such as low-pressure flares or a butane
lighters (Figure 11.21) for example, can have a TAE on
the order of 1 × 10−9. For every order of magnitude that
the TAE changes, the SPL will change by 10 dB. For
example, the difference in SPL for a TAE of 1 × 10−6 and
1 × 10−9 is 30 dB.
Figure 11.22 is a photograph of an engineer collect-
ing noise levels from a flare using a noise meter. Data
collection is important because it allows for the TAE
of flares to be determined; this information can then
be used to model the level of combustion roar emitted
from a flare. To demonstrate how the TAE is deter-
FIGURE 11.21
mined from experimental data consider the following Shadow photograph of a burning butane lighter. (From Baukal, C.E. (ed.),
example: The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
W(watts)
  Lp + 10 log 10 ( 4πr 2 ) − 10.5  

= 1 × 10 −12 antilog 10  
  10  

(11.8)
where
Lp is the SPL in dB (100 dB for the example)
r is the distance from the flame in feet (400 ft for
the example)
Substituting these values into Equation 11.8 gives

W = 1792 W. The TAE is then calculated to be
Acoustical power
TAE =
Thermal power
FIGURE 11.22 1792 W
Engineer measuring flare noise. (From Baukal, C.E. (ed.), The John = = 1.2 × 10 −6
1465 × 10 6 W
Example 11.2
Since the TAE is on the order of magnitude of
Given: A flare burning a waste gas stream with a 1 × 10−6, one would expect that the flame would
heat release of 5000 × 106 Btu/h (1465 MW). Noise be highly turbulent and momentum-dominated.
measurements show that the SPL 400 ft (120 m)
from the flame is 100 dB. Estimate the TAE of the
flare flame. Figure 11.23 shows data obtained from field mea-
Solution: The sound power emitted from the surements of the PWL emitted from a number of
flame, W, can be determined as follows (see also industrial flares operating at different loads.7 This
Equations 11.3 and 11.4): data includes combustion roar as well as the noise
175
EF
EF-SSA
165
EF-SSS
155
LwA (dB)
145
EF-SSS-NC
135
GF-SSS-NC
125
TAE = 1.9E-8
115 TAE = 4.5E-5
10 100 1,000 10,000
Qcombust (MW)
FIGURE 11.23
PWL LW calculated from measured noise data, plotted versus heat release rate, Q̇combust, for different types of industrial flares under various
operating conditions. EF, elevated (single-point) flare; GF, (enclosed) ground flare; SSA, smoke suppression by air; SSS, smoke suppression by
steam; NC, equipped with advanced noise control.
Noise 285
POI@ 100 m POI@ 100 m POI@ 100 m
POI@ 50 m POI@ 50 m POI@ 50 m
(a) (b) (c)
>60.0 dB (A) >80.0 dB (A) >100.0 dB (A)

>65.0 dB (A) >85.0 dB (A) >105.0 dB (A) Point source
>70.0 dB (A) >90.0 dB (A) >110.0 dB (A) Barrier
>75.0 dB (A) >95.0 dB (A) >115.0 dB (A) Receiver
FIGURE 11.24
Predicted sound pressure field contour plots for a multipoint LRGO flare system. (a) First stage in operation; (b) stages 1 and 2 in operation;
(c) all stages in operation.
emitted from valves, injectors, smoke suppression not only of the individual burners and their combustion
devices, etc. during data collection. The two lines and jet noise emissions, but also of the arrangement of
in the figure represent constant TAE corresponding the burners in the flare pit and the effect of the radiation
to the maximum and minimum values of the data set; fence surrounding the burner array. Figure 11.24 shows
the minimum TAE is 1.9 × 10 −8 and maximum TAE is the calculated sound pressure field contour plots for a
4.5 × 10 −5. A central problem with using this approach multipoint, so-called linear relief gas oxidizer (LRGO)
for determining the TAE lies in applying it to a whole flare system at different operating conditions.
flare system instead of the combustion process alone. In summary, it can be concluded that the TAE con-
Doing so means that TAE becomes a “lumped param- cept applied to flare systems will usually only allow a
eter” incorporating all effects that have an impact very rough estimate of the actual noise emissions and
on the acoustical behavior of the flare, i.e., noise the associated effect of these emissions on persons in
emissions from valves, injectors, smoke suppression the neighborhood of the flare. In some cases, the results
devices, etc. as well as any noise control measures can be dramatically different than the actual situation.
installed. Alternative approaches to evaluating TAE A key issue in developing reliable noise prediction tools
can be found in the literature.8 for flares that are more generally applicable in a broad
Another problem with using the TAE concept to range of operating conditions lies in a proper treatment
predict combustion noise is that it does not take the of the individual sources that contribute to the overall
frequency characteristics of the noise emissions into noise emissions of a flare system.
account. Frequency characteristics are very important
for the design of adequate noise control measures and
11.3.1.2 Flare Combustion Instability Noise
will also determine the human perception of the noise
emissions since the ear’s sensitivity for noise is fre- If a flame lifts too far above a flare tip, it can become unsta-
quency-dependent (see Section 11.1). ble. An unstable flame will periodically lift and then reat-
Additional challenges arise for spatially extended flare tach to the flare tip and create a low-frequency rumbling
systems that can no longer be treated as point sources. noise. Typically, this rumbling noise occurs in the fre-
Prediction of the noise emissions of a grade-mounted quency range of 5–10 Hz and is usually called combustion
multipoint flare system requires the correct modeling instability. Being as low in frequency as it is, combustion
110
Over-steamed flare
105
100
Noise level (dB)

95
90
Normal operating steam

85
flare
80
75
70
10 100 1000
Frequency (Hz)
FIGURE 11.25
SPL emitted from a steam-assisted flare operating at normal conditions and at over-steamed conditions (combustion instability). (From Baukal,
C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
instability noise is usually inaudible, but can travel over velocity of the air-fuel mixture, and various turbu-
several miles without being substantially attenuated by lence parameters such as the turbulent length scale and
the atmospheric air. When there are reports of shaking the intensity. This section does not attempt to discuss these
walls and windows of buildings in the vicinity of a flare, it prediction techniques in detail, but gives a broad and
is usually due to combustion instability. general discussion of combustion roar and combustion
If too much steam or air is added in the base of a flare instability noise using some of the results from these
tip, the waste gas stream can be over-aerated causing the studies.
flame to periodically lift and reattach at the tip. This peri- Figure 11.14 is a plot showing a typical noise spectrum
odic lifting and reattachment of the flame from the flare emitted from a burner operating under normal condi-
tip is the mechanism responsible for creating the low- tions in a furnace. Notice that the noise spectrum has two
frequency rumbling noise. Combustion instability noise peak frequencies associated with it; the high-frequency
can usually be reduced by lowering the steam flow rate noise contribution is from the fuel gas jets while the low-
to a steam-assisted flare or by lowering the blower air frequency contribution is from the combustion roar. As
flow rate to an air-assisted flare. Figure 11.25 graphically with combustion roar emitted from flares, burner com-
depicts a typical steam-assisted flare operating under bustion roar is associated with a smooth broad band
both normal conditions as well as over-steamed condi- spectrum having relatively low conversion efficiency
tions.9 Note that the combustion noise frequency shifts from chemical energy to noise (TAE): in the range of 1 ×
substantially to a lower region and the level dramatically 10−9 to 1 × 10−6. However, the combustion noise spectrum
increases when the flare is over-steamed. associated with a burner and a flare is not similar. The
reason is that a flame burning in the open atmosphere
will behave differently compared to a flame that is burn-
11.3.1.3 Burner Combustion Noise
ing in an enclosed chamber such as a furnace.
Like flares, burner combustion noise is an unwanted The combustion roar associated with flares typically
sound associated with combustion roar and combustion peaks at a frequency of approximately 63 Hz while the
instability. In many situations, the combustion noise can combustion roar associated with burners can vary in the
be the dominant source of noise emitted from a burner. 200–500 Hz range. Burner noise can have a spectrum
Combustion roar and combustion instability are quite shape and amplitude that can vary with many factors.
complex by nature. The literature contains a variety of These factors include the (1) internal shape of the fur-
combustion noise and combustion instability prediction nace, (2) design of the burner muffler, plenum, and tile,
techniques for burners operating in a furnace. Most of (3) acoustic properties of the furnace lining, (4) trans-
these prediction techniques are based on experimental mission of the noise into the fuel and air supply system,
studies that attempt to correlate the acoustic power radi- and (5) transmissive and reflective characteristics of the
ated by the burner/furnace geometry, laminar burning furnace walls and stack.
Noise 287
11.3.1.4 Burner Combustion Instability Noise of the turbulence with the solid constructed parts of the
fan create the broadband noise. This broadband noise is
Combustion instability within a furnace is characterized
of the dipole type, meaning that the noise is directional.
by a high amplitude low-frequency noise often resem-
On the other hand, the discrete tones are created by the
bling the puffing sound of a steam locomotive. This
periodic interactions of the rotating blades and nearby
type of noise can create significant pressure fluctuations
upstream and downstream surfaces. Discrete tonal
within a furnace that can cause damage to the structure
noise is usually the loudest at the frequency at which a
and radiate high noise levels to the surroundings.
blade passes a given point. The tonal frequency is eas-
Figure 11.26 is a plot showing the SPL for a gas burner
ily calculated by multiplying the number of blades times
operating under normal conditions and with instabil-
the impeller rotation speed in revolutions per second.
ity. It is obvious that the SPL increases substantially
The broadband and discrete tonal noise emitted
when the operation is accompanied by instability.
from fans can radiate from both the suction and pres-
Combustion instability noise has a high efficiency of
sure side of a fan and through the fan casing. The
conversion of chemical energy to noise (TAE). Typically
noise can radiate downstream through the ducting
the TAE from burner combustion instability is in the
and discharge into the environment at an outlet. Fan
range of 1 × 10−4.10
and duct systems should include provisions to con-
The oscillations caused by combustion instability are
trol this noise if residential areas are located nearby.
naturally damped by pressure drop losses through the
Installation of mufflers and silencers on the suction
burner and furnace, and therefore cannot be sustained
and the discharge sides of the fan, as well as wrapping
unless energy is provided. These steady oscillations are
of the casing and the ducts are common methods for
sustained by energy extracted from the rapid expansion
reducing fan noise.
of the air-fuel mixture upon reaction. Over the years,
furnace operators have used several techniques in an
attempt to eliminate combustion instability. Some of 11.3.3 Gas Jet Noise
these techniques include modifying the (1) furnace stack
height, (2) internal volume of the furnace, (3) acousti- Gas jet noise is very common in the combustion indus-
cal properties of the furnace lining, (4) pressure drop try and in many instances it can be the dominant noise
through the burner by varying the damper position, (5) source within a combustion system. The noise created
fuel port diameter, (6) location of the pilot, and (7) flame when a high-speed gas jet exits into an ambient gas usu-
stabilization techniques. ally consists of two principal components: gas jet mix-
ing noise and shock-associated noise.11
11.3.2 Fan Noise

11.3.3.1 Gas Jet Mixing Noise
The noise emitted from industrial fans typically consists
of two noise components: broadband and discrete tones. Studies have shown that a high-speed gas jet exiting
Vortex shedding of the moving blades and the interaction a nozzle will develop a large-scale orderly pattern as
115
110
With instability
105
100
SPL (dB)
95
90
85
Normal operation
80
75
70
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
Frequency (Hz)
FIGURE 11.26
Burner SPL normal and with instability. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Region of
maximum jet
mixing noise 30°
15°
Jet centerline
Nozzle 0°
Locations where orderly patterns are developing 15°

FIGURE 11.27
Development of orderly wave patterns within a high-speed gas jet. 30°
(From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Region of
Press, Boca Raton, FL, 2001.) maximum jet
mixing noise
shown in Figure 11.27. This orderly structure is known
FIGURE 11.28
as the “global instability” or “preferred mode” of the jet. Illustration showing the region of maximum jet mixing noise. (From
The presence of both the small-scale turbulent eddies Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
within the jet and the large-scale structure is respon- Boca Raton, FL, 2001.)
sible for the gas jet mixing noise.
pressure, the gas jet develops a structure of shock waves
The source of gas jet mixing noise begins near the
downstream of the nozzle as shown in Figure 11.29. The
nozzle exit and extends several nozzle diameters down-
critical pressure of a gas jet typically occurs at a pres-
stream. Near the nozzle exit, the scale of the turbulent
sure of 12–15 psig (0.8–1 barg), depending on the gas
eddies is small and predominantly responsible for the
composition and temperature. These shock cells consist
high-frequency component of the jet mixing noise.
of compression and expansion waves that repeatedly
The lower frequencies are generated further down-
compress and expand the gas as it moves downstream.
stream of the nozzle exit where the large-scale orderly
Using Schlieren photography, several investigators have
pattern of the gas jet exists.
seen as many as seven shock cells downstream of a noz-
Gas jet mixing noise consists of a broadband frequency
zle. These shock cells are responsible for creating two
spectrum. The frequency at which the spectrum peaks
additional components of gas jet noise: screech tones
depends on several factors including the (1) diameter of
and broadband shock-associated noise.
the nozzle, (2) Mach number of the gas jet, and (3) tem-
Screech tones are distinct narrow-band frequency
perature ratio of the fully expanded jet to the ambient
sounds that can be described as a “whistle” or “screech.”
gas. In the flare and burner industry, gas jet mixing noise
The literature reports that these tones are emitted from
typically peaks somewhere between 2,000 and 16,000 Hz.
the fourth and fifth shock cells downstream of the noz-
The overall SPL created by gas jet mixing depends on
zle exit as shown in Figure 11.30.12
several variables including the (1) distance from the gas jet,
(2) angle of the observer relative to the gas jet centerline,
Nozzle
(3) Mach number, (4) fully expanded gas jet area, and (4) exit plane
density ratio of the fully expanded jet to the ambient gas.
The maximum overall SPL of gas jet mixing noise occurs
at an angle between approximately 15°–30° relative to the
centerline of the gas jet as illustrated in Figure 11.28.11
As one moves in either direction from this angle, the
noise level typically drops off significantly. For example,
the overall SPL created by gas jet mixing can be reduced
as much as 25 dB when one moves from an angle of
maximum noise level (15°–30°) to an angle directly
behind the nozzle (180°).
Shock waves
11.3.3.2 Shock-Associated Noise
FIGURE 11.29
When a flare or burner operates above a certain fuel
Photograph showing shock waves downstream of an air jet. (From
pressure, a marked change occurs in the structure of Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
the gas jet. Above a certain pressure called the critical Boca Raton, FL, 2001.)
Noise 289
Shock cells (diamonds)
1 2 3 4
Shock noise from fourth shock cell
Nozzle exit
Noise level
Broadband
Broadband Screech tones

Shock associated
noise spectrum
Mixing noise
spectrum
Noise frequency
FIGURE 11.30
Screech tone emissions.
The sound waves from these shock cells propagate radiate downstream through the pipe and exhaust into
upstream, where they interact with the shear layer at the environment at an outlet and/or radiate through the
the nozzle exit. This interaction then creates oscillating pipe wall not only into the space near the valve itself,
instability waves within the gas jet. When these insta- as illustrated in Figure 11.31, but also at relatively large
bility waves propagate downstream they interfere with distances from the valve.
the fourth and fifth shock cells causing them to emit the Usually, butterfly valves and ball valves are noisier
screech tones. Screech tone noise is not highly direc- than globe valves. Butterfly valves and ball valves
tional (monopole noise source), unlike gas jet mixing
noise. Sound radiating from flare
tip to surroundings
Broadband shock-associated noise occurs when
the turbulent eddies within the gas jet pass through
shock waves. The shock waves appear to suddenly
distort the turbulent eddies which creates a noise
that can range over several octave bands. The broad- Sound radiating from valve
to surroundings
band, shock-associated peak frequency noise typically
occurs at a higher frequency than the screech tone
peak frequency.
Pressure
relief or steam
valve
11.3.4 Valve and Piping Noise
When a gas flowing steadily in a pipe encounters a valve,
a change in the flow pattern and pressure will occur that
can create turbulence and shock waves downstream of
the valve. Typically, when valves are partially closed, Low pressure
High pressure
side
creating a reduction in flow area, the small flow passage side
behaves much like an orifice and produces jet noise. As FIGURE 11.31
discussed earlier, turbulence and shock waves create Noise radiating from a valve. (From Baukal, C.E. (ed.), The John Zink
mixing noise and shock-associated noise. This noise can Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
typically have a smaller vena contracta than a globe roar and gas jet noise. Inhibiting the rate at which the
valve operating at the same pressure drop which air and fuel streams mix can reduce the level of com-
results in higher levels of mixing and shock-associated bustion roar; however, this noise abatement technique
noise. As a general guideline when the pressure ratio generally tends to reduce the smokeless performance
across a valve is less than approximately 3, the mix- and increase thermal radiation and flame length.
ing noise and shock-associated noise are within about Reducing the mixing rate of the air and fuel stream in
the same order of magnitude. However, for pressure order to lower combustion roar levels usually does not
ratios greater than 3, shock noise usually dominates justify the accompanying sacrifices in the performance
mixing noise.13 There are several methods used for of a flare.
reducing the noise emitted from a valve. These include In such cases, enclosed flares may provide one solu-
sound-absorptive wrapping of the pipes and the valve tion. Enclosed flares are designed to completely hide a
casings and the installation of silencers between the flare flame in order to reduce noise and thermal radia-
valve and the connected pipes. In addition, special tion levels. The design of these flare systems typically
low-noise valve designs with multiple pressure drop consists of an insulated enclosure with a wall around the
stages exist that generate less noise than standard air inlets, as shown in the photograph in Figure 11.32.
designs. These types of flares can substantially reduce noise
emissions as compared to open elevated flares.
There are several abatement techniques commonly
used to reduce the gas jet noise emitted from flares.
Such techniques include mufflers, water injection, and
modifications to the nozzle geometry. Mufflers are most
11.4 Noise Abatement Techniques commonly used on steam-assisted flares to abate the
There are three places noise can be reduced: at the source, high-pressure steam jet noise as shown in Figure 11.33.
in the path between the source and personnel, and on the In most flare systems, steam is supplied to nozzles at a
personnel.14 The ideal place to stop noise is at the source. pressure of 100–150 psig (7–10 barg). These high-pressure
There are several techniques used in the flare and burner steam jets produce high-frequency mixing and shock-
industry to reduce the noise at the source, but these tech- associated noise. A number of flare muffler styles have
niques have limitations. Ear protection can reduce noise been used in the industry with varying degrees of noise
relative to the personnel using it; unfortunately, a plant abatement performance. Many of these mufflers are
operator cannot ask a surrounding community or work- designed with a fiber material several inches thick placed
ers within a nearby office building to wear ear protection on the inside. Mufflers usually do a good job of absorbing
when the noise levels become a problem. The most com- the high-frequency steam jet noise, as demonstrated by
mon method for reducing noise is in the path between the data in Figure 11.34. This plot shows the noise spec-
the source and personnel, using silencers, plenums, and trum emitted from a steam-assisted flare operating with
mufflers. The purpose of this section is to discuss the and without a muffler on the lower steam jets. The data
most common and effective noise abatement techniques
utilized in the flare and burner industry.
Enclosed flare helps
reduce noise
11.4.1 Flare Noise Abatement Techniques
The following individual sound sources can contribute
to the overall noise emissions of flares:
1. Combustion process
2. Gas jet and flow noise of the released gases
3. Smoke suppression equipment
a. Jet and flow noise by steam and air injection Wall at bottom of
enclosure helps
b. Combustion air fans reduce noise
4. Pilot burners
5. Valves (gas and steam side) and connected piping
FIGURE 11.32
As previously discussed the two principal sources of Photograph of two enclosed flares. (From Baukal, C.E. (ed.), The John
noise emitted from industrial flares are combustion Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Noise 291
Muffler
Steam nozzle
Muffler surrounding steam
nozzles
FIGURE 11.33
A steam-assisted flare with a muffler.
100
95
90
SPL (dB)
85
80
Without muffler
75
With muffler
70
100 1,000 10,000 100,000
Frequency (Hz)
FIGURE 11.34
Steam jet noise emitted with and without muffler.
clearly show that mufflers are more efficient at absorb- high-pressure flares by injecting water into the waste gas
ing the higher noise frequencies than the lower ones. stream near the flare nozzle exit (see Figure 11.36).15 The
Instead of using a muffler, noise from steam injector water injection method appears to substantially reduce
nozzles in steam-assisted flares can also be decreased by the shock-associated noise as shown in Figure 11.37. This
reducing the amount of steam required to ensure smoke- plot depicts the noise spectrum emitted from a John Zink
less combustion. This method of (indirect) noise control high-pressure flare operating with and without water

does not call for any mufflers or other sound absorbing injection. Schlieren photography shows that water injec-
devices: its working principle is that less noise is gener- tion does not eliminate the downstream shock cell struc-
ated because of the lower steam pressure and flow rate. ture, but does appear to alter its appearance. This suggests
Figure 11.35 shows the PWL spectra of the noise emit- that water injection suppresses the feedback mechanism
ted from two types of steam-assisted flares. The data responsible for growth of the gas jet instability that leads
are for the two flares burning the same type and flow to screech tones.
rate of fuel with the steam flow adjusted as to just obtain Gas jet noise reduction using water injection is more
smokeless combustion. To do so, the more efficient XP pronounced when flaring high-molecular-weight gases
flare tip uses less steam and, therefore, emits less noise. as compared to low-molecular-weight gases at the same
In high-pressure flaring applications, gas jet noise can be operating pressure. Test data show that high-molecular-
the major source of noise. In 2000, the John Zink Company weight gases are more dominated by screech tone noise
developed a unique method for reducing gas jet noise from than low-molecular-weight gases operating at the same
From measurements/calculations the A-weighted gas jet noise level. Gas jet noise emitted
140
from high-pressure flares usually peaks at a frequency
Flare tip between approximately 2,000 and 16,000 Hz. The peak
Standard (133 dBA) frequency is a function of several variables, but is most
XP (124 dBA) affected by the diameter of the nozzle. For example, a
120 1-in. (25 mm) diameter gas jet nozzle will peak at a fre-
quency between 2000 and 4000 Hz, whereas a 1/4-in. (6.4
LWA (dBA)
mm) diameter gas jet nozzle will peak between 8,000 and
16,000 Hz. To the human ear, a group of several smaller-
100 diameter gas jet nozzles will appear quieter than a single
larger nozzle operating at the same pressure and mass
flow rate; the primary reason being that the group of
smaller nozzles will peak at a higher frequency, where the
human ear is less sensitive. Designing a flare with many
80
small-diameter nozzles is not always practical or eco-
63 125 250 500 1k 2k 4k 8k
nomical to build. Some large-capacity flare designs would
Frequency (Hz) require several thousand nozzles to substantially reduce
FIGURE 11.35 the gas jet noise.
Example for noise abatement in steam-assisted flares by reduc-
ing the amount of steam required to ensure smokeless combustion.
A-weighted PWL for a standard flare tip (overall PWL 133 dBA) and a 11.4.2 Burner Noise Abatement Techniques
modern John Zink Company XP flare tip (124 dBA). Values measured
(XP) and calculated based on measurements (standard tip).
Typical sound sources contributing to the overall noise
emissions of burners used in industrial heaters and fur-
pressure, which explains why gas jet noise reduction, naces are as follows:
using water injection, is more pronounced when flaring
high-molecular-weight gases. • Combustion process
It is very common in the flare industry to design a flare • Gas jet noise of the fuel gas at the burner
using several small-diameter nozzles in order to reduce
• Pilot burners
• Combustion air fan
• Steam or air injection for fuel atomization
• Control valves and connected pipes
The most important of the aforementioned sound

sources, typically, are the (1) combustion roar from
the combustion process itself, which resides in the fre-
quency range of approximately 100–1000 Hz, and the (2)
gas jet noise, which typically ranges between 4,000 and
16,000 Hz. The mid-to-high-frequency noise is the most
annoying and damaging to the ear. Several techniques
have been used to suppress the noise emitted within the
mid-to-high frequencies. Four common techniques used
to reduce noise in industrial burners are the following:
1. Sound absorption in the burner plenum

2. Mufflers at air inlets of natural draft burners
3. Acoustically optimized furnace wall construction
4. Acoustical treatment of the air ducts in forced
draft burners
Figure 11.38 shows a plot of the SPL as a function of the

frequency for a burner operating with and without a
FIGURE 11.36 muffler. Clearly, without the muffler, the noise level is
Water injected into a high-pressure flare. higher especially in the higher frequency region.
Noise 293
110
105
100
Noise level (dB)

95
90
85
80
No water injection, 116 dB, 114 dBA
75
Optimum water injection, 113 dB, 101 dBA
70
10 100 1,000 10,000 100,000
Frequency (Hz)
FIGURE 11.37
Noise spectrum from a high-pressure flare with and without water injection. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
85
Unmuffled
80 A-weighted
75 burner noise
70
SPL (dBA)
65
60
55
50
Muffled A-weighted
45 burner noise
40
31.5
63
125
250
500
1,000
2,000
4,000
8,000
16,000
Frequency (Hz)
FIGURE 11.38
Sound pressure versus frequency for a burner operating with and without a muffler. (From Baukal, C.E. (ed.), The John Zink Combustion
The performance of burner mufflers can be opti- common plenum chambers for groups of burners can be
mized through their size, but also through design. used as shown in Figure 11.40. Properly designed plenum
For example, Figure 11.39 shows a muffler located at chambers can provide a reduction in burner noise emis-
the air intake of a natural draft burner in two differ- sions in addition to the reduction achieved by the mufflers.
ent designs. Although these two mufflers are similar
in size, the muffler shown on the left side will provide
11.4.3 Valve and Piping Noise
better performance than the muffler shown on the
Abatement Techniques
right side. By designing the muffler with a 90° turn,
the sound wave is required to impact the muffler wall Valve and piping noise abatements include sound-
and change directions several times before exiting the absorptive wrapping of the pipes and valve casings,
muffler. Each time the sound wave impacts the wall installation of silencers between the valves and the
and changes direction, energy is lost resulting in lower connecting pipes, and the use of low-noise valve
noise emissions. designs with multiple pressure stages. Acoustical pipe
When space is limited around a furnace, designing each lagging is similar to thermal pipe insulation. However,
burner with a muffler can be a problem. In some cases whereas acoustical pipe lagging also provides excellent
Better Often unfavorable
Gas nozzle
Silencer
FIGURE 11.39
Illustration showing two different muffler designs. Although the mufflers are similar in size, the design on the left side provides better noise
reduction performance than the design on the right side.
Furnace
Floor burners
Silencer Silencer
Air Air
Plenum
chamber
Air Air
FIGURE 11.40
Illustration showing a common plenum chamber for floor burners in a furnace.
thermal insulation, many thermal insulations provide

only poor noise control. Rigid insulations for cold ser-
vice (such as foam glass installed on smaller-diameter
pipes) can actually aggravate the noise situation by
conducting the noise to the outer surface. Although
Valve noise
acoustical energy radiated per unit area of insulated
and jacketed pipe is less than for the same noninsu-
Without noise control
lated pipe, the surface area of an insulated and jacketed
pipe is greater. The product of these two factors can
SPL (dBA)
cause larger-diameter jacketed pipes to radiate more

noise than bare pipes.16
Piping requiring acoustical treatment in a typical pet-
rochemical plant is often in cold service. These lagging
systems have to be both thermal and acoustical insu-
lators. For that reason, fibrous insulation followed by With noise control
an outer leaded aluminum jacket is commonly used.
Sometimes, very noisy pipes need a layer of impreg- 63 125 250 500 1k 2k 4k 8k
nated vinyl sandwiched between layers of fibrous insu- Frequency (Hz)
lation, called a septum system.16
FIGURE 11.41
Figure 11.41 shows an example of noise emitted from Noise emissions from a steam control valve. “Without noise control”
a steam control valve located at the base of an elevated, refers to an uninsulated pipe equipped with a butterfly valve. “With
steam-assisted flare. In the original configuration, a noise control” refers to insulated pipe equipped with a low-noise valve.
Noise 295
standard butterfly valve is used and neither the valve nor Furnace wall
the connecting steam piping is insulated. During these
tests the waste gas flow rate was very low; therefore the Burner Burner Burner
flare noise emissions were not significant. The data show A B C
that the noise emissions are dominated by the valve noise
1A 1B 1C
with a peak frequency of 4000 Hz. After installing a low-
noise valve and acoustically insulating the steam pipes,
the noise emissions at frequencies of 1000 Hz and above 3 ft 5 ft
are significantly reduced. Noise emissions at 500 Hz and
below remain unchanged because these are dominated 2
by the combustion noise coming from the flare tip.
11.4.4 Fan Noise Abatement Techniques

Fan noise can usually be addressed similar to valve and FIGURE 11.42
piping noise: Illustration used for burner noise example. (From Baukal, C.E. (ed.),
1. Silencers can be installed at the suction and Solution: First, find the PWL, LW, emitted from
pressure sides of the fan particularly for fans each burner, assuming that the noise is emitted at
communicating with the atmosphere on either the muffler exit at points 1A, 1B, and 1C. Assume
the suction or the pressure side and thereby cut that the noise spreads over a uniform sphere from
down on noise coming out of these portals. each of these points. The PWL can be calculated
as follows (see Equation 11.4):
2. Acoustically enclose the fan casing to address
noise radiated from or transmitted through the  1 
Lw = Lp − 10 log 10  − 10.5 (11.10)
casing surface.  4πr 2 

3. Acoustically isolate the ductwork leading to
and from a fan. where
Lp is the SPL
At the design stage, one may consider the use of low- r is the distance from the source (ft)
noise motors and the use of impellers with more blades
The noise level 3 ft (1 m) from burner B (location 2)
and reduced tip speed, etc. is 85 dB when it is operating alone. From Equation
11.10, LwB = 95.03 dB. Assuming that all burners are
operating at the same conditions, the PWL must be
95.0 dB for each one. The SPL contribution, Lp, can
now be calculated at location 2 when burner A is
11.5 Analysis of Combustion operating alone by solving Equation 11.10 for Lp:
Equipment Noise
11.5.1 Multiple Burner Interaction  1 
Lp = Lw + 10 log 10  + 10.5 (11.11)
 4πr 2 

A burner manufacturer will typically guarantee a burner
noise level at a location 3 ft (1 m) directly in front of the For this case LwA = 95.03 and r = (52 + 32)0.5 = 5.83 ft.
muffler. When several burners are installed in a furnace, Substituting these values into Equation 11.11 gives
however, the noise level 3 ft (1 m) from the burner may be LpA = 79.2 dB. This is the SPL contribution emitted
higher than for a single burner due to the noise contribu- from burner A measured at location 2. Since the
tion from surrounding burners. The purpose of this sec- distance from burner C to location 2 is the same, we
tion is to give an example that illustrates the noise level know that the SPL contribution from burner C at
increase due to noise emitted from surrounding burners. location 2, LpC, is also 79.2 dB. The total SPL at loca-
tion 2 can be determined by adding the SPL contri-
Example 11.3 bution from each burner (79.2 dB + 79.2 dB + 85 dB).
Given: Assume a furnace with a simple burner con-

figuration, as illustrated in Figure 11.42, with burner ( )
Lptotal = 10 log 10 10 0.1LpA + 10 0.1LpB + 10 0.1LpC = 86.8 dB
B operating alone, and the noise level is 85 dB at
location 2.
For this example, the noise level will be approxi-
Find: How is the noise level determined at loca- mately 1.6 dB higher when all the burners are
tion 2 when all burners are operating? operating than if burner B is operating alone.
11.5.2 High-Pressure Flare Notice that the gas jet mixing noise is a broadband
frequency spectrum, while the screech noise occurs
Figure 11.43 is a plot showing the SPL spectrum of a
over a fairly narrow bandwidth.
high-pressure flaring event burning natural gas in a
The screech noise would not exist if the flare were
3.5 in. (8.9 cm) diameter flare tip. The symbols and the
operated below the critical gas pressure. Below the
lines represent the noise spectrum gathered using a
critical gas pressure, shock waves, which cause screech
real-time sound-level meter and mathematical mod-
noise, do not form. The summation of the combustion
eling results, respectively. The SPL spectrum consists
roar, gas jet mixing noise, and screech noise provides
of two major peaks; a low-frequency peak that corre-
the total SPL prediction emitted from the flare.
sponds to the combustion roar and a high-frequency
The OASPL determined experimentally and calcu-
peak that corresponds to the gas jet noise. The inter-
lated using the mathematical model is summarized
mediate peak is a result of piping and valve noise.
in Table 11.9. Notice that in this particular example
Notice that the combustion roar peaks at a frequency
the OASPL, on a dBA scale, is dominated by the gas
of approximately 63 Hz, which is typical for a stable
jet noise. If this 3.5 in. (8.9 cm) diameter flare were
burning open flare.
designed with several smaller-diameter ports, having
Figure 11.44 is a plot showing the noise contribu-
the same total exit area, then the gas jet noise would
tions separately based on the mathematical model.
120
115
110
105
Noise level (dB)
100
95
90
85
Experiment: 115 dB, 109 dBA
80
Model: 114 dB, 110 dBA
75
70
10 100 1,000 10,000 100,000
Frequency (Hz)
FIGURE 11.43
SPL spectrum for high-pressure flaring. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
110
105
100
Total
Noise level (dB)
95
Screech tones
90
85
Jet mixing Combustion
80 roar
75
70
10 100 1,000 10,000 100,000
Frequency (Hz)
FIGURE 11.44
Noise contributions separately based on the mathematical model. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca
Raton, FL, 2001.)
Noise 297
TABLE 11.9 64 dBA) amounts to the sound intensity reduction equal

OASPL Determined Experimentally and Using the to one-tenth of its intensity at 3000 ft (910 m) without
Mathematical Model atmospheric attenuation. Hence, it should be noted that
measurements at large distances may vary significantly
Jet Mixing Screech Combustion
Noise Noise Roar Total on different days for the same equipment, if the atmo-
spheric conditions are significantly different.
Model
The effect of wind is even stronger and the SPL at
dB 105.7 105.1 113.0 114.3
1000 m (3300 ft) distance from a noise source may differ
dBA 105.2 105.3 97.4 108.6
by more than 20 dB for downwind conditions as com-
Experiment pared to upwind conditions.
dB — — — 113.7
dBA — — — 109.2
shift to higher frequencies. If the diameter of these

Glossary
ports were small enough to substantially shift the fre- Absorption: Conversion of sound energy into another
quency of the gas jet noise, then the combustion noise form of energy, usually heat, when passing
would dominate on the dBA scale. through an acoustical medium.
Absorption coefficient: Ratio of sound absorbing effec-
11.5.3 Atmospheric Attenuation Example tiveness, at a specific frequency, of a unit area of
acoustical absorbent to a unit area of perfectly
Figure 11.45 shows an example of the SPL received from a absorptive material.
flare calculated at various distances and for different atmo- Acoustics: Science of the production, control, transmis-
spheric attenuation. Notice at a distance of 1500 ft (460 m) sion, reception, and effects of sound and of the
from the flare, the noise level peaks at about 80 dBA, while phenomenon of hearing.
at 3000 ft (910 m) the peak reduces to about 74 dBA if atmo- Ambient noise: All-pervasive noise associated with a
spheric attenuation is not taken into account. When the given environment.
atmospheric attenuation is taken into account, depend- Anechoic room: Room whose boundaries effectively
ing on the ambient temperature and humidity level con- absorb all incident sound over the frequency
sidered (see Section 11.1.3.2), there is further reduction in range of interest, thereby creating essentially
noise levels. It is important to note that the contribution in free field conditions.
each case is significant. Given the particular atmospheric Audibility threshold: Sound pressure level, for a speci-
conditions in this example, the attenuation has created a fied frequency, at which humans with normal
significant difference. The 10 dB attenuation (from 74 to hearing begin to respond.
90 Overall dB level
80 86 dB
80 dBA
70
SPL (dB)
80 dB
60
74 dBA
50
79 dB
40 64 dBA
30
31.5
125
500
2000
8000
Frequency (Hz)
Noise spectrum at a distance of 1500 ft without atmospheric attenuation

Noise spectrum at a distance of 3000 ft without atmospheric attenuation
Noise spectrum at a distance of 3000 ft with atmospheric attenuation
(60°F, 90% rel. humidity)
FIGURE 11.45
Effect of distance on flare noise.
Background noise: Ambient noise level above which Pink noise: Broadband noise whose energy content is
signals must be presented or noise sources inversely proportional to frequency (−3 dB per
measured. octave or −10 dB per decade).
Decibel scale: Linear numbering scale used to define a Power spectrum level: Level of the power in a band
logarithmic amplitude scale, thereby compress- 1 Hz wide referred to a given reference power.
ing a wide range of amplitude values to a small Reverberation: Persistence of sound in an enclo-
set of numbers. sure after a sound source has been stopped.
Diffraction: Scattering of radiation at an object smaller Reverberation time is the time (in seconds)
than one wavelength and the subsequent inter- required for sound pressure at a specific fre-
ference of the scattered wavefronts. quency to decay 60 dB after a sound source is
Diffuse field: Sound field in which the SPL is the same stopped.
everywhere, and the flow of energy is equally Root-mean-square (RMS): The square root of the arith-
probable in all directions. metic average of a set of squared instantaneous
Diffuse sound: Sound that is completely random in phase; values.
sound that appears to have no single source. Sabine: Measure of sound absorption of a surface. One
Directivity factor: Ratio of the mean-square pressure metric sabine is equivalent to 1 m2 of perfectly
(or intensity) on the axis of a transducer at a cer- absorptive surface.
tain distance to the mean-square pressure (or Sound: Energy transmitted by pressure waves in air
intensity) that a spherical source radiating the or other materials which is the objective cause
same power would produce at that point. of the sensation of hearing. Commonly called
Far field: Distribution of acoustic energy at a much greater noise if it is unwanted.
distance from a source than the linear dimen- Sound intensity: Rate of sound energy transmission
sions of the source itself. See also diffraction. per unit area in a specified direction.
Free field: An environment in which there are no reflective Sound level: Level of sound measured with a sound
surfaces within the frequency region of interest. level meter and one of its weighting networks.
Hearing loss: An increase in the threshold of audibil- When A-weighting is used, the sound level is
ity due to disease, injury, age, or exposure to given in dBA.
intense noise. Sound level meter: An electronic instrument for
Hertz (Hz): Unit of frequency measurement, represent- measuring the RMS of sound in accordance
ing cycles per second. with an accepted national or international
Infrasound: Sound at frequencies below the audible standard.
range, that is, below about 16 Hz. Sound power: Total sound energy radiated by a source
Isolation: Resistance to the transmission of sound by per unit time.
materials and structures. Sound power level: Fundamental measure of sound
Loudness: Subjective impression of the intensity of a power, defined as
sound.
Masking: Process by which the threshold of audibil- P
Lw = 10 log dB
ity of one sound is raised by the presence of P0

another (masking) sound.
Near field: That part of a sound field, usually within where
about two wavelengths of a noise source, where P is the RMS value of sound power (Watts)
there is no simple relationship between sound P0 is 1 × 10−12 (Watts)
level and distance. Sound pressure: Dynamic variation in atmospheric
Noise emission level: dBA level measured at a speci- pressure. The pressure at a point in space minus
fied distance and direction from a noise source, the static pressure at that point.
in an open environment, above a specified Sound pressure level: Fundamental measure of sound
type of surface; generally follows the recom- pressure defined as
mendation of a national or industry standard.
Noise reduction coefficient (NRC): Arithmetic average P
Lp = 20 log dB
of the sound absorption coefficients of a mate- P0

rial at 250, 500, 1000, and 2000 Hz.
Phon: Loudness level of a sound, numerically equal to where
the SPL of a 1 kHz free progressive wave, which P is the RMS value (unless otherwise stated) of
is judged by reliable listeners to be as loud as sound pressure (Pascal)
the unknown sound. P0 is 2 × 10−5 (Pascal)
Noise 299
Sound transmission loss: Ratio of the sound energy 9. W. Bussman and J. White, Steam-assisted flare testing,
emitted by an acoustical material or structure John Zink Co. Internal Report, September 1996.
to the energy incident upon the opposite side. 10. A.A. Putnam, Combustion Noise in the Handheld Industry,
Standing wave: A periodic wave having a fixed distri- Battelle Columbus Laboratories, Columbus, OH.
11. L.L. Beranek and I.L. Ve’r, Noise and Vibration Control
bution in space that is the result of interference
Engineering, John Wiley & Sons, Inc., New York, 1992.
of progressive waves of the same frequency and
12. H. Shen and C.K.W. Tam, Numerical simulation of the
kind. Characterized by the existence of maxi- generation of axisymmetric mode jet screech tones,
mum and minimum amplitudes that are fixed AIAA J., 36(10), 1801, October 1998.
in space. 13. L.L. Beranek, Noise and Vibration Control, McGraw Hill
Thermoacoustic efficiency: A value used to character- Book Co., New York, 1971.
ize the amount of combustion noise emitted 14. Allied Witan Co., Noise Facts and Control, Company bro-
from a flame. Defined as the ratio of the acousti- chure, North Royalton, OH, 1976.
cal power emitted from the flame to the total 15. W.R. Bussman and D. Knott, Unique concept for noise
heat release rate of the flame. and radiation reduction in high-pressure flaring, OTC
Ultrasound: Sound at frequencies above the audible Conference, Houston, TX, 2000.
16. L.D. Frank and D.R., Dembicki, Lower plant noise with
range, that is, above about 20 kHz.
lagging, Hydrocarb. Process., 71(8), 83–85, 1992.
Wavelength: Distance measured perpendicular to
the wavefront in the direction of propagation
between two successive points in the wave,
which are separated by one period. Equal to the
ratio of the speed of sound in the medium to the
fundamental frequency.
Weighting network: An electronic filter in a sound level Bibliography
meter that approximates, under defined condi-
tions, the frequency response of the human ear. Brief, R.S. and R.G. Confer, Interpreting noise dosimeter results
The A-weighting network is most commonly based on different noise standards, Am. Ind. Hyg J., 36(9),
677–682, 1975.
used.
Crow, S.C. and F.H. Champagne, Orderly structure in jet tur-
White noise: Broadband noise having constant energy
bulence, J. Fluid Mech., 48(3), 547–591, 1971.
per unit of frequency. Diserens, A.H. Personal noise dosimetry in refinery and chem-
ical plants, J. Occup. Med., 16(4), 255–257, 1974.
Energy Conservation Resources Board, Noise Control,
Directive 038, dated February 16, 2007, http://www.
ercb.ca/docs/documents/directives/Directive038.pdf,
accessed 1 August 2012.
Gharabegian, A. and J.E. Peat, Saudi petrochemical plant noise
References
control, J. Environ. Eng., 112(6), 1026–1040, 1986.
1. A.P.G. Peterson, Handbook of Noise Measurement, 9th edn., HFP Acoustical Consultants, Effect of flow parameters
GenRad, Westford, MA, 1980. on flare stack generator noise, Proceedings of Spring
2. ISO 3744, Acoustics—Determination of sound power Environmental Noise Conference: Innovations in Noise
levels and sound energy levels of noise sources using Control for the Energy Industry, Banff, AB, Canada, April
sound pressure—Engineering methods for an essentially 19–22, 1998.
free field over a reflecting plane, 2010. IEC 61672-1, Electroacoustics—Sound level meters—Part 1:
3. W. Daiminger, K.R. Fritz, E. Schorer, and B. Stüber, Specifications, 2002.
Ullmann’s Encyclopedia of Industrial Chemistry, Vol. B7, ISO 1683, Acoustics—Preferred reference values for acoustical
VCH, Weinheim, Germany, 1995, pp. 384–401. and vibratory levels, 2008.
4. API, Recommended Practice 531M: Measurement of Noise ISO 532, Acoustics—Method for calculating loudness level, 1975.
from Fired Process Heaters, Reaffirmed Edition, American ISO 1996-1, Acoustics—Description, measurement and assess-
Petroleum Institute, Washington, DC, August 1995. ment of environmental noise - Part 1: Basic quantities
5. A. Thumann, and R.K. Miller, Secrets of Noise Control, and assessment procedures, 2003.
Fairmont Press, Atlanta, GA, 1974. ISO 3744, Acoustics—Determination of sound power levels
6. O.C. Leite, Predict flare noise and spectrum, Hydrocarb. and sound energy levels of noise sources using sound
Process., 68, 55, 1988. pressure—Engineering methods for an essentially free
7. Müller-BBM, GmbH: Noise emissions of different flare field over a reflecting plane, 2010.
systems—field measurements taken in various refineries ISO 8297, Acoustics—Determination of sound power lev-
and petrochemical plants. Proprietary data, 1990–2004. els of multisource industrial plants for evaluation of
8. Verein Deutscher Ingenieure, VDI 3732, Standard noise lev- sound pressure levels in the environment—Engineering
els of technical sound sources—Flares, VDI guideline, 1999. method, 1994.
ISO 9614-1, Acoustics—Determination of sound power lev- Ribner, H.S. Perspectives on jet noise, AIAA J., 19(12), 1513, 1981.
els of noise sources using sound intensity—Part 1: Roberts, J.P., Gas Combustion Noise, PhD thesis, London
Measurement at discrete points, 1993. University, London, U.K., 1971.
ISO 9613-1, Acoustics—Attenuation of sound during propaga- Seebold, G. and A.S. Hersh, Control flare steam noise,
tion outdoors—Part 1: Calculation of the absorption of Hydrocarb. Process., 51, 140, 1971.
sound by the atmosphere, 1993. Selle, G.K. Steam-assisted flare eliminates environmental con-
ISO 9613-2, Acoustics—Attenuation of sound during propa- cerns of smoke and noise, Hydrocarb. Process., 73(12),
gation outdoors—Part 2: General method of calculation, 77–78, 1994.
1996. Shivashankara, B.N., W.C. Strahle, and J.C. Henkley,
Lang, W.W. Ed., A commentary on noise dosimetry and stan- Combustion noise radiation by open turbulent flames,
dards, Proceedings of the Noise Congress-75, Gaithersburg, Paper 73-1025, AIAA Aero-Acoustics Conference, Seattle,
MD, September 15–17, 1975. WA, 1973.
Maling, Jr., J.C. Ed., Start-up silencers for a petrochemical com- Straitz, J.F. Improved flare design, Hydrocarb. Process., 73(10),
plex, Proceedings of the International Conference on Noise 61–66, 1994.
Control Engineering, Honolulu, HI, December 3–5, 1984. Thomas, A. and G.T. Williams, Flame Noise: sound emission
Powell, A. On the noise emanating from a two dimensional jet from spark-ignited bubbles of combustion gas, Proc. Roy.
above the critical pressure, The Aeronautical Quarterly, 4, Soc., A, 294, 449, 1966.
103, 1953. Zwicker, E. and H. Fastl, Psychoacoustics—Facts and Models,
Reed, R. Furnace Operations, Gulf Publishing, Houston, TX, 1981. Springer Verlag, Berlin, Germany, 1990.
12
Combustion Controls, Burner Management,
and Safety Systems
Vladimir Lifshits
CONTENTS
12.1 Introduction.................................................................................................................................................................. 302
12.2 Functions of Combustion Controls and Burner Management Systems............................................................... 302
12.3 Types of Combustion Controls: Controls Schemes................................................................................................. 305
12.3.1 Single Point Positioning Controls.................................................................................................................. 306
12.3.2 Parallel Positioning Controls with Multiple Actuators.............................................................................. 308
12.3.3 Metering Controls............................................................................................................................................ 309
12.3.4 Parallel Positioning Combustion Controls with Trims................................................................................311
12.4 Combustion Controls for Systems Requiring Minimum NOx Emissions: Ultralow NOx Controls................. 312
12.5 Special Cases..................................................................................................................................................................314
12.5.1 Controls for Applications with Fuels of Variable Composition.................................................................314
12.5.2 Systems with Flue Gas Recirculation............................................................................................................ 315
12.5.3 Boilers with Multiple Burners........................................................................................................................ 315
12.6 Modulating Combustion Air Flows...........................................................................................................................316
12.6.1 Fans with Variable Frequency Drives............................................................................................................316
12.6.2 Modulation of Flows with Inlet Vane Dampers...........................................................................................316
12.7 Sizing Controls Components.......................................................................................................................................316
12.8 Burner Management Systems......................................................................................................................................318
12.9 Layers of Safety............................................................................................................................................................ 319
12.10 Setting Combustion Controls: Commissioning of Combustion Systems............................................................ 322
12.11 Feedwater Controls...................................................................................................................................................... 323
12.12 Sensors Often Used in Modern Combustion Controls........................................................................................... 324
12.12.1 Pressure Transmitters.................................................................................................................................... 324
12.12.2 Temperature Measurements......................................................................................................................... 324
12.12.2.1 Thermocouples................................................................................................................................ 324
12.12.2.2 Resistance Temperature Detectors............................................................................................... 327
12.12.2.3 High Velocity Thermocouples...................................................................................................... 327
12.12.3 Flowmeters...................................................................................................................................................... 328
12.12.3.1 Orifice- and Nozzle-Based Flowmeters....................................................................................... 328
12.12.3.2 Vortex Shedding Flowmeter......................................................................................................... 330
12.12.3.3 Coriolis Flowmeter......................................................................................................................... 330
12.12.3.4 Turbine Flowmeters....................................................................................................................... 330
12.12.3.5 Ultrasonic Flowmeters................................................................................................................... 330
12.12.4 Emissions Analyzers..................................................................................................................................... 330
12.12.4.1 In Situ Oxygen Analyzers............................................................................................................. 330
12.12.4.2 Portable Emission Analyzers........................................................................................................ 331
12.12.4.3 Emissions Analyzers, CEMS......................................................................................................... 331
12.12.5 Wobbe Index and Combustion Air Requirement Index Analyzers....................................................... 332
Nomenclature.......................................................................................................................................................................... 333
References................................................................................................................................................................................. 333
301
The main components of a typical BMS are the con-

troller performing all the logical steps, different pres-
12.1 Introduction
sure switches, different on/off sensors (position end
This chapter gives a relatively short overview of com- switches), and flame detection devices. The safety shut-
bustion controls without going in depth on many of the off valves (SSVs) and vent valves may also be considered
design details extensively discussed in numerous spe- as part of a BMS.
cialized books1–4 with1 being a very good example on Simply stated, the function of combustion controls
the subject. This chapter is primarily written to describe is to modulate the burner load while maintaining the
many practical issues related to the accuracy, reliabil- required fuel-to-air equivalence ratios and, in many
ity, and safety of the entire boiler control system that cases, the amount of flue gas recirculated from the
design engineers and users of the equipment need to be exhaust back to the burner (flue gas recirculation [FGR]).
aware of when defining system requirements for spe- Ideally, the controls would also steer the process to the
cific conditions, comparing alternative solutions, and maximum robustness with respect to allowing for suffi-
operating the systems. The importance of these issues ciently wide variations of different uncontrolled param-
is continuously increasing with more demand to oper- eters and various deviations in the components and
ate combustion equipment with minimum possible sensors performance.
NOx emissions—a goal that may be in potential conflict The boiler combustion controls are integrated with
with reliability and safety unless adequately addressed the unique conditions of the process and configured
both in the system design and maintenance. The low- to operate in two modes. One is automatic, where the
est NOx can be achieved when the flame is diluted load is modulated based on the process requirements.
with inert gases and/or heavily staged to the point that The second is manual when the load is set manually.
many deviations of the regime may result in combus- In either mode of operation, the boiler is continuously
tion instabilities or a loss of flame. Minimizing these supervised by the BMS. The main function of the con-
deviations in real systems whose operation is affected trols in boilers is maintaining required steam pressure
by many primary and secondary process variables in the system when the consumption of steam varies.
(PVs) is the task of combustion controls. Some environ- Other process control functions may be in maintaining
mental regulations, like BACT (best available control certain temperatures or the load, steam quality or the
technology) and MACT (maximum achievable control percentage of steam in the two-phase flow of steam and
technology), by definition have become continuously water, etc.
moving targets for the combustion systems. These mov- The combustion controls may also perform self-
ing targets should be considered with the associated checking functions of the quality of its performance
analysis of the costs to the industry and society that and, in some cases, the “health” of the burner and diag-
often is not properly completed. As a result, these mov- nostics. Depending on the severity of abnormal con-
ing targets have caused rising complexity of the com- ditions, if detected, the combustion controls can send
bustion controls described here. signals to the BMS to generate alarms or emergency
shutdowns. The combustion controls may be designed
with the ability to impose different limitations to the
operation, like setting temporary limits for the load
modulation, for example.
The complexity of the combustion controls depends
12.2 Functions of Combustion Controls
on the requirements for the system and can easily vary
and Burner Management Systems by an order of magnitude and more.
The tasks of the burner management system (BMS) are The major components of the combustion controls can
to safely start and stop the burners and monitor the be divided in six groups:
supply of utilities and the presence of the flame dur-
ing operation, whereas the combustion controls sys- • Flow and hydraulic resistance controlling devices or
tem is responsible for load modulation according to elements—Different valves and dampers with
the demand while meeting the process performance driving actuators and, in some cases, variable
requirements from the standpoint of reliability, overall frequency drives (VFDs) of the fans. These
efficiency, and environmental regulations. NFPA 85— components in many cases are equipped with
Boiler and Combustion Systems Hazard Code3 that is feedback modules. In some cases, the controls
followed by many industries—specifies in detail the may contain manually adjustable components,
many steps that the BMS shall perform and also states dampers, or valves that will remain in fixed
many basic requirements to the combustion controls. positions after the burner tuning is complete.
Combustion Controls, Burner Management, and Safety Systems 303
• Measurement devices (sensors of PVs)—Pressure, to modulation” switch is set “on” to allow for fully auto-
temperature, species concentration transmit- matic or manual modulation of load.
ters flowmeters, etc. These devices provide pri- There may be several levels of manual operations. For
mary inputs to the combustion control system. example, the load can be set manually with the rest of the
Examples and description of many primary controls functioning in automatic and making adjust-
input devices are given in Section 12.12. ments for variables that still may fluctuate over time. In
• Controller—A device that processes all the this case, the boiler output cannot be set higher than the
information from the sensors, process and process demand unless the output bypasses the process.
manual inputs, and inputs from the feedback With more manual intervention, any or all outputs to
modules; generates outputs to the actuators; the flow’s controlling devices may be in manual. This
performs some accuracy monitoring func- level of manual operation is usually reserved for setting
tions; communicates with the BMS; and may and tuning the combustion system and not available to
have capability of recording the history of the the boiler operator.
operation. The device is usually a loop control- As requirements to maximize system efficiency and
ler, a dedicated programmable logic controller minimize pollution become even more stringent, the
(PLC), or a part of the distributed control sys- demand for the accuracy of the combustion controls
tem (DCS) at the site. increases requiring more and more complicated com-
bustion controls. The drive to reduce the NOx emissions
• Input-output (I/O) modules—Devices convert sig-
to lowest achievable levels results in special burners
nals from the sensors to a format that can be
designed for using substantial amounts of FGR often
processed by the PLC.
in the amounts approaching combustion stability lim-
• Human-machine interface (HMI) device—Allows its. Systems with such burners require control of the
the user to conveniently set the controls and process with minimum deviations from the optimum
observe the operation. The most common type of regimes that may be defined not only with respect to
HMI for PLCs is a touch screen display that must the load but also with respect to many other usually sec-
be programmed for the specific process. The ondary variables. Examples of these variables are atmo-
HMI can be programmed with multiple screens. spheric conditions, properties of fuel (in some cases),
Some of the HMI combustion controls screens effects of the system thermal inertia, hysteresis effects
will be open to the process operators. Others will in the controls elements, gradual changes in the system
be normally locked and reserved for use by per- characteristics due to physical wear of the components,
sonnel responsible for the combustion controls contamination with some deposits, etc.
set up and tuning. Some physical HMIs can be The complexity of the combustion controls increases
programmed for interfacing with the BMS and where the load needs to modulate quickly. When the
the combustion controls. In DCS systems, HMIs changes are slow, the effect of many variables can be
for the overall process controls also are used for compensated without loss of accuracy in dynamic situ-
the combustion controls. ations. An example is a boiler on standby serving as a
• Some combustion controls systems also include backup to another power generation device that has a
multiple devices, primarily pressure gauges likelihood of a sudden trip and stoppage, contributing
and temperature indicators, not necessarily output to a power grid. In this case, a typical require-
wired into the combustion controls, but serv- ment to ramp up load rate from minimum to high fire
ing as convenient visual indicators of the pro- would be just a few minutes.
cess. Pressure gauges usually cost an order of A substantial increase in complexity of the controls
magnitude less than pressure transmitters. The may also be caused by variations in fuel composition
readings from these gauges may be very useful coupled with low emissions requirements. Additional
during the controls tuning and troubleshooting issues may need to be adequately resolved by the con-
of possible problems. trol of multiple burner boilers when individual burn-
ers need to be added or removed from service with the
Operation of the controls starts shortly after the burner boiler online when the burners—typically low NOx
light-off and ends with the burner shutdown triggered burners—cannot simply ride out sudden substantial
by the BMS. However, to some degree the controls are changes in the fuel-to-air ratios.
also engaged in the sequence of steps governed by the With the expanding complexity of the applications,
BMS when preparing the system for burner light-off combustion controls may easily become the most com-
and after its shutdown. Right after the burner light- plicated part of the overall system as well as the most
off, the operation is in manual mode. Then a “release expensive after the boiler itself.
In addition to combustion controls, boilers require In industrial and utility applications, a common way of
well-designed feedwater controls. The feedwater con- describing and documenting combustion controls func-
trols are tasked to minimize fluctuations in the steam tions and algorithms is with the use of logic diagrams.
drum water level that inevitably happen when the The common format used for these diagrams was devel-
boiler load changes. An excessive rise in the water oped by the Scientific Apparatus Makers Association.
level in the drum may cause significant carryover of They are called SAMA diagrams. SAMA diagrams show
water into the steam line that may be very undesirable the flow of information within the controls system from
for the steam user, while the drying of the drum may the sensing element, like a pressure transmitter, usu-
be damaging to the boiler. The feedwater controls usually shown on the top, to the final control element, like
ally work in parallel with combustion controls with a valve or damper, shown on the bottom. The language
minimum direct interaction except that they usu- of these diagrams is shown in Figure 12.1 along with the
ally use the same controller. The feedwater controls accepted rules that will be followed in different exam-
communicate with the BMS providing alarm and trip ples in this chapter. The language contains elements of
signals when the water level deviates outside of the Boolean algebra convenient for describing discrete logic
assigned limits. steps. Complete SAMA diagrams show processing of
Out = analog
Signal Generates constant Or A
Analog PT
input signal from
A analog signal
OR Out Out = A|B
external source generator (adjustable by user)
gate B
input
A A = signal input
Out failure A A
And gate AND Out Out = A and B
A B B
Out = A except When B = 0, Out = A
Transfer T C Not
Low/high that out will be When B = 1, Out = C
A Out Out = inverse of A
B C bound within low (invert)
limit limit B and high
limit C Out
Out
A Signal connection examples
A Out = lower of
the two signals,
A or B Add Σ B Out = A + B
Low select B <
LO
C = true if A is
C lower than B
Out Each represents one signal with Each represents two
Out multiple connection points. independent signals that
A Out = higher of A do not connect.
the two signals,
High select A or B
B > Subtract Δ B Out = A – B Analog
HI C = true if A is
C higher than B signal
Out Out Discrete
Σ A signal
B n A Out = average of
Average A and B Discrete
D
Out Multiply × B Out = A × B input
A PI controller
Discrete
D
A = process value Out output
B Δ B = setpoint
C TC TR D When C = 1, out A Input
PI E MAN tracks D
HMI A
When E = 1, PI
from HMI
K ∫
action stopped Divide ÷ B Out = A / B Output
and out holds A HMI
Out last value to HMI
Out Signal
A Master
Out = rate-of- name demand
change limited A. A
Rate limiter B TR When B = 1, out
Signal A drives
track A without Analog device external to
any limiting output FCV control system
Out
A
Out = preset
function of A,
Function usually by a
f (x)
generator multisegment
straight line
characterizer
Out
FIGURE 12.1
Symbols of SAMA logic diagrams.
all the necessary Boolean inputs (0 and 1) from the BMS (steam pressure in this case) and its usually manual
and manual Boolean and analog inputs to the controls target value— setpoint (SP)—by gradually changing
from the operator. The diagrams, however, leave out an the burner firing rate. The perceived complexity of
important part of the controls—design of HMI screens. the logic here is due to showing interactions with the
BMS during the purge and light-off cycles and check-
ing different limits set for process and for the burner
operation that are parts of different layers of safety dis-
cussed later in this chapter.
A basic loop controller combines three control princi-
12.3 Types of Combustion Controls:
ples in one unit. These include a proportional controller,
Controls Schemes integral controller, and a controller that includes a dif-
In a combustion control system, a main variable— ferential function together called a PID controller. The
master demand defining the burner firing rate—comes function of a PID controller can be explained with the
from a manual setting or is determined automatically following generic equation between the controller output
based on the demand. An element of a SAMA logic dia- (CV), PV, and the target value of the variable also called
gram showing generation of the demand signal is illus- the SP. A PID controller defines CV as a sum of some
trated in Figure 12.2. constant (K0), a value proportional to the instantaneous
In the heart of the logic is a basic feedback control error E = (PV − SP), a value proportional to the integral
loop, also called a controller, working toward mini- of E over time, and a value proportional to the speed of
mizing the discrepancy between the measured PV change in E at the moment:
t
Auto
 dE 
control
active
Steam pressure
set point
Steam drum
pressure

∫
CV = K 0 + K1E + K 2 E dt + K 3  
t0
 dt 

(12.1)
HMI HMI PT
with the units for CV, PV, and E as percentages in the

Service access A HMI range from 0% to 100%. Equation 12.1 can be written
Service access with a slightly different notation:
A
A
 t 
 1 dE 
Low limit High limit


 R
t0
∫
CV = BIAS + GAIN E +   E dt + K 3
dt 

Δ
HMI
BMS override TC TR
Man Master
PID manual demand manual The same equation after differentiation is as follows:
K ∫
HMI
dCV  dE   GAIN 
= GAIN ×   ×  ×E
T A dt  dt   R 
Service
access Service d 2E
access + (GAIN × K 3 ) × (12.2)
HMI HMI dt 2
Min demand Max demand

All the components in the aforementioned equations
have physical meanings. The first component in (12.2)
HMI
represents some initial, often average, value of the
BMS override TR
output. The second component represents a portion
of output proportional to the error. This component
by itself is not sufficient to accurately control even a
HMI
steady-state operation as E will ultimately be settled
Master at the value
demand
FIGURE 12.2 E=
(CV − BIAS) (12.3)
Steam pressure control logic. GAIN
While it may be tempting to use high GAIN for more positioning, the output to the first control element—
accuracy, it must be limited due to process time lag that a fuel valve, for example—is defined by the process
can cause the system to continuously hunt or oscillate. and usually comes from some PID loop set to main-
The third component in (12.1)—the integral part—as tain the desired steam pressure or another PV. The
better seen from (12.2), defines the speed of change of the output of the first PID loop itself may be fluctuating or
output (CV) to be proportional to the error. The output set manually depending whether the main process is
changes faster when the error is large and more slowly automatically or manually controlled. The target posi-
when the error is small. The R/GAIN factor is called reset tions/outputs for all other control elements like other
time and has the dimension of time. The reset time is modulating valves and dampers are defined instan-
proportional to the time it would take to reduce the error taneously based solely on the position of the first ele-
by ~2.7 times (the number e which is the base of natural ment. With fully metering controls, all the flows are
logarithms) if there are no new disturbances to the pro- first defined based on the process demand usually via
cess with the proportionality factor being an inverse of a PID loop. Then the fuel valve and air damper posi-
the effect of 100% change in CV on E. This is by far the tions are defined by other PID loops based on com-
most important component in PID loops controlling flow parison of the metered and target values for the flows
rates. Setting a shorter reset time accelerates the controls creating cascading PID loop controls configuration. In
response, but for the robustness of the controls algorithm hybrid systems with partial metering, for example, not
the R/GAIN must be longer than inertia (or time lag) in all the flows are metered. The outputs to the elements
the system response to changes made by the controls. The defining flows that are not metered are defined by one
integral part by itself may be insufficient in dynamic situa- of the metered flows. Other hybrid systems use paral-
tions. For example, when the demand increases at a steady lel positioning with adjustments—trims based on the
rate, E may stay constant with CV continuously lagging inputs from flow sensors (or inputs indicative of flows)
the demand. However, with the help of the proportional and sensors of other PVs.
part of the output, CV will be able to change faster.
The last component makes adjustments to the output
12.3.1 Single Point Positioning Controls
based on the acceleration of the change in E. It may help to
speed up the reaction of the controls in some cases but is The simplest type of parallel positioning system is a
seldom used as it easily makes controls unstable and very system where dampers and fuel valve(s) are mechani-
susceptible to any system noise. So, a common type of cally linked and connected to a single actuator. Such a
controller in combustion applications is actually a PI type. system is also called single point positioning (SPP) and
When directly controlling the flows with the CV pro- is shown schematically in Figure 12.3.
portional to the actual flow, the GAIN must always be About 50 years ago this was the most common type of
substantially less than 100%. Reference [4] provides an combustion controls in single burner industrial boilers.
overview and details on setting optimum parameters of Mechanically, the connections were usually such that a
PID loops for different processes including combustion single actuator was driving a rotating shaft supported
controls. by sliding bearings called a jack shaft, with the rotation
As the parameters of the PID loops may need to be span of ~90°–120°. The jack shaft had a few clamped
changed during the system set up and tuning, in many arms, each connected to a fuel valve or a damper via
cases, it may be important to program the controls in a ball joint and adjustable in length with mechanical link-
way to achieve bumpless transfer during these changes. ages similar to those used in aircrafts. Such systems
Bumpless transfer eliminates disturbance to the process provided wide latitude for kinematic adjustments to the
at the moment system parameters of the PID loop get initial and final points of valves and damper movements
changed. In order to achieve bumpless transfer, the out- by varying the rotation of the arms on the jack shaft, dis-
put CV of the controller should be maintained by chang- tances on the arms where ball joints were mounted and
ing BIAS every time parameters of the PID loop change. finally the length of linkages. In the simplest systems,
This procedure is also done when switching from man- using a modified “trial and error” procedure of setting
ual to auto or back by setting the bias to insure the CV is linkages can produce a rough air/fuel balance.
set to the current setting. In a few cases, however, main- The only actuator in the system may be driven based
taining the bias is more important than elimination of on the demand generated by a controller with a PID con-
the transitional disturbance. trol loop operating based on the input from the process,
Generally, combustion control schemes that convert or set manually by the operator. Cam plates, or their mul-
the firing demand to the actual required positions of tipoint adjustable equivalents, were soon used to achieve
fuel control valves (FCVs), dampers, etc. are classified more accurately required openings of FCVs for load lev-
as parallel positioning or metering types with many els between the maximum and minimum. Figure 12.4
hybrid systems using features of both. With parallel shows an example of a plug type fuel valve integrated
Master
demand
f (x)
Power unit
Jackshaft
Mechanical
linkage Adjustable
cam
Air Fuel Fuel

control damper control valve #1 control valve #2
FIGURE 12.3
Schematic of SPP.
with a cam made as a series of adjustable and lockable

screws installed through a preinclined plate. The slope
of the plate usually matches 20%–80% of the plug travel
from fully closed to fully open position to minimize the
span of adjustments with the screws.
SPP control systems were also expanded to add a
rather crude control of the FGR by connecting the same
actuator via additional linkages to the FGR controlling
damper (a damper in the FGR line or a damper creating
some negative pressure at the fan inlet). This worked
satisfactorily in early systems using small amounts of
induced FGR, typically less than 10%, that was mainly
needed for operation above 40%–50% load to reduce
peak NOx emissions.
The advantages of SPP controls are its simplicity, low
cost and, in most cases, the ability for fast load modula-
tion without lead/lag effects and loss of accuracy. SPP
controls are also reliable and safe, which is reflected in
the simplicity of the BMS.
The disadvantages of SPP are many. The most impor-
tant is low accuracy, as changes in ambient conditions
(mostly combustion air and fuel temperatures) cause
deviations to the originally set air/fuel ratios and mak-
ing it necessary to set the burner operation upfront with
FIGURE 12.4
somewhat higher excess air (stack O2), etc. The second
COEN AC (adjustable characteristic) fuel valve designed for SPP disadvantage is difficulty shaping the cam or mechani-
controls. cally adjusting multiple points, especially if the FCVs
are oversized or when extended range of operation, device toward the requested position. The time constant
typically more than 6–1, is needed. With the SPP there of this loop (reset time R, for example) is usually much
is also no easy ability to document exact setting of the shorter than the characteristic time of the process PID
points, creating the necessity of resetting the burner if loop defining request for the load.
the components are worn off, replaced, or serviced after The process of setting all the functions takes place
the initial set. during the initial commissioning of the combustion
system and involves manually setting the boiler over
the load range by moving it from point to point. The
12.3.2 Parallel Positioning Controls
functions are usually a combination of stitched together
with Multiple Actuators
linear segments defined between identical points in the
Having separate actuators for the valves and dampers load. The number of segments in the function may vary.
eliminates the need for mechanically linking the modu- Typically 12–16 points are sufficient to accurately control
lating elements of the system. The concept of parallel the boiler. It is a good practice for the setting accuracy to
positioning in this case is in creating a set of functions keep identical load segments and number of points for
defining the position of all the fuel air and FGR control- all functions. Increasing the number of points usually
ling elements as functions of load (or fuel flow) as illus- does not add appreciably to the accuracy of the set, but
trated in Figure 12.5. may add some convenience to the setting process at the
Such a concept allows a much higher degree of flexibil- expense of increased time needed to document the set.
ity than SPP. In most cases, these actuators are equipped Another advantage of this type of controls is its speed.
with positioners or function generators that define the With the speed of modern processors, the outputs to all
actuator position. The actuator uses this positioner or actuators are computed and sent at the same moment,
function generator with its internal control PID loop or minimizing any delays in moving actuators to the
its electromechanical equivalent to actually drive the desired positions. Synchronous sending of the outputs to
Master
demand
f (x) f (x)
Purge Purge
signal signal
from BMS from BMS
T 100% T 0%
Air manual Fuel manual

control active Manual air control active Manual fuel
(service access) control station (service access) control station
(service access) (service access)
HMI HMI
HMI HMI
T A T A
HMI HMI
Air control Fuel control

device device
FIGURE 12.5
Parallel positioning of fuel and air.
the positioners does not provide instant response of the Fuel Air
actuators due to the limited speed of response that may flow flow
be substantial for large geared drive dampers. The inter- Master
FT demand FT
nal electronics of the device may also delay its reaction,
similar to the action of a PID loop, so the speed of load
modulation will still be limited, but typically much faster
than any industrial combustion system requirements.
Setting some number of these multiple functions for
the actuators is possible using loop controllers, however,
it is very inconvenient and time consuming. Much bet- f–12(x) f –11(x)
ter suited systems are those that can set actuator func-
tions via convenient HMI devices. It can be a touch >HI >LO
screen display, a personal computer (PC), or controls
integrated into a DCS.
In their basic form, as described before, the parallel f1(x) f2(x)
positioning controls with multiple actuators rely exclu-
sively on the repeatability of the valves’ and dampers’ Δ Δ
characteristics. Malfunction of any of the actuators (or TC TR TC TR
switching an actuator off) will result in deviations in the MAN MAN
fuel, air or FGR flows that may result in a loss of flame. K ∫ K ∫
To minimize the chance of this happening, all actuator
positioners should be equipped with position feedback
modules with set limits for deviation alarms and trips Air control Fuel control
substantially adding complexity to the system. Practical device valve
experience shows that even with the best brands of
products keeping all the deviations between the out- FIGURE 12.6
puts and position feedbacks to below 2% of the span is Fully metered controls with cross limiting between fuel and air.
not reliable. At the same time, even 2% deviation at low
firing rates can cause substantial error in the flow ratios considerations) or in a step by step process during the
often exceeding 10% and even 15%. For this reason, burners’ commissioning.
parallel positioning systems with multiple actuators in In some cases, a third tier PID loop is used to make
their basic form are seldom used unless the system is corrections to the SPs of the second tier. Alternatively,
enhanced with additional sensors allowing trimming the controls may be configured with another modulat-
of the actuator positions to achieve better accuracy and ing device like a VFD of the fan, a pressure regulator
reduced reliance (or no reliance) on the position feed- in the fuel supply line upstream of the FCV or a second
back information. damper acting to make corrections based on the action
of another parallel controller.
For slow-changing processes, the concept allows con-
12.3.3 Metering Controls
trolling the combustion process almost as accurately as
Using metering controls, the flows are measured as the the flowmetering elements themselves. If the process
name implies. One flow is set to the process demand changes quickly (fast change in load, for example), cas-
via a PID controller. This measured flow defines SPs for cading PID loops in the controls may cause substantial
other flows by applying to F1(x) and F2(x) characteriz- dynamic errors. Optimizing parameters of the loops
ing functions to the demand signal, prompting relevant may reduce the errors, but it can still be a limiting fac-
actuators to move according to the outputs of second tor on how fast the load may modulate. Addition of
tier (cascading) PID loop controllers as illustrated in the rate of change limiting blocks to the controls algo-
Figure 12.6. rithm allows automatically limiting magnitudes of the
The output of the PID loop demand controller dynamic deviations to the fuel-to-air and FGR ratios
(DC) (Figure 12.2) affects the SPs of the next tier flow whether during manual or automatic load modulation.
controller (FC) also operating with a PID loop. The However, slowing down the controls below the actual
argument x in the F(x) functions is the demand usu- demand of the process will result in temporarily insuf-
ally proportional to the fuel flow, so that F1(x) func- ficient or excessive burner load that may cause other
tion may be just a scaling multiplier. The functions problems including system trips.
are appropriately scaled ratios of the flows over the The main advantage of fully metering controls
load range defined upfront (based on theoretical is that with sufficient slowness the controls will
maintain their accuracy even with poorly repeatable magnitude and accuracy of measurement, but carries a
controlling elements. substantial penalty of increasing the pressure upstream
Having metering controls allows adding additional of the device that may not be available. For the air flow,
layers of safety to the process. From the safety stand- the impact on increased pressure may require increasing
point, fuel-rich operation of the boilers is more danger- the power of the combustion air fans (see Chapter 13).
ous than fuel lean. If the flame is lost due to becoming With the cost, space, and flow energy loss limitations, it
too lean—insufficient fuel—the consecutive actions of is usually impractical to design the system such that the
the BMS will result in the fuel cutoff and then purg- peak pressure signal would be more than a few inches
ing the system with air, causing further dilution of the of water column at peak flows for air measurements.
residual unburned fuel components, making it impos- There are certain requirements to the length of the
sible to reignite. On the other hand, the loss of flame straight run in the section of the duct upstream and
under the rich fuel condition followed by the fuel cut downstream of the metering orifices or nozzle. For
off will initially result in more favorable conditions for accurate measurements, the overall straight run may
fuel reignition and possible explosion as the flow of air need to be as long as ~24 duct equivalent diameters.
continues. The requirements for the length of the straight run can
In order to minimize the chance of dynamic errors be reduced with special flow conditioning devices and
resulting in fuel-rich combustion conditions, the meter- multiple pressure sensing points, but they also add to
ing controls are usually cross limited between fuel and the costs and possibly to the parasitic losses of valuable
air, making sure that in dynamic situations the devia- pressure.
tions toward fuel-rich conditions are limited to only In combustion systems, the difficulty is usually with
a few percentage points from the normal operation. accurate and repeatable metering of the air and FGR
As illustrated in Figure 12.6, the cross limiting holds, without creating substantial unrecoverable pressure
increasing the fuel valve opening or reducing the air losses through the metering device and without the use
flow when an instantaneous air deficiency of more than of bulky and expensive arrangements.
a specified amount is detected. Most of the air and FGR flow measuring devices are
Measuring flows over extended load changes while based on processing of the pressure signal generated
maintaining the accuracy presents substantial chal- by devices based on the Pitot tube principle, piezoelec-
lenges as many flowmeters are rather limited in range. tric ring, metering orifice type restriction, or venturi-
Extensive literature exists on the subject of flow mea- shaped nozzle. Using other types of air flow metering
surements, see Ref. [5] as a good example. devices in boilers historically did not withstand the
Metering of fuels, whether in gaseous or liquid form test of time due to reliability, accuracy, and mainte-
is usually easier than metering combustion air. The fuel nance issues. An example is the heated wire type mass
flows are typically 10–20 times less in mass, so its flow- flowmeters extensively marketed about 30 years ago
metering devices are relatively more compact than the air that went out of common use in industrial combustion
flowmeters. Secondly, fuels are usually available with subsystems later.
stantial pressure allowing operation with higher veloci- The most efficient way of measuring combustion air
ties and less concern to unrecoverable pressure losses in flows is with the use of fan inlet duct silencers with
the metering device. A variety of different devices can be multiple passages, each resembling a flat venturi. In
used for metering fuel flows described in Section 12.12. addition to reducing noise, the silencer can easily be
However, predominantly orifice plates or venturi nozzles shaped to perform flow conditioning and fit in a rela-
are used due to their simplicity and reliability. tively small space.
The differential pressure signal across an orifice plate Another approach is using piezoelectric rings placed
or generated by a venturi nozzle is proportional to the in the inlet cone of the combustion air fan. The benefits
second power of fluid velocity through the element that of the ring are compactness and minimum parasitic
gets rapidly reduced as the load goes down. For added losses of pressure. The ring will generate a signal char-
accuracy, the signal may need to be corrected for tem- acteristic of the flow going into the fan that may include
perature and, in some cases, for pressure. The effects induced FGR. The drawbacks are the limited signal
of the Reynolds numbers on the device outputs may magnitude and potential errors due to unconditioned
also need to be considered. The practical range of these or poorly conditioned flow upstream of the fan inlet
devices depends on the magnitude of the signal that can and possible temperature nonuniformity in cases with
be generated at high fire, the accuracy of the pressure induced FGR. These errors may be a source of consider-
transmitters, and the magnitude of different distur- able nonrepeatability of the measurements.
bances. Increasing the fluid velocity through the meter- Due to a variety of noise issues dealing with pres-
ing element reduces its cost and increases the signal sures of less than ~0.05–0.1 in. (1.3–2.5 mm) of water
column is not reliable. Measurements at these low There are several common types of trims that can be
pressure levels become difficult even if the character- added to parallel positioning controls. One is metering
istics of a pressure transmitter allows it. Due to the trim configured as illustrated in Figure 12.7. This type
energy loss limitations, it is also common to have a of controls may also be called fully metering controls
maximum pressure signal magnitude at high fire in with parallel prepositioning and cross limiting between
the range of 2–4 in. (5–10 cm) of water column. This, fuel and air.
along with reasonable requirements to accuracy at When properly applied, these controls may react to
the low end, puts a limit on the metering controls the load changes almost as fast as parallel position-
turndown to about 4–1 to 7–1. If higher turndowns ing controls and maintain the accuracy and safety fea-
are required, the controls logic is designed to switch tures of the metering controls. Some dead band in the
at the low end to parallel positioning where require- cross limiting action defined by a percentage basis in
ments for the accuracy of maintaining the flow ratios the schematic by δ is necessary for the scheme to func-
can be more relaxed. In some specific cases, the range tion. When δ is set too low, the action of the cross limit-
of pressure-based metering controls can be extended ing may contribute to slowing the rate of load changes
with the addition of a variable restriction—a damper, below the demand of the process. On the other hand,
for example—that can be put in a substantially closed excessive δ may under some adverse conditions reduce
position with a characterized resistance before the the accuracy of the controls.
flow drops below 20%–25% of the maximum. Another type of a common trim is stack O2 trim,
It is important to understand that in many cases, using measured stack O2 to trim the excess air level
the absolute accuracy of flow measurements is not as delivered to the burner. With this, a PID loop is added
important as its repeatability. Usually the majority to adjust the air flow to the combustion air to minimize
of all systematic errors can be eliminated during the the difference between the measured stack O2 and its
burner setting process. In cases where actual flows target value set as a function of load. The advantage of
are also important, the system itself can often be used this trim is its potential accuracy if the oxygen probe
for calibration of the flowmeters during the commis- is properly maintained. The disadvantage of this trim
sioning process. For example, with a known fuel flow, is that it reacts with an additional delay becoming sub-
its stoichiometric air requirements and oxygen con- stantial especially at reduced loads. It may take con-
centration in the exhaust, it is easy to determine the siderable time—up to 60 s or more in some cases—for
mass flow of combustion air reaching the burner and the changes in the fuel-to-air ratio at the burner to be
use it for the air metering calibration when necessary. reflected at the stack O2. As a result, the O2 trim feed-
Of course this calibration method would only work back loop must be set to react slowly. The effects of
well when leakages of air in or out of the system are delays may be reduced when the trim is applied in
insignificant. addition to the metering trim. Having multiple trims
in the system, however, often requires proper consid-
eration of possible interactions between different PID
12.3.4 Parallel Positioning Combustion
loops and analysis of root causes for the stack O2 and
Controls with Trims
air fuel ratio deviations. In a similar manner, metering
In this arrangement, the overall controls will operate as trimming actions can be applied to make adjustments
parallel positioning with adjustments to maintain the to the FGR flow or to the combined flow of air and FGR.
target ratios computed based on the information com- The physics of the process including the fan laws (see
ing from the sensors (flowmeters, pressure transmit- Chapter 13) need to be considered in order to design
ters, temperatures, and oxygen sensors). The trims are fast and reliable ways of trimming the flows. Going into
applied via the actions of the PID loops that can be con- specific details of these controls and their optimization
structed in multiple ways. The best, however, are those is outside the scope of this chapter. Accuracy of prop-
arrangements that minimize the action of cascading PID erly designed parallel positioning controls with trims
loops, leaving only minor corrective trim functions. In depends much less on the repeatability of the flow con-
order to achieve the trimming functions, they should trolling components than with parallel positioning only
be constructed with the proper consideration of the controls and can negate the effects of gradual slippage
physical effects of different variables and causes of flow in the dampers or valve characteristics. Ideally, only the
deviations from target values. Ultimately only the small hysteresis type effects in the valves and dampers may be
nonrepeatability in the actuator positions may need to the main remaining source of inaccuracies in dynamic
be compensated with the PID loops, allowing accurate situations. The hysteresis effects must be minimized in
controls performance even when the demand under- all cases as these errors will continue to create limits on
goes rapid changes. the dynamic accuracy of the controls.
Air Fuel
Master flow flow
demand
FT FT
A
÷ ÷
Δ ÷
f3(x)
Σ
>HI >LO
1.0 >HI ÷
1.0 >LO Manual

× trim
HMI
Manual trim
Metering active 1.0
Manual air Manual fuel Δ
trim range
HMI D TC TR A
control f1(x) control f2(x) (service MAN
active Manual air active Manual fuel Metering trim
access) K ∫
control control active
HMI HMI
HMI HMI 0.15 HMI D
T 50%
A
T A T A
HMI
× Δ
Air Fuel
control control
device valve Σ 1.0 ÷
FIGURE 12.7
Fully metered controls with parallel prepositioning and cross limiting between fuel and air.
All these specifics make operation of ultralow NOx

burners more sensitive to any deviations of the speci-
12.4 Combustion Controls for Systems
fied firing regimes from its optimum settings.
Requiring Minimum NO x Emissions: The excessive deviation from initial optimal settings
Ultralow NO x Controls may reveal itself in multiple ways. First, loss of com-
Maintaining minimum emissions and maximum effi- pliance with the emission requirements may occur.
ciency requires precision combustion controls for the Second, substantial instabilities, vibration, and intense
fuel air ratio and FGR rates. The main mechanisms acoustical noise may occur and ultimately loss of flame
of controlling NOx emissions (see Chapter 10) to very and emergency shutdown. In some cases, loss of flame
low levels are FGR, fuel or air staging, utilizing lean can take place without any apparent warnings.
premixed combustion when firing gaseous fuels, and In general, the lower the NOx emission the burner is
in furnace recirculation of reduced temperature com- designed and set to operate, the narrower is its perfor-
bustion products into the combustion zone. Achieving mance envelope with respect to deviations in the excess
very deep reductions in NOx emissions in comparison air and the FGR rates. Figure 12.8 gives an example of
to uncontrolled levels (up to 90%–95% reductions in operating ranges and performance envelope of one of
some cases) requires using a combination of these tech- early versions of a single zone lean premixed burner
niques and special burners. Some of these burners are when firing natural gas with ambient temperature air.
described in other chapters of this book (see Chapters Figure 12.9 gives ranges and performance envelope of
14 and 15). The ultralow NOx burners are typically more a more refined lean premixed burner with auxiliary
complicated, have multiple fuel zones and may require fuel zones.5 Deviations from the shown limits toward
varying fuel distribution between the zones with load. higher excess air or higher FGR would first create flame
60
Flammability limit in a well stirred reactor
50
50% firing rate

40
25%–40% firing rate
5
FGR, %
30 NO
x=9
7
pp Operation at
m,
cor minimum fire
r. 3
%
12
O
2
20
15
10
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
Air to fuel stoichiometric ratio
FIGURE 12.8
Operating ranges and NOx performance of a single zone lean premixed burner.
60
Flammability limit in a well stirred reactor
50
Over 50% firing rate
40
2
25%–40% firing rate

FGR, %
30 NO
x =9
7
pp
m,
co
rr.
12
3%
20 O
2
15
10
Operation at
minimum fire
(8, ..., 15%)
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
FIGURE 12.9
Operating ranges and NOx performance of a lean premixed burner with auxiliary fuel zones.
instability—barely anchored to the burner flame— fuel pressure, these low Reynolds numbers are rarely
possible low frequency pulsation, and a complete loss the case, except as may occur at very low loads.
of flame. Operating with insufficient air or FGR also If the fuel composition varies within small ranges,
may result in flame pulsation, often with higher fre- but the combustion controls system does not account
quencies, as the mixture of air flue gas and fuel becomes for this variation, the burner heat input may be deter-
more flammable. As seen from these examples, operat- mined with the error proportional to the deviation
ing ranges of ultralow NOx burners narrow down as in WI. In most cases, the change in PVs will call for
the load reduces. This puts additional demands on the the load to be brought to the level to match the heat
accuracy of combustion controls at the low end where it demand and the whole process may run just fine. The
is already more difficult to achieve. concentration of excess oxygen in the exhaust, how-
The required accuracy and repeatability of the ever, will deviate but may be corrected if the stack O2
ultralow NOx controls also prompts designing it with trim is applied.
additional layers of safety and redundancy for mini- Direct measurements of WI are based on calorimet-
mizing chances of emergency trips due to actual loss ric methods that are rather slow (over 20 s delay), but
of flame and nuisance trips. These additional layers of may be used in lieu of in parallel with stack O2 when
safety will be described in Section 12.8. placed well upstream in the fuel line, or when accurate
computations of the heat input are needed. Knowing the
WI, however, does not necessarily allow the combustion
process to accurately compute the air required, as the
stoichiometric air requirements are not exactly propor-
tional to the heat input. For example, stoichiometric air
12.5 Special Cases requirement for the same heat input of carbon monoxide
(CO) is 29% less than for methane (CH4) if based on LHV
12.5.1 Controls for Applications with Fuels and 21.1% less if based on HHV. The same numbers for
of Variable Composition hydrogen (H2) are 22% and 17%. For higher saturated
Controlling combustion of fuels with variable com- hydrocarbon gases (C2H6, C3H8, etc.) the deviations from
position creates additional difficulties due to chang- methane (CH4) are less than 2% and somewhat higher
ing requirements to the fuel air ratios, rate of FGR if for unsaturated hydrocarbon gases like ethylene (C2H4)
used, computations of the heat output, etc. Knowing and propylene (C3H6).
or measuring some bulk fuel characteristics, however, To resolve this issue, a different characteristic called
may help to design the controls that account for most the combustion air requirement index (CARI) is used
of these effects. in the controls to maintain the air fuel ratios. CARI is a
One such characteristic is Wobbe index (WI) of the dimensionless ratio of stoichiometric volumes of air and
fuel. It is defined as a ratio of the fuel heating value per fuel multiplied by the square root of ratio of air to fuel
unit volume (H) to the square root of its specific gravity specific gravities:
(ρf) at some normalized conditions with respect to the
pressure and temperature: V  ρa
CARI =  a  × (12.5)
V
 f  stoich ρ f
H
WI = (12.4)
ρf 0.5
When the air flow is controlled as a function of the fuel
metering signal generated based on the orifice or nozzle
H can be defined based on the high heating value (HHV) differential pressure (temperature and pressure com-
or low heating value (LHV) of fuel. pensated) multiplied by CARI, the stoichiometric ratio
For a combustion device operating with fixed geom- between the fuel and air will not be a function of fuel
etry of all the fuel passages (usual case), the heating composition.
value of the fuel flow delivered with a certain pres- Short of analytical instruments that would be poorly
sure to the burner will be closely proportional to WI. suited for integration with industrial combustion con-
The same relation holds between the heating value of trols, CARI can be determined with special instru-
fuel passing through a metering orifice and the orifice ments operating on the principle of catalytic oxidation
differential pressure. Exceptions are flow rates with of fuel and measurements of O2 in the exhaust. The
relatively low Reynolds numbers (less than ~4000) in methodology has a certain degree of theoretical error
the passages with the area primarily defining the flow in converting the measured O2 to CARI especially for
rate. In industrial systems operating with substantial the fuels with high CO, as explained in more detail in
Section 12.6. However, the measurement of approxi- simple, but generally results in increased static pres-
mated CARI can be substantially faster than measur- sure requirements from the combustion air fan with
ing calorimetric measurements of WI. exceptions when a small amount of FGR is driven by
Using CARI instead of WI leaves some uncertainty in suction inside the fan inlet cone. Another advantage
computing the burner heat input, but in most cases, the of the arrangement is that even without the modulat-
exact knowledge of the heat input is less important as ing dampers, the FGR flow may be substantially self-
the firing rate is adjusted to the demand and, in many controlled as increased air flow through the fan results
cases, can be determined indirectly based on produc- in increased suction and near proportional increase in
tion rate, efficiency, calculations, etc. Alternatively, the FGR flow. In reality, modulating dampers are still
WI can be estimated using the correlation between needed in most cases for overcoming stack draft effects,
the approximately measured CARI and WI that holds changing the percentages of FGR flow (computed as a
relatively closely for many practical fuels containing portion of created combustion products) with load, and
mostly hydrocarbons, but deviates somewhat when H2 compensating for the changes in FGR temperature.
and CO are present in fuel in substantial quantities (see FGR can also be added into the system downstream of
Section 12.12.5). the combustion air fan with the help of a separate FGR
The necessity of using either CARI or WI instru- fan—forced FGR. This arrangement potentially can be
ments in the controls needs to be carefully evaluated. more energy efficient than induced FGR. This will be
If the changes in fuel composition are relatively slow, the case when the addition of forced FGR impacts the
input from a stack O2 probe may be sufficient for mak- combustion air fan static pressure requirements less
ing adjustments to the air fuel ratio in the controls. At than the induced FGR arrangement. Another possible
the same time measurements of the combustion air flow benefit is that forced FGR may be more effective (less
and fuel (combined with the stack O2) would allow com- FGR for the same NOx reduction) than induced when
puting the actual (not approximated) CARI that has a added to specific areas of the flame. However, the forced
better correlation with WI. FGR schemes are substantially more complicated and
The aforementioned considerations leave very few require a second fan. This option is often considered in
niches for using CARI and WI analyzers in boiler com- retrofit applications, as it may help to reuse the exist-
bustion controls instead of stack O2 trims combined ing combustion air fan and add only a smaller FGR fan,
with air fuel metering controls. One niche is when or when the burners are specifically designed to oper-
fast reaction to changing fuel composition is needed; ate with a separate FGR flow injected directly into the
another is when readings of stack O2 poorly represent burner. From the controls standpoint, forced FGR cannot
actual excess air delivered to the burners. be substantially self-controlled. On the contrary, with-
The real difficulty when designing combustion con- out restricting the FGR flow or reducing it by slowing
trols for systems with variable composition fuels is the fan (fans with VFD), the FGR flow would increase
maintaining low NOx emissions. Different fuels gen- due to the generally increased fan static pressure and
erate different levels of thermal and prompt NOx and reduced pressure in the system. This and the fact that
may have very different flammability characteristics the burner operation may be sensitive to the FGR rate
impacting operation of low NOx burners. Navigating puts increased demand on the accuracy and reliability
through this complexity is usually done on a case by of controlling the flow of forced FGR.
case basis and requires knowledge of the specific burner
limitations.
12.5.3 Boilers with Multiple Burners
A specific analysis of a possible range of fuel composi-
tion variation is usually needed to achieve better utili- Operation of boilers with multiple burners with the
zation of approximated CARI and WI or stack O2 for the requirement to be able to add or remove burners from
purposes of combustion controls. service without shutting the system off or going first to
low fire presents additional challenges to the combus-
tion controls. Boilers with multiple burners usually have
12.5.2 Systems with Flue Gas Recirculation
a single FCV for all burners and shutoff valves at each
There are several ways of recirculating a portion of the burner. The air flow is also controlled at a common place
flue gas from the system exhaust back to the burner with isolation on-off dampers at each burner. These boil-
(FGR). The simplest is induced FGR when FGR flow ers are usually equipped with metering controls that can
is mixed with combustion air upstream of the burner. also factor in the number of burners in service as well as
The FGR flow in this case is driven by a negative pres- the number of burners open for the combustion air flow.
sure created by controlled restrictions in the passage The problems with adding or removing burners from
of the air flow upstream of the fan. The concept is very service can be explained with the example of a two
burner boiler. Suddenly starting or stopping the fuel the only means of controlling the delivery of combustion
flow to a second operating burner would instantaneously air and not to rely on the dampers. However, one needs
change the fuel to air ratio to the other burner with a fac- to be very careful when deciding to use only the VFD
tor of two creating upset conditions that will gradually instead of the dampers for controlling the combustion
be corrected as the controls react to balance the air and air. Excessive reliance on VFD fans may reduce pulsation
fuel to the burners in service. A similar disturbance takes dampening effects in the system and cause combustion
place during opening or closing of one burner air isolation instabilities. In many low NOx burners, the propensity to
damper but on a slightly smaller scale as some air in the generate instabilities increases at lower loads. It is recom-
amount as high as 50% passes through the closed damper mended that all VFD applications include at least one air
(a certain amount of air shall always be passing through side damper.
the out-of-service burner to protect it from high tempera-
ture furnace gas and thermal radiation). These transi-
12.6.2 Modulation of Flows with Inlet Vane Dampers
tional regimes present no problems when the burners are
designed to operate within these created disturbances. Using an inlet vane damper (IVD) mounted at the air
For boilers with low NOx burners, however, the dis- inlet of the combustion air fan for the flow modulation
turbances may be too high to reliably operate with these also allows some energy savings when compared to the
created disturbances without losing the flame and shut- dampers downstream of the fan. Fan manufacturers
ting down the burners. Certain steps can be taken in the often state that for avoiding surge effects, fans shall oper-
design and logic of the boiler controls that will minimize ate only to the right of the maximum static pressure on the
the magnitude of deviations from normal operating con- fan performance curve.6–11 If a surge develops, it becomes
ditions and/or shorten the duration of the upset by orders impossible to accurately control the flow. However, in
of magnitude. The most obvious steps are minimizing most combustion applications with modulating loads,
the fuel to air ratio prior to opening air to the additional the requirement to operate only on the declining part
burner and then reducing the ratio prior to opening the of the fan characteristic is impossible to meet and is not
next burner fuel isolation valve. When one of the burners always required. Typically a curve peak is in the range
needs to be taken out of service, the steps are minimiz- from 40% to 70% of the maximum demand. Controlling
ing the fuel air ratio prior to closing the burner fuel isola- the flow with an IVD seemingly improves the situation
tion valve and then maximizing it prior to isolating the as with partial IVD opening the peak shifts to lower
air. In case of a loss of a burner due to a detected loss of flows. However, the shift may still not be enough to
flame, the fuel and air to this burner need to be closed at operate only to the right of the peak over the entire load
the same time. The entire situation becomes immensely range. This is fine as surging is less of a problem at lower
more complicated with high FGR rates and/or very low loads.10 At the same time, removing a closely coupled fan
NOx operation. The controls may require feed forward discharge damper negatively changes acoustical interac-
information from the BMS and provide fast reaction to tion between the fan and the duct downstream, increas-
reduce FGR/air rates depending on the number of burning the likelihood of surge problems. Designing energy
ers in service. Other options would include parallel posi- efficient controls for the combustion air flows that will be
tioning action to the FCV computed based on the known free from surge problems requires certain expertise and
characteristic of the valve. analysis applied in each specific case.
12.6 Modulating Combustion Air Flows

12.7 Sizing Controls Components
12.6.1 Fans with Variable Frequency Drives
Selecting appropriate sizes for combustion controls
With substantial advances in reliability, performance, and components—fuel valves and dampers—is important
cost reduction of variable frequency drives (VFD) of AC for the accuracy of the controls and, in many cases, for
motors, the use of VFD fans for supplying combustion the amount of energy that would be required from the
air is becoming more and more common. In systems that combustion air fan or a fuel pump.
operate at partial loads most of the time, the VFDs offer When designing combustion systems, engineers usu-
substantial energy savings. Some users are even requiring ally have a very good idea regarding the maximum flow
VFDs at their sites. The modern VFD fans can operate with a valve or a damper needs to pass and some approximate
a very wide turndown and, when equipped with dynamic idea regarding the limitations on how much pressure
breaking, can be controlled with high accuracy. These loss across the flow controlling element they can afford
qualities make it tempting to use fan speed modulation as without restricting the system capacity. It is a common,
but very poor practice to size components based on the like this is indicative of insufficient pressure available
line sizes. The criteria for the line size optimization are for the controlling element, usually a very small portion
different from the controls elements size optimization. of the total available pressure. In piping trains, oversiz-
So, while the sizing of lines and controlling devices is ing the lines and adding gradual reducers upstream and
related, the priority should be given to the controls ele- downstream of the valve, increasing the supply pressure,
ments sizing. In the end, some lines may have the same or revising the burner design to lower pressure require-
size as the valves or dampers, but not necessarily. In ments are often better alternatives than compromises
many cases, a properly sized element will be smaller than with controllability of the process.
the lines. With a properly selected device, its modulation Valve manufacturers often strive to achieve a close to
range between minimum and maximum flows during linear characteristic of the flow rate or so-called equal per-
the system operation shall not be substantially less than centage characteristic with respect to its position/opening
the full range of the available movement. Regardless of that is normally defined for a fixed pressure drop across
the size, the controlling device should have good repeat- the valve. However, in combustion systems, this is rarely
ability, minimum hysteresis effects, etc. the objective and a fixed pressure drop across the valve is
A properly sized controlling device in combustion seldom the case. The demand for accuracy of controlling
applications is a device with the smallest cross section flows at reduced loads generally increases as the goal of
that meets the limitations for pressure loss in the maxi- the controls is to maintain the necessary ratios more than
mum open position with more or less linear characteris- the actual flows. At the same time as the load goes down,
tics in the mid-range and a gradually reducing slope of the pressure downstream of the valve usually decreases
the flow characteristic when approaching the fully closed and the valve pressure drop increases, adding difficulty
position. The valve industry would use a term called to maintaining the accuracy at low end.
equal percentage to describe a characteristic like this.6 Properly sized butterfly type valves with airfoil-
An oversized device will be in less than 50%–60% open shaped blades closely matching the valve body internal
position at the maximum flow. When a valve or damper diameter are usually the best for meeting these criteria.
is oversized, a small change in its opening substantially Butterfly valves with additional seals over the perim-
changes the flow, putting higher demand on the position- eter of the valve opening increase the valve tightness in
ing accuracy that may be difficult or impossible to meet. closed position, but may negatively affect its character-
Selecting a smaller size device in this case easily solves the istic when approaching fully closed position. A typical
problem. A valve or damper may not be adequate for the characteristic of a butterfly type flow controlling device
process, even if not oversized, when the flow characteris- is shown in Figure 12.10.
tic (flow vs. opening) has a very shallow asymptotic slope Simple plug type valves usually do not have as good
over a substantial range starting at 40%–50% of the open- a flow characteristic as butterfly valves. In order to
ing, for example. The effect is similar to the prior valve improve its characteristic, the final flow passages in
example that does not have equal percentage character- these valves are usually made in a V shape with only
istics where most of the changes to the flow are taking the apex available for the flow before the valve closes
place within a smaller portion of the travel. A situation completely as shown in Figure 12.4.
160%
140%
Flow through the valve as a percent
120%
of nominal flow
100%
80%
60%
40%
20%
0%
0% 25% 50% 75% 100%
Valve opening %
FIGURE 12.10
Typical characteristic of a butterfly type valve.
When sizing the valves, consideration also needs to be of many system components prior to the burners’
given to the noise created by high velocity flow, erosion, light-off as well as supply of sources—fuel, combus-
corrosion, condensation of liquids and, in some cases, tion air, etc.—with adequate pressure prior to and dur-
to the external and internal icing resulting from tem- ing the operation. As the system undergoes a light-off
perature changes due to local adiabatic cooling effects. sequence, each step of this sequence gets confirmed
Similar concerns engineers face when sizing the damp- by a specific switch or indicator. The typical steps are
ers with much higher emphasis, however, to minimiz- as follows:
ing parasitic losses of pressure that would need to be
overcome with selection of higher static pressure com- • Confirming the initial closed position of the
bustion air fans. SSVs and open position of the vent valve (where
Sizing the dampers is usually more difficult than applicable).
sizing the control valves. Valves come in distinct • Confirming the fuel supply pressure and other
sizes with closest sizes often varying up to two fold utilities (atomizing steam or air, for example)
in the passing capacity. Dampers, on the other hand, are within the design limits.
can be sized exactly for the application. There are,
• Confirming that combustion air flows through
however, some design constraints related to the size
the system in sufficient quantity and duration
of connecting ducts that may require the addition of
when purging the system.
properly designed transitions. So, the ducts them-
selves need to be carefully sized when designing the • Confirming that the fuel valves and air and
system. Contrary to the valves, dampers often have a FGR controlling elements and the drafts control
rectangular cross section and for their compactness (if applicable) are in positions corresponding to
and mechanical strength often have multiple damper the light-off load of the burner.
blades connected together by mechanical linkages. • Confirming the presence of the pilot flame.
Multiblade dampers have much better characteristics • Confirming the vent valve has closed.
with appropriately designed fixed plates separating
• Confirming the opening of the SSVs.
flow passages around each rotating blade, making
the overall damper to perform as a butterfly damper. • Confirming the pilot has shut down.
The fixed plate may also be designed to perform some • Finally, after the fuel valve is closed and the
flow conditioning function.7 SSVs open and the burner igniter/pilot is shut
Additional concerns with sizing the dampers and down, the BMS starts monitoring the presence
ducts are possible strong disturbances to the flows of the burner flame as well as supplies of critical
created by the dampers. At partial loads, the dampers utilities.
become the prime consumers of static pressure gener-
ated by the fans creating strong jets inside the system All the aforementioned steps are well formalized by the
when the useful flows are reduced and more suscep- applicable safety codes like NFPA.3
tible to disturbances. Unless properly dissipated, these In the earlier days of BMSs, limit switches and inter-
effects may create additional complexity for control- locks were usually hard-wired in series to the master
ling the combustion process and negatively affect the fuel trip relay. This provided a high level of confidence
performance. that a limit going off-normal would cause the burner
to trip by directly causing power to be removed from
all of the fuel shutoff valves. Later, as electronic pro-
grammable logic devices became popular, these hard-
wired trip circuits were still used, but the logic system
12.8 Burner Management Systems was added primarily to handle sequencing (replacing
hardwired timers), alarm detection and a redundant
It is common to think that the BMS provides the first means of shutting down the system. Eventually, as
layer of safety to the combustion system. However, logic systems came into more mainstream usage and
the very first layer of safety is the boiler or furnace experience has been gathered to prove their reliabil-
itself that has to be designed to withstand some ity, the hardwired limit circuits have been eliminated.
combustion instabilities and even minor puffs with- Limit switches and interlocks are now usually wired
out sustaining extensive mechanical damage to the directly to logic system inputs. The benefit of this is
structure. that much more advanced alarming and diagnostics
The BMS provides several main layers of safety to the can be obtained than through the traditional hard-
combustion system. Its first layer monitors functioning wired approach. However, with the elimination of
the ability for these limits to directly de-energize the

fuels valves (via the master fuel trip relay), the logic
12.9 Layers of Safety
system must be designed so that it can provide at least
the same level of confidence of detecting a limit or As discussed previously, the first layers of safety are the
interlock going off-normal. Programmable control- boiler construction followed by a BMS system. However,
lers, often of the same type as used for the controls, even a structurally sound boiler and a properly func-
are used for implementing all the steps of BMS logic. tioning BMS designed per modern safety codes together
For additional safety, the BMS controllers are often may not be sufficient in some cases to guarantee com-
required to not be integrated with the combustion plete safety of operation.
controllers.3 In the United States, BMS systems are typically
Flame monitoring can be considered as the most designed to react within 2–4 s after the scanners stop
critical layer of safety in modern boilers. It relies on detecting the flame. Systems can also be designed to react
flame detectors—typically optical flame scanners in one second. It may take another second or a fraction
allowing operation to continue only with continuing of it for the SSVs to respond, stopping the fuel flow com-
detection of the flames. There are two types of flame pletely. Due to the naturally highly fluctuating structure
detectors: optical scanners and flame rods; these of the flame, a reliable detection of the loss of flame event
devices are designed to be fail-safe. The flame rods in less than 2 s may be difficult. In any case, current BMS
detect the presence of the flame based on gas ioniza- controls have reaction times that are high compared to
tion by the flame that is in contact with the electrode the flow of fuel and air into the boiler in the range of 1–2 s.
(rod) and the ground (usually any of the burner metal Many modern boilers have a very compact design with
parts). Optical scanners monitor intensity and pulsa- the heat release rate of over 100 kBtu/h/ft3 (over 1030 kW/
tion signature of radiation emitted by the flame in m3) based on the furnace volume. In such a furnace, deto-
different spectrum ranges from infrared to ultraviolet nation of high fuel flow accumulated even in 1 s after it
depending on the type of fuel fired. This monitoring actually stops burning may cause a spike in pressure of
takes place at a distance from about 1 ft to as long as over 1 atm (1 bar) that would dissipate over a ~50–100 ms
10 ft (3 m). Modern scanners can be very sophisticated duration or faster. This is more than sufficient to create
tunable devices equipped with fail safe and frequent substantial damage to the boiler and surrounding area.
self-checking functions. Flame scanners in general Thus, deviations in the system leading to such emer-
are considered superior and more reliable than flame gency shutdowns should be avoided as much as possible
rods and definitely more versatile. Figure 12.11 shows even if it leads to some potential nuisance trips.
COEN i-scan™ flame scanner mounted through a ball The situation may also be problematic in case of high-
joint to the burner front plate. Figure 12.12 shows some fire emergency shutdown caused by the closure of SSV
details of the signal processing by the i-scan scanner due to nuisance trips as well. Such trips may be initiated
done with manual or automatic setting: DC and AC by the combustion controls system based on the magni-
components of the flame signature, frequency spec- tude of detected abnormalities other than the actual loss
trum computed using fast Fourier transform algo- of flame, or by the BMS due to a failure to control the
rithm, automatic selection of frequency range with load, steam pressure, water level, pressure of supplied
the best signal to noise ratio, and safeguards against utilities, etc. Besides interruptions to the production pro-
different kinds of operator/installation errors and cess, a sudden substitute of the high temperature com-
hardware malfunction. bustion products with about 5 times smaller volumetric
FIGURE 12.11
COEN i-scan™ flame scanner.
FIGURE 12.12
Example of the setup panel of COEN i-scan™ flame scanner.
flow of cold air would cause an implosion situation excessive combustion air or FGR, the bulk flammability
with a pressure drop on the order of 0.05 atm (~20 in. of the mixture coming into the furnace is reduced. This
WC) followed by a smaller positive pressure spike. The reduces the likelihood of mixture detonation within
characteristic time for the implosion process is on the the next few seconds after the loss of flame. The loss of
same order of magnitude as during detonation inside flame due to excessive combustion air or FGR does not
the furnace. Such an implosion will not likely cause any guarantee that the created mixture will not reignite in
major structural damage, but may cause intense shak- the furnace from high temperature objects like refrac-
ing of the boiler walls, breaking of some small welds, tory or the remaining pockets of high temperature
cracks in refractory packing, etc. Figure 12.13 shows an combustion products, but the intensity of the result-
example of computed pressure signatures inside a typi- ing pressure spike would be reduced. It is important to
cal 150,000 lb/h (68,000 kg/h) package boiler that would mention that in systems with FGR, delays in shutting
be developed after a loss of flame at high fire with and the burner down on the loss of flame longer than the
without a subsequent detonation computed using com- residence time of FGR passing through the system will
pany proprietary programs.6 result in replacement of the FGR flow at the burner with
It is reasonable to expect/require that all new boilers a blend of air and unburned fuel making the overall
be built to withstand, with good safety margins, implo- flow coming into the furnace more flammable than at
sion during high fire shutdown without sustaining any the moment the flame was lost.
mechanical damage. The costs of making boilers to meet It is difficult with the possibility of multiple failure
this criterion would be insignificant in comparison with modes to design an absolutely safe and highly reliable
the overall system costs. industrial combustion system at the same time with-
The actual loss of flame during firing may take place out exceeding reasonable cost limitations. It is possible,
when the operating regime deviates from normal however, to enhance the safety by adding additional
toward too fuel-rich or too fuel-lean conditions or when layers of checks and balances using information derived
the FGR rate becomes excessive. If the flame is lost due to from available combustion controls sensors or just a few
0.15 0.3
0.125
0.1 0.2 Furnace pressure

after flame
Furnace pressure, atm (gauge) 0.075
detachment from
0.05 0.1 the burner w/o
detonation, atm
0.025 (left axis)
0 0 Peak furnace
pressure during
–0.025 detonation
depending on its
–0.05 –0.1 delay after the
–0.075 flame detachment,
atm (right axis)
–0.1 –0.2
–0.125
–0.15 –0.3
0.00 0.05 0.10 0.15 0.20 0.25 0.30
Time from flame detachment, s
FIGURE 12.13
An example of furnace pressure after the moment of flame detachment with and without subsequent detonation for high fire operation of a
typical 150,000 lb/h package boiler.
additional sensors added to the system as described a signal prompting changes in the FGR or alarming
next. or even shutdown. In some cases, combustion systems
If the flame is lost due to the mixture being too fuel can be designed with additional sensors allowing for
rich and then the fuel flow is cut off, the continuing cross checking of the burner “health,” making sure its
flow of air through the burner will immediately make mechanical integrity and performance are not com-
some volume of gas in the furnace more flammable promised. All these techniques have their plusses and
again creating conditions for possible reignition from minuses and need to be selected based on the system
some source in the furnace. Acknowledging this inher- specific requirements, configuration, demand for accu-
ent danger of loss of flame with fuel-rich conditions, the racy, reliability of the components, etc. and weighed
safety codes require combustion controls with metering against the drawbacks of increased complexity and its
of fuel and air to cross limit the flow of fuel to the flow of impact on the availability of the system.
air when changing loads. Combustion controls may also One of the relatively new techniques that can add
be designed to monitor deviations in the air fuel ratio both safety and reliability to the system is a combus-
regardless on its direction. The excessive deviations tion instability detector. In most of the cases, this will
may first cause an alarm and then a trip if a higher devi- provide a few seconds advanced warning before a flame
ation threshold is reached. This works somewhat dif- failure. If designed as a fully integrated separate device,
ferently than classical cross limiting. The cross limiting it may work like a BMS switch in parallel with the flame
may slow and even freeze the load change completely in detector. However, in most cases, it is simpler to inte-
the case of a valve or a damper malfunction. The freeze grate the instability detector into the combustion con-
in load may consequently cause excessive steam output trols logic. With the latter, the instability detector can
with a trip on high steam pressure or process tempera- also be tasked to automatically alter combustion control
ture or a loss of output below the process demand. settings, making an attempt to avoiding a possible sys-
In systems where it is possible for the FGR to devi- tem trip altogether. In any case, the instability detector
ate by excessive amounts, the controls can be designed should not be a substitute for the flame detector.
with direct or indirect monitoring of these deviations. It should be recognized that setting and tuning of the
The proper FGR rates can be assured by many differ- combustion controls may be a very complicated pro-
ent means short of direct metering the flue gas flow. cess performed with all or some of the controls func-
For example, in systems with induced FGR, limits can tions in manual mode. This opens the possibility of
be set to the deviations in the position of the inlet box input errors. Causes of the errors may be HMI malfunc-
damper critical for the FGR flow. Excessive deviation in tion, instrument/sensor calibration, actuator/position-
the oxygen concentration in the combined flow of air ers malfunction, some initial errors in the algorithm
and flue gas delivered to the burner may also generate or simply human mistakes. To minimize the chance of
these errors, the controls logic may include additional beginning of the burner setup process. This step usu-
safety checks—barriers. These barriers may be of sev- ally ends when the BMS opens fuel SSVs.
eral kinds. One is rate and range limiters to different The next and usually the most time-consuming and,
outputs and computed values. Another is to utilize step sometimes, less formalized step is establishing burner
change limiters screening new inputs introduced via operation at light-off. It may require multiple adjust-
HMI and rejecting those that have a high likelihood of ments to the valves and dampers at and near light-off
human error. In addition, some critical inputs through positions and repeats of the burner light-off sequence
HMI may be designed as requiring two steps: first, input until reasonably stable combustion after shutting down
itself and second, its verification. the pilot is established. The initial adjustments can be
Adding all these steps in the logic increases its com- completed with the controls of the valves and dampers
plexity. In most cases, it is much easier to program and in automatic mode by making changes to the lookup
troubleshoot the logic loaded on to a dedicated PLC tables at or near the light-off load.
by the same company that developed the combustion With a stable flame at light-off established, the regime
controls, rather than to pass all the logic diagrams at light-off needs to be tuned to the desired load, excess
(SAMA) on to a “third party” to do the programming air, and FGR rates. The tuning may be done with all the
and commissioning at the site using a DCS or another actuators in manual mode and then setting the points
type of PLC. in lookup tables based on the manual inputs to the
actuators. Alternatively, the adjustments can be made
only with the burner load in manual mode and all the
additional functions of the controls in automatic mode.
With the alternative approach, the changes are made
12.10 Setting Combustion Controls: directly to the lookup tables at the light-off point. The
recommendation is to tune with everything but the
Commissioning of Combustion Systems
load in automatic mode.
In this section is described a typical process of setting The next steps will be a gradual increase of the load.
burner combustion controls for an application that Again, it can be accomplished with all the actuators in
demands considerable accuracy for the operation due to manual mode until the next point in the load has been
the performance requirements. The attention here pri- established, and then recording this point in the lookup
marily goes beyond setting parameters of different PID tables. Or, it can be made with the controls in automatic
controllers that are described in many books specializ- mode. Both ways have their advantages. With the con-
ing more generally in the controls.1,2,4 trols in manual mode, it is easy to make a selection of
Setting industrial combustion controls in general and the argument for the next point depending on the mag-
especially low and ultralow combustion controls is usu- nitude of changes to the valves’ and dampers’ positions.
ally a very tedious and sometime complicated process. With everything in automatic mode, the changes are
It typically involves the following steps. made via changes in the lookup tables for the next point
The first step would be going through cold runs, with a higher load. Sometimes when increasing the load
checking functioning of the dampers and valves mak- between two points, the required combustion perfor-
ing sure there is no excessive deadband or hysteresis mance cannot be reached with any fixed setting for the
in positioning, checking responses from all the sensors next point. This indicates the need to reduce the step
and making necessary adjustments, calibration, trouble- in the lookup tables between the two loads because the
shooting, and checking that all the utilities are supplied original spacing between the points is too far apart for a
with adequate pressure. This step also involves prelimi- good approximation of nonlinear functions with linear
nary setting or checking default settings of all lookup segments. With the usually limited number of points,
tables (functions approximated by stitching together adding an intermittent point would require losing one
linear segments), especially at a few points adjacent to of the preliminary SPs. This is fine as the first approxi-
the light-off load. This step also includes checking all of mation for load points was an estimate.
the BMS functions and components including operation When the controls are parallel positioning with a
of the burner pilots and possibly operation of the flame trim, it may be important to tune the points such that
detection scanners. During this step, the system may the action of trim would be at a neutral level (the neutral
need to be equipped with additional temporary sen- level is often 50%) and all the deviations are at or close
sors and instruments like O2, NOx, CO, and combustible to zero.
analyzer(s), gauges or liquid manometers. The presence The first run up of the burner from low to high fire
of a stack O2 probe and/or CEMS (continuous emission usually results in a rough set that is far from final. So,
monitoring systems) may reduce the need for portable the next recommended step is refinement of the set by
analyzers. However, the CEMS may not be ready at the smoothing functions specified by the lookup tables.
For that, the functions need to be plotted. These func- are very seldom known) may also be a cause of trips in
tions reflect the actual characteristics of the controlling situations where the system would actually be capable
elements that are in most cases smooth functions. Any of continued operation.
initial irregularities in the functions, unless backed by
the physical nature of the process, are usually within
deviations that are still acceptable for achieving the per-
formance. Smoothing of the functions improves the set
by increasing the latitude of deviations of the controlled
12.11 Feedwater Controls
variables as well as deviation in uncontrolled variables
that is important for improving the robustness of the set. The task of the feedwater control in drum type boilers is
The next step in the setup process is verification of the to minimize fluctuations in the steam drum water level
performance when the load descends with the controls that inevitably happen during boiler load changes. An
in automatic mode and further fine tuning of the lookup excessive rise in the water level in the drum may cause
tables. a significant carryover of water into the steam line that
The following two very important steps involve test- may be very undesirable for the steam user, while the
ing stability and margins of the operation and setting drying up of the drum may be damaging to the boiler.
alarm and trip limits for the deviations monitored by the The feedwater controls usually work in parallel with
combustion controls. This involves selecting several lev- combustion controls with minimum interaction but
els of load that may be most representative for the tests, may use the same controller. It communicates with the
switching some functions of the controls to manual and BMS when the water lever deviates outside of the limits
deviating from the SP in both directions. The magnitude initiating alarms or the overall system trips.
of these deviations should not be less than the necessary A change in the steam production rate affects the vol-
margins in the controls accuracy and the effect of all the umetric amount of steam (or water) phase in the tubes
unaccounted variables, but may be limited by some signs creating potentially fast swelling effects of the water
of combustion instabilities. If instabilities are an issue, level in the tubes that carries into the steam drum when
the controls set may need to be revised by shifting the the load increases and opposite effects when the load
points away from the unstable operation. Some addi- decreases. So, substantial changes in the drum water
tional revisions to the set, to the controls or to the overall level would happen even if the feedwater input into the
mechanical system may need to be considered if these boiler would be exactly compensating for the amount
shifts still do not improve the minimum deviations to a of steam produced. The magnitude of fluctuations in
satisfactory level. At the end of the controls setting pro- the amount of water in the drum depends on the rate
cess, the alarm and trip levels for the deviations should of changes in the steam consumption, steam pressure,
not be wider than the tested margins while the trip mar- boiler design, and the quality of the feedwater control
gins may be temporarily kept higher. algorithm. It should be recognized that in many prac-
The final step will be testing and setting limits to the tical systems, even a perfect feedwater control algo-
rate of load changes. It is preferred that at maximum rithm would not be capable of maintaining a constant
change rates the system does not generate deviation drum level. For example, it is easy to see that with a fast
alarms. Review of the parameters of the PID controllers enough change in the steam output and increase in the
may be necessary at this point to make the controls faster, steam phase in the tubes, the water level can rise even
while in prior steps the parameters could have been with the feedwater flow reduced to zero. This will soon
changed to a slower response if faster setting resulted in be followed by a drop in the drum level. Another exam-
cycling. These tests will also allow checking adequacy of ple is when steam output exceeds its production in the
the feedwater controls for the process. Excessive fluctua- tubes, resulting in the drop of pressure in the drums.
tions in the drum water level may require revision of the If the drop is substantial, the steam phase in the tubes
parameters of the feedwater controls algorithm, or reduc- will expand causing an initial rise in the drum water
tion in the limits to the rate of load changes. level even when the feedwater is much less than the
It is necessary to point out that the process of set- amount of steam coming out. At the same time, the ini-
ting the controls as outlined earlier is somewhat ideal. tial response of the system will increase the feedwater
The time and cost limitations may be factors limiting flow due to the increased differential pressure across the
the actual extent of following the steps. Furthermore, feedwater valve. The opposite is true when the situation
in systems where the cost of emergency trips is reverses and the drum level drops. Slightly oversizing
extremely high, some refinery operations, for example, the feedwater supply system relative to the maximum
going through the stability tests may increase the like- design steam output helps to reduce downswings in
lihood of nuisance trips during the tests. Setting trip the drum water levels, but one should be careful not to
level deviations with safety margins (the exact limits diminish the accuracy of the oversized valve response.
Most boilers have economizers that heat feedwater and short on the order of about 10–15 s, factoring the firing
allow design of the system with higher efficiency due to rate in the controls may help to keep the controls more
reduced heat losses with the flue gas. An additional ben- steady with fluctuating steam pressure.
efit of the economizer is better control of the steam pres-
sure as fluctuations of the feedwater flow rate coming
into the drums will have much less effect on the instanta-
neous steam production. In order to avoid the possibility
of the economizer overheating, systems with economiz- 12.12 Sensors Often Used in Modern
ers always require continuous modulation of the feedwa-
Combustion Controls
ter flow and not an on-off running pump. The minimum
requirements of the feedwater flow rate actually depend 12.12.1 Pressure Transmitters
on the firing rate of the burner. The presence of an econo-
Pressure transmitters are some of the most common
mizer also reduces fluctuations of the steam pressure in
sensors in industrial combustion controls. A pressure
the drums as a much lower percentage of heat is needed
transmitter is usually used to provide an analog pres-
for the fluctuating flow of water entering the drums to
sure signal. These devices use a diaphragm coupled to a
reach saturated steam temperature.
variable resistance, which modifies the 24 V DC loop cur-
These previously described fluctuations of the level
rent from 4 to 20 mA output linear within the range the
often define the limits for how fast the load changes can
transmitter is calibrated. In recent years, these devices
be tolerated or define the size of the most expensive boiler
have become more accurate and sophisticated, with
part—the steam drum. In both cases, the boiler may not
onboard intelligence, easy calibration capabilities, digi-
have much design margin in terms of tolerance to poor
tal output capabilities, etc. The modern pressure trans-
quality of the feedwater flow (or drum water level) control.
mitters are very reliable, available in all the necessary
In a once-through steam generation system, there is no
ranges in a wide variety of configurations and materi-
steam drum. Algorithms for controlling the feedwater rate
als, and can be used in almost any service. Some of them
depend on the process. With the boiler output feeding a
can be remotely checked and recalibrated. Figure 12.16
steam turbine, for example, the firing rate and feedwater
shows a typical pressure transmitter installed on a line.
control algorithms should meet the objectives of delivering
steam with the required temperature and pressure com-
ing out of the superheater. Once-through boilers may be 12.12.2 Temperature Measurements
controlled by directly using the steam demand setting for
12.12.2.1 Thermocouples
the feedwater flow. The firing rate then will be set based on
the required steam quality (percentage of steam phase in a Thermocouples are the second most widely used sensor
two phase outgoing flow), or based on the steam tempera- in combustion applications. They are used for measur-
ture coming out of the superheater. Extensive literature ing temperatures. Whenever a wire is exposed to ther-
exists on the feedwater controls in boilers and especially mal gradient, a voltage gradient is also developed. This
in drum boilers (see Refs. [1,2] as examples). voltage is a function of the material of the wire. When
The simplest algorithms use only the drum water level two dissimilar wires come into contact on one end and
signal for defining the opening of the feedwater valve there is a temperature gradient along the wires, a volt-
called single-element or simple feedback feedwater con- age on the other end is created. This phenomenon is
trol operating as shown in Figure 12.14. The algorithm called the Seebeck effect. The magnitude of that voltage
can be improved by adding steam flow output providing varies approximately linear with temperature differ-
feed-forward action. Further improvements can be made ence between the wires’ junction and the temperature
by measuring the feedwater rate with an algorithm as at the voltmeter and is well known for specific pairs
shown in Figure 12.15. This three-element feedwater con- of metals used in thermocouples. The survivability of
trol, the most often used in the industry, has an advan- metal wires also defines the applicable temperate range
tage over two-element control as it negates the effects of and environment for which a specific type of the ther-
the feedwater valve characteristic that can vary substan- mocouples can be used. With temperatures ranging up
tially if the differential pressure across the valve varies. to 2400°F (~1300°C) rather inexpensive K type (chromel-
Because of the thermal inertia of the boiler, its steam alumel) thermocouples are usually used. Several other
production lags changes in the burner firing rate. While types are also available for lower ranges up to 1400°F
not normally done, additional benefits to the feedwater (760°C). For somewhat higher temperatures, substan-
controls may be achieved by integrating the firing rate tially more expensive platinum—platinum rhodium
input from combustion controls into the feed-forward alloy—thermocouples are available. Figure 12.17 shows
logic of feedwater control. Besides potential elimination a typical thermocouple assembly. These assemblies
of the delays caused by thermal inertia that are relatively vary in length and diameter of the shield where the
Drum level Drum level

setpoint
LT
A A A # DrmPvBadQual
To BMS
DrmSpLoLim DrmSpHiLim HMI
HMI
Δ
TC TR
Drm1PidTrk MAN
Drm1PidMan K ∫
%
Drum level
control
Drm1PidAct
manual station
A A
FdwOut
(No others)
HMI
f (x)
FV
#
Feedwater flow
control valve
FIGURE 12.14
Single-element feedwater control.
thermocouple wires are placed. Often thermocouples module mounted inside the PLC panel (usually configured
are not placed in the flow directly, but inside a thermow- for multiple thermocouple inputs), or a local device placed
ell exposed on the outside to the flow; see Figure 12.18. in the vicinity of the thermocouple outlet scaling the signal
The thermowell also allows replacement of the thermo- usually to 4–20 mA output. In the latter case, the combi-
couple on a live line without discharge of the flowing nation of a thermocouple and a secondary device may be
gas to the atmosphere and prevents the sensing element called a temperature transmitter. If the distance between
from suffering the corrosive or erosive effects of the pro- the thermocouple junction and the secondary device is
cess being measured. However, a thermowell slows the substantial, some other wires, called compensation wires,
response of the instrument to changing temperature, are used to connect to the thermocouple wires and the
may introduce other errors when measuring tempera- secondary device. The compensation wires are usually of
tures of gases and should be used with care. much lower cost than thermocouple wires and may have
When connecting a thermocouple to a secondary device, lower impedance. These wires should correspond to the
the device should be set up for the type of thermocouple type of thermocouple wires and have to be connected with
employed. The secondary device processes the incoming the proper polarity. Not using the proper compensation
voltage and adds its own temperature determined with a wires will introduce an error proportional to the tempera-
resistance temperature detector (RTD) to generate the cor- ture difference between the junctions of thermocouple
rect output. The secondary device may be a standard input wires and the secondary device.
326
Drum level Drum level Steam flow Feedwater flow

setpoint
LT FT FT
A A A # FT
DrmPvBadQual # StmFlwBadQual # FdwPvBadQual
To BMS To BMS To BMS
HMI HMI
DrmSpLoLim DrmSpHiLim
HMI Δ HMI A
StmFlwFdFwdGain
Δ
Scale
Drm3PidTrk TC TR
MAN
K ∫ %
Drm3PidMan
% 0 100 Δ
FdwPidTrk TC TR
MAN
Scale FdwPidMan K ∫
Drm out trk val
% Feedwater
control
Σ FdwManAct
manual section
A A A A
FdwSpLoLim FwdSpHiLim
FdwOut
HMI
FdwSp HMI
f (x)
FV
#
Feedwater flow
control valve
FIGURE 12.15
Three-element feedwater control.
FIGURE 12.16
Pressure transmitter (left) and pressure gauge (right). (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
FIGURE 12.17
Thermocouple. (From Baukal, C.E. (ed.), The John Zink Combustion
Thermocouples can be wired to some standard input FIGURE 12.18

devices that are part of the PLC panel, or to a local sec- Thermowell. (From Baukal, C.E. (ed.), The John Zink Combustion
ondary device converting the thermocouple voltage to Handbook, CRC Press, Boca Raton, FL, 2001.)
temperature based on the thermocouple type.
these reasons, temperature transmitters are much less

12.12.2.2 Resistance Temperature Detectors
frequently used than thermocouples.
Alternative to thermocouples are RTDs that work on a
principle that different materials vary their electrical
12.12.2.3 High Velocity Thermocouples
conductivity differently with temperature. When con-
figured to generate a scaled signal (typically 4–20 mA Also known as suction pyrometers, the design of veloc-
range), the RTD also forms a temperature transmitter. ity thermocouples attempts to minimize the inaccura-
In some cases, these transmitters may be more accu- cies in gas temperature measurements caused by radiant
rate than thermocouples, but have a lower temperature heat exchange between the thermocouple joint and wires
range and are substantially more expensive. Due to and remote objects at temperature different than the
Water inlet
Thermocouple
Outlet for gas

analysis (if needed)
Water discharge
Compressor air or
steam for eductor
FIGURE 12.19
High-velocity thermocouple.
surrounding gas. A simple shielding of the thermocouple 12.12.3.1 Orifice- and Nozzle-Based Flowmeters
from “viewing” the objects may not necessarily help as
Historically most flows in combustion applications
the shield temperature may not be the same as the gas.
are measured with the help of orifices. This method is
Minimizing the size of thermocouple wires and junction
very popular due to its simplicity. Placing the orifice
helps to reduce the errors, but makes the thermocouple
in the fluid flow causes a pressure drop across the ori-
more susceptible to damage and deterioration over time.
fice—orifice ΔP—that is measured between standardly
A practical solution is to combine a small cylindrical
located pressure taps. In order for the flow measure-
shield around the thermocouple junction (spaced from it
ments to be accurate, the pipe or duct upstream and
and wired) and force surrounding gas to pass with high
downstream of the orifice should not have any turns
velocity around the junction and through the shield using
over some distance specified as the number of equiva-
a pump or ejector as shown schematically in Figure 12.19.
lent duct diameters. The minimum distance depends
For high temperature measurements, the design should
on the required accuracy of the measurements, ratio β
be such that the compressed air for the ejector cools the
of the orifice to the line diameters and extent of dis-
thermocouple. If the concept as shown in Figure 12.19 is
turbances to the flow in the adjacent parts of the line.
used, the exhaust will be discharged such that it has no
Typically a total length of straight run between 10
effect on the flow passing through the shield.
and 25 line diameters is needed for the metering ori-
A high velocity thermocouple is often integrated with
fice placement with about 30% of this distance being
the gas sampling. In this case, the sample is pulled
upstream of the orifice plate. The requirements for the
through the shield and the thermocouple with a pump
length of the straight run increase with an increase
and the whole supporting structure is water-cooled. An
in β. When measurement accuracy is important, select-
alternative concept with an external small ejector pump
ing β > 0.6 without a thorough evaluation of differ-
may be used to eliminate discharge from the ejector into
ent sources of errors is not recommended. This space
the system. Using high velocity thermocouples provides
requirement is the main drawback of metering orifices
a high degree of precision in gas temperature measure-
in combustion applications. Another drawback is that
ment. High velocity thermocouples are very seldom
parasitic pressure losses are a major part of the mea-
integrated into the combustion controls due to the com-
sured ΔP.
plexity, maintenance costs, and limited need.
In SI units, the base fundamental equation for com-
puting the flows based on the orifice ΔP measurements
12.12.3 Flowmeters is usually written in the following form:
There are many different types of flowmeters and many
reasons to use one or another for a given application. The
following is a list of some of the more common types of
flowmeters, how they work, and where they are used.
Fm = Y × C × (2 × ∆P × ρ ) × A
f (12.6)
where TABLE 12.1

Fm is the mass flow rate Difference in C with Flange Taps between a 20 in. (50 cm)
Y is the net expansion factor used in case of compress- Diameter Pipe and 2 in. (5 cm) Diameter Pipe (Independent
ible flows of Re)
C is the orifice flow coefficient that is a function of Difference in C between a 20 in. Dia Pipe and 2 in. Dia Pipe
the ratio β between the orifice and the pipe or duct β (Independent upon Re)
diameter and the Re number
0.2 −5.5E−05
ΔP is the pressure drop across the orifice
0.25 −1.4E−04
ρf is the specific gravity of the flow upstream of the
0.3 −2.8E−04
orifice
0.35 −5.3E−04
A is the orifice cross-sectional area
0.4 −9.2E−04
0.45 −1.5E−03
If the orifice measures gas flow and is installed at a
0.5 −2.4E−03
location with variable gas temperature and/or pressure,
0.55 −3.6E−03
the relevant temperature and pressure should be mea-
0.6 −5.5E−03
sured and properly accounted when computing specific
0.65 −8.3E−03
gravity of the flow in flow calculations. For noncom-
0.7 −1.3E−02
pressible flows (most liquids), Y = 1. Deviation of Y from 0.75 −1.9E−02
1 becomes significant for gases only when the pres-
sure drop across the orifice becomes significant relative
to the flow absolute pressure. Graphs in Ref. [7] allow magnitude of a small difference in flow coefficients
determining of Y for different cases. between a large 20 in. (50 cm) pipe and a small 2 in.
When measuring flows through pipes, the meter- (5 cm) pipe. The graphs and the table were computed
ing orifice is often mounted between standard orifice based on the tables provided in Ref. [8]. The aforemen-
flanges with integrated passages for connections to tioned information along with the reference literature
a pressure transmitter. These flange taps are located is sufficient in configuring an orifice-based flowmeter
differently than pressure taps at 1 pipe diameter for the application.
upstream and 0.5 pipe diameters downstream of the Orifice meters are very accurate and highly reliable
orifice as described in Ref. [7] and generate slightly but have a narrow range due to the quadratic power
smaller ΔP but are more convenient to use. Figure relation between ΔP and flow and have the drawbacks
12.20 shows orifice flow coefficient C for large pipe mentioned earlier. They are widely used for measur-
flange taps for different β and Re. Table 12.1 shows the ing liquid, steam, or gas flows. The requirements for
0.790
b = 0.3 b = 0.4 b = 0.5 b = 0.6 b = 0.7
0.770
0.750
0.730
0.710
0.690
C
0.670
0.650
0.630
0.610
0.590
0.570
0.550
1.E+04 1.E+05 1.E+06
Re-Reynolds number based on the pipe dia
FIGURE 12.20
Flow coefficient C for square edge orifices with flange taps for larger pipes per ASME MFC #M1989 (β-ratio of the orifice dia. to the pipe
i nternal diameter).
the length of straight runs can be reduced with flow 12.12.3.4 Turbine Flowmeters
conditioning devices. Industry also offers orifice flow-
The main part of a turbine flowmeter is a bladed rotor
meters integral with temperature and pressure mea-
spun by the passing flow of fluid. The rotor is usually
surements and local processing of the three signals
designed with a minimum resistance to the flow both
to one signal proportional to the mass flow rate. The
due to the helically shaped blades and due to the way of
price of these devices, however, is considerably higher.
suspending the rotor. With this, the speed of rotor rota-
A venturi nozzle-based flowmeter works similar
tion is closely proportional to the flow with one revolu-
to the orifice flowmeter, but uses different locations
tion representing passage of the volume corresponding
for the pressure taps and different charts for the flow
to the length of one revolution of the screw. In reality
coefficient. The theory of using nozzles is explained
some frictional and rotational drug forces cause devia-
in Refs. [7,9]. The advantage of the venturi nozzle is
tion from exact proportionality, limiting the range of
lower parasitic pressure losses, but it comes with a
accurate measurements. Typically a magnetic sensor is
substantial increase in cost and much lower availabil-
used to measure the rotor speed.
ity. For these reasons, the nozzles are not used often
except in cases when the nozzle can be custom speci-
fied and built. 12.12.3.5 Ultrasonic Flowmeters
When sound travels through a moving medium (fluid),
12.12.3.2 Vortex Shedding Flowmeter the speed of the medium affects the time of sound prop-
A vortex shedding works on the principle of measur- agation between the points spaced diagonally across
ing the frequency of instabilities generated in the wake the duct. By measuring propagation times and the dif-
of a turbulent flow around a bluff body (vortex shed- ference in propagation times of ultrasonic impulses in
der), placed in the path of the fluid. As the fluid passes upstream and downstream directions the flow velocity
by, vortices form and break off with a frequency closely can be computed.
proportional to the velocity around the bluff body with An alternative concept is based on the Doppler effect
a proportionality factor being a weak function of the and can be used when the flow contains sonically reflec-
characteristic to the bluff body Reynolds number. For tive materials, such as solid particles or bubbles, moving
a reliable and strong signal not sensitive to possible along with the flow without slippage. In this case, the
uncontrolled disturbances in the flow, the shedder is flow velocity is derived from the difference in the fre-
usually designed to block somewhere between 30% and quencies of sound waves reflected from the flow mov-
35% of the flow passage, thus making it a lower pressure ing away and toward the receiver and the velocity of
loss device than an orifice plate. As the vortex shedding sound through the medium. These meters are reason-
flowmeter measures the velocity, knowing the specific ably accurate, have a fairly wide range, are reasonably
gravity of the flow is also necessary. Placement of vortex priced, and are highly reliable.
shedding flowmeters with respect to straight distances
upstream and downstream is similar to the metering
12.12.4 Emissions Analyzers
orifices with a high β. The advantage of vortex shedding
flowmeters over orifices is a wider range of flows close In combustion applications, often there is a need to mea-
to 10–1 with consistent accuracy. They are reasonably sure concentrations of O2, CO, combustibles other than
priced, highly reliable, and useful in liquid, steam, or CO, NOx emissions (NO and NO2), and sulfuric gases
gas service. (SO2, SO3). Different models of gas analyzers used for
these purposes can be classified as in situ analyzers car-
rying out the analysis at the probe, and extractive ana-
12.12.3.3 Coriolis Flowmeter
lyzers that remove a sample from the process and then
In these flowmeters, a dense fluid, usually a liquid, runs condition it by cooling and removing moisture before
through a U-shaped tube or a pair of tubes that are the analysis. In situ analyzers, primarily O2 analyzers,
being vibrated by an attached transducer around the are used in combustion controls. Extractive analyzers
fixed inlet and outlet points in a plane perpendicular to are mostly used in CEMS and during the combustion
the plane of the U. The flow of the fluid causes the tube system setting and testing.
to twist because of the Coriolis force. The magnitude of
the twisting momentum and deflection of U from the
12.12.4.1 In Situ Oxygen Analyzers
plane shape caused by it is proportional to the mass
flow. These meters are highly accurate and have a wide In situ oxygen analyzers—oxygen probes—are often used
range. They are generally more expensive and their reli- in combustion controls systems primarily for measuring
ability is not as good as some other types. oxygen concentration in the combustion products. The
most common in situ analyzers operate on the principle

of electromotive force generated across a zirconium oxide
tube, a solid electrolyte, exposed on one side to gas in
which the partial pressure of oxygen (P1) is measured and
to a referenced gas on the other side with a known partial
pressure of oxygen (P2). The potential difference E gener-
ated by the cell is defined by Nernst equation and is pro-
portional to the logarithm of oxygen pressure ratio on the
different sides of the cell and to its absolute temperature:
P
E ~ T × ln  1  (12.7)
 P2 

For proper operation, the cell needs to be heated above

~850°C (~1500°F).
As seen from the Nernst equation, the accuracy of the
probe in absolute numbers increases substantially with
a reduction in oxygen content. The accuracy is lower
when the deviations between the oxygen content in the
gas in question and the reference gas are small. In some
cases, the effects of ambient humidity as well as varia- FIGURE 12.21
tions in pressure between the process and reference View of a portable TESTO analyzer.
sides need to be considered along with using other than
air reference gases. portable instruments are very reliable and can reason-
These in situ probes are usually designed with a ably withstand the industrial environment. Yet care
heated sensing element protected from the flow by a should be taken in protecting the instrument from
ceramic diffusion assembly and a flame arrestor that weather elements, direct exposure to sun, and over
cause additional dead time for the probe response to saturation of its measuring cells with excessive emis-
changing O2 concentration. The probes are sensitive sions. The instruments require frequent replacement
to the presence of combustible gases like CO, H2, and of moisture removing filters and need to be periodi-
hydrocarbons reacting with oxygen at or near the cell cally calibrated with calibration gases. When applying
surface. Some other gases and materials containing the instruments, it is important not to overpressure
halogens (chlorine), sulfur, and lead may permanently the sample line as it may reduce accuracy and even
damage-poison the cell. When properly applied, zir- cause some damage.
conia oxide oxygen probes work reliably, but require
periodic calibration with the frequency depending on
12.12.4.3 Emissions Analyzers, CEMS
the needed accuracy of the response and specifics of the
application. The probes are rather expensive, especially A variety of more accurate stationary analyzers is used
when the options of pressure equalization and dry where accurate emissions measurements—primarily
instrument supply are added. when NOx —is needed. The most accurate analytical
instruments for NOx measurements operate by mea-
suring the chemiluminescence of NO. These analyzers
12.12.4.2 Portable Emission Analyzers
typically have an option of passing the sample through
Portable instruments, like the one shown in Figure a catalytic converter transforming NO2 to NO prior to
12.21, are standard tools used during the burner passing it through the metering cell. This option allows
commissioning process and checking of the equip- it to measure both the NO and NO2 together defined
ment performance. The cells for measuring NO, NO2, as the NOx. When properly periodically calibrated, the
CO, combustibles, and SO2 in portable analyzers instruments are very accurate and report emissions
are usually of the electrochemical type. The instru- with the accuracy close to 1%. The instruments usually
ments require periodic calibration with sample gases. operate in an air conditioned environment.
Their accuracy depends on the frequency of calibra- Chemiluminescent NOx analyzers are widely used
tion—once a month is usually acceptable for getting in CEMS that are designed to undergo frequent peri-
accuracy within 5% of the measured values and bet- odic self-calibration cycles using bottles of compressed
ter accuracy with more frequent calibration. Modern calibration gases. The CEMS are also equipped with
continuously operating coolers for moisture removal system. So the use of CEMS is mostly limited for infor-
from the sample gas. The CEMS are expensive both in mation and reporting purposes.
the initial costs as well as in maintenance.
It may be very tempting to use CEMS outputs to con-
12.12.5 Wobbe Index and Combustion Air
trol the combustion process (i.e., using NOx readings
Requirement Index Analyzers
for trimming the FGR). However, this is normally not
done as the incorporation of CEMS into the combustion WI analyzers are calorimeter-based instruments mea-
controls would add a large amount of rather sensitive suring fuel calorific value, and specific gravity for gases
instrumentation and may reduce the reliability of the also provide outputs of WI defined as a ratio of the fuel
25
C4H10
20
Approximated CARI
C3H6
15 CH4
C2H4
10
H2
CO
0
0 2 4 6 8 10 12 14 16 18
Wobbe index
FIGURE 12.22
Error in CARI approximation based on wet O2 measurements optimized for CH4 and CH4 blend with N2 Error = (approximated CARI)/(actual
CARI) * 100% – 100%.
1.3
1.2
CO
Ratio of Wobbe index to approximated CARI
1.1
Air orifice equiv. area
—————————————— = 50
Fuel orifice equiv. area
1
0.9
CH4 + CO H2 C2H4
0.8 C3H6
CH4 + CO2
0.7
CH4 + N2 CH4 C2H6 C3H8 C4H10
0.6
0.5
0.4
0 2 4 6 8 10 12 14 16 18
Wobbe index, MJ/(kg/Nm3)0.5
FIGURE 12.23
Correlation of WI (defined via LHV) and approximated CARI (approximation based on wet O2 measurements) for hydrocarbon fuels, CO, H2,
and blends of CH4 with N2, CO2, and CO in 50/50 (vol.) ratio.
calorific value to the square root of specific gravity of CEMS continuous emissions monitoring system
gas under standard or normal temperature and pres- CO carbon monoxide
sure. The response of the calorimeters is somewhat CV controller output
slow, over ~20 s, so their use for the combustion controls DCS distributed control system
for fast-changing processes may have some limitations. FC flow controller
The instruments of this type are reliable, but relatively FCV fuel control valve
expensive. FGR flue gas recirculation
Another type of instrument provides measurements HHV higher heating value
of CARI defined earlier in this chapter. Knowing the HMI human-machine interface
CARI is usually more important than WI for the pur- I/O input/output
pose of designing accurate combustion controls, but the IVD inlet vane damper
use of CARI measuring instruments should be properly MACT maximum achievable control technology
justified. CARI measurement instruments operate on NOx nitrogen oxides (e.g., NO, NO2)
the principle of catalytic oxidation of fuel and measure- PC personal computer
ments of wet O2 in the exhaust. The methodology has PID proportional, integral, differential
a certain degree of theoretical error (not counting the PLC programmable logic controller
loss of accuracy due to instrument sensitivity, calibra- PV process variable
tion errors, etc.) caused when converting wet O2 mea- RTD resistance temperature detector
surements to CARI. Figure 12.22 shows the magnitude SAMA Scientific Apparatus Makers Association
of this error computed for different hydrocarbons such SP set point
as CO, H2, and blends of CH4 with N2, CO2, and CO. SSP single point positioning
As can be seen from this chart, the correlation holds SSV safety shutoff valve
within 1.5% error for hydrocarbons, but breaks down VFD variable frequency drive
in cases of high CO content. The error is within ±5% for WI Wobbe index
fuels containing up to ~50% of CO2 and including cases
of pure H2. In cases of fuels with high CO content, the
methodology results in substantial errors that may be
as high as 21%. In most practical cases, the magnitude
of possible errors is much smaller as the variability in References
fuel composition does not approach extreme cases and 1. G. Dukelow. The Control of Boilers, 2nd edn., Instrument
can be narrowed. Society of America, Pittsburgh, PA, 1991.
These instruments also lose accuracy when dealing 2. C. E. Baukal (ed.). The John Zink Combustion Handbook,
with fuels highly diluted with inert gases. This is due to John Zink Co LLC, CRC Press, Boca Raton, FL, 2001.
reduction in O2 after oxidation of the fuel. The response 3. NFPA 85. Boiler and Combustion Systems Hazard Code, 2011
rate of the device can be down to a few seconds—much edn., NFPA, 1 Batterymarch Park, Quincy, MA, 2011.
faster than from a calorimeter. There is not enough 4. F. G. Shinskey. Process Control Systems, Application, Design,
information yet to assess reliability of available CARI and Tuning, 3rd edn., McGraw-Hill, New York, 1988.
5. V. Lifshits and S. Drennan, Development of an Ultra Low
measurement instruments. The instruments at this
NOx Burner with widened Stability Limits, American
point are still rather expensive. Flame Research Committee, Maui, Hawaii, 1998.
Measurements of CARI also allow computing of WI 6. S. Londerville. Sources and solutions of burner related rum-
using correlations between the approximated CARI and bling problems in boilers, ASME FACT-Vol 10, Combustion
WI. However, as seen from Figure 12.23, this correlation Modeling and Burner Replacement Strategies, Nayak, R.V. and
is more approximate. Again the most deviating case is Mousa, N.A., eds., Book G00523, pp. 39–45, 1990.
for gases with high CO. For HC fuels, the error is within 7. Flow of Fluids through valves, fittings, and pipe,
±6% of the average, but increases to ~12% if tuned to Technical Paper No. 410, Crane Company, Signal Hill,
eliminate error in CH4. CA, 1988.
8. Fisher Controls Company. Control Valve Handbook, 2nd
edn., Fisher Controls, Marshalltown, IA, 1977.
9. R. W. Miller. Flow Measurement Engineering Handbook,
McGraw-Hill Book Company, New York, 1983.
10. R. Jorgensen. Fan Engineering, an Engineer’s Handbook on
Nomenclature Fans Their Applications, Buffalo Forge Company, Buffalo,
NY, 1983.
BACT best available control technology 11. P. S. Dickey and H. L. Coplen. A study of damper charac-
BMS burner management system teristics, Transactions of the ASME Spring Meeting, Atlanta,
CARI combustion air requirement index GA, March 31–April 3, 1941.
13
Blowers for Combustion Systems
John Bellovich and Jim Warren
CONTENTS
13.1 Introduction.................................................................................................................................................................... 335
13.2 Applications.................................................................................................................................................................... 335
13.3 Types of Blowers for Combustion Systems................................................................................................................ 336
13.4 Fan Arrangements......................................................................................................................................................... 336
13.5 Design Considerations.................................................................................................................................................. 337
13.5.1 Fan Control......................................................................................................................................................... 346
13.5.2 Materials of Construction................................................................................................................................. 347
13.5.3 Motors and Drives............................................................................................................................................. 348
13.5.4 Couplings and Belts........................................................................................................................................... 348
13.5.5 Bearings and Lubrication................................................................................................................................. 349
13.5.6 Vibration and Installation................................................................................................................................. 350
13.5.7 Shaft Seals........................................................................................................................................................... 350
13.5.8 Noise Considerations........................................................................................................................................ 352
13.5.9 Filtration.............................................................................................................................................................. 352
13.6 Operational Costs.......................................................................................................................................................... 352
13.7 Inspection and Testing.................................................................................................................................................. 353
13.8 Maintenance and Troubleshooting............................................................................................................................. 355
References................................................................................................................................................................................. 357
to introduce the basic concepts of blowers for combus-

tion services and give awareness to the critical issues.
13.1 Introduction
The intricacies of blower design are not discussed.
Compared to industrial air handling or air condi- To gain more knowledge on the topic, consult the
tioning blowers, blowers and fans used in the service Fan Engineering Handbook by the Howden/Buffalo fan
of combustion systems have some unique require- company.1 Other valuable resources include any of the
ments that make design, installation, and opera- Air Movement and Control Association International
tion a little more difficult. An incorrect fan design or (AMCA) standards.2,3 Some good general references
material selection can have a disastrous effect on the related to this topic are also available.4–8
performance of an entire facility. It is very important to
understand the service of the fan as well as its perfor-
mance, in order to properly run a combustion system.
The terms “blower” and “fan” are used interchange-
13.2 Applications
ably in the context of this book. Blowers and pumps
are very similar in the fact that they both push a fluid Typical applications for fans in combustion systems
from one point to another by putting more energy into include the following:
the system. Pumps push liquids and fans push gases.
Fans for combustion systems can be used to introduce • Forced and induced draft fans for fired heaters,
air into a combustion zone, prevent smoke formation, burners, boilers, and thermal oxidizers
move exhaust gases, cool or reoxygenate a stream, • Introduction of air for smoke suppression of
purge a stack, and so on. The purpose of this section is flare systems
335
• Cooling or purging of flare systems driven, the motor is usually inside the fan housing, and
• Movement of waste gases into the burner or the unit would have to be disassembled for work on the
flare system motor. Some vane axial fans have an extended shaft to
move the motor out of the housing, but these are gener-
• Reoxygenation and quenching of thermal oxi-
ally for large ventilating applications.
dizer flue gases
Backward-curved and air-foil fans usually offer the
highest efficiency. Radial and radial-tipped fans are
usually used in applications that have particulates (see
Table 13.1). See Figure 13.4 for an example of a blower
used for purging. See Figure 13.5 for an example of a
centrifugal blower used for a landfill flare application.
13.3 Types of Blowers for
Combustion Systems
There are primarily two types of blowers used in the com-
bustion industry, centrifugal and axial. Centrifugal fans 13.4 Fan Arrangements
(see Figure 13.1), also known as radial fans, are very simi- Arrangements are the AMCA’s designations for the
lar to centrifugal pumps. The air enters the center of the basic configurations of fans.3 There are two consid-
impeller and is ejected radially outward and “pushed” erations that need to be made before finalizing what
by the impeller blades through the housing and out of the arrangement is required, installation and maintenance.
exit. Centrifugal fans usually have the same basic con- Arrangements 1 through 3 (as seen in Figure 13.6) require
figuration with options concerning what type of impel- a separate pedestal for the fan housing and the driver.
ler and drive system is used (see Figure 13.2 for different This reduces the up-front equipment cost, but because a
impeller designs). Other variations of the centrifugal fan separate foundation needs to be poured for the driver,
will include high pressure fans and multistage turbo the installation cost is higher. After installation, align-
blowers. Axial fans are more like an airplane propeller ment will be necessary for direct drive units using these
that pushes the air along in the direction of the axis of arrangements. These arrangements are less often used
rotation. Axial fans fall into two categories, vane axial in the combustion industry and typically used for large
and tube axial. Vane axial fans (see Figure 13.3) have a set motor sizes of above 300 hp (224 kW). Arrangements 4, 7,
of internal guide vanes, while tube axial fans have none. and 8 offer the convenience of mounting the driver on a
Which type of fan is used will depend on the application. common frame with the fan for a higher equipment cost.
Centrifugal fans typically can generate higher pressures These arrangements are more common. Arrangements
than vane axial fans, and the motor can be maintained 9 and 10 are belt-driven configurations where alignment
outside of the fan. Vane axial fans are high volume low and installation are simply a matter of preference.
pressure fans, are usually lower in cost, and are typically The second factor to consider with the different arrange-
a little more efficient. Unless the vane axial fan is belt ments is the location of the bearings. The locations of the
bearings will determine the extent of maintenance to
be done on the machine. From a maintenance point of
view, arrangements 4 (seen in Figure 13.7) and 5 fans are
the simplest. In arrangements 4 and 5, the fan impeller
is mounted directly to the motor shaft, and there are no
bearings other than the motor bearings. Arrangement
5 is similar to arrangement 4 except that the motor is
mounted directly to the fan housing. This may be per-
fectly acceptable for smaller fans. Arrangements 1, 2,
8, 9, and 10 are what is known as overhung, that is, the
wheel is cantilevered on the shaft and the bearings are
both on the same side. There is an inboard and outboard
bearing. Because both bearings are on the same side of
the wheel, the bearings are subjected to uneven wear.
Arrangements 3 and 7 have the fan wheel in between the
bearings, which results in more even bearing wear and
FIGURE 13.1 longer life. This arrangement is the best in terms of main-
Centrifugal fan. tenance of bearings, but has a higher associated cost.
Blowers for Combustion Systems 337
Paddlewheel Backward inclined

Open design/no shroud Static efficiency to 80%
60%–65% static efficiency Low to medium tip speed
Inexpensive design capabilities
Good for high temperature or
highly erosive applications
Medium to high pressure
Radial blade Backward curved

Static efficiency to 75% Medium to high tip speed
High tip speed capabilities capabilities
Reasonable running clearance High efficiency to 83%
Best for erosive or sticky Clean or dirty airstreams
particulate Solid one-piece blade design
Radial tip Airfoil

Static efficiency to 75% Static efficiency to 87%
Medium to high tip speed Medium to high tip speed
capabilities capabilities
Running clearance tighter than Relatively tight running
radial blade but not as critical as clearances
backward inclined and airfoil
Good for high particulate airstream
Axial flow
Forward curved (sirrocco) High volume, low pressure
35%–50% static efficiency
Smallest diameter wheel for a
High temperature furnace
given pressure requirement
recirc. applications
High volume capability Reversing flow capability
55%–65% static efficiency Airflow parallel to shaft axis
Often used for high
temperatures
FIGURE 13.2
Fan wheel designs. (Adapted from Robinson Fans, Zelienople, PA.)
13.5 Design Considerations

When selecting a fan for a combustion application, the
design of the entire system needs to be taken into account.
Location of the fan in the system is very important in
determining what kind of fan will be required. If the fan
is located at the beginning of the system and “pushing”
the air or gas through, then it is considered a forced draft
system. If the fan is located at the back end of the system
and is “pulling” the flue gas through, then it is considered
an induced draft system. A system with both a forced draft
fan and an induced draft fan is called a balanced draft
system. Once it is determined where the fan is located in
the system, then all the process conditions that can affect
the fan selection need to be determined. Final selection is
based on flow rate, composition of the gas, range of oper-
ating temperatures, elevation, and inlet/outlet pressure
required. Because the area around a flare system is a harsh
FIGURE 13.3 environment with a sterile zone that limits access to per-
Vane axial fan. sonnel, vane axial fans are preferred for flares.
TABLE 13.1
Relative Characteristics of Centrifugal Blowers
Forward-Curved
Radial Radial Tip Backward-Curved
Efficiency Medium Medium High
Tip speed High Medium Medium
Sizea Small Medium Large
Initial costb Small Medium Large
HP curve Medium rise Medium rise Power limiting
Accept corrosion coating Excellent Fair to poor Good (thin coat)
Abrasion resistance Good Medium Medium
Sticky material handling Good Poor Medium
High temperature capability Excellent Good Good
Running clearance Liberal Medium Minimum required
Operation without diffuser Not as Must use Good efficiency
efficient
Noise level High Medium Low
Stability/non-surge rangec Medium Poor Medium
20%–100% 40%–100% 20%–100%
a Size is based on fans at the same speed, volume, and pressure.
b Cost is based on fans at the same speed, volume, and pressure.
c More a function of operating point along a curve than fan type.
FIGURE 13.4
Purge air blower on the side of a combustion chamber.
FIGURE 13.5
Temperature and elevation can greatly affect the per- Multistage high speed centrifugal blower for a landfill application.
formance of a fan. The fan itself should be sized using
the maximum flow rate at the highest temperature. and not standard or normal volumes. Row two, or actual
Table 13.2 shows the effect that temperature and eleva- cubic feet per minute (ACFM), shows that the volume of
tion (pressure) have on a volume of air. The first row air, or gas, is proportional to the ratio of the absolute
shows the flow rate in standard cubic feet per minute temperatures. As the temperature increases, so does
(SCFM). This number is always the same because it is the volume and vice versa. Also note that the volume
corrected back to standard temperature and elevation, increases with respect to the elevation, that is, the atmo-
in this case 68°F (20°C) at sea level. However, because sphere does not push down on a volume of gas as much
of the effects of temperature and elevation on volume, at 3000 ft (900 m) above the sea level as it does at the sea
blower vendors tend to want to work in actual volumes level. So, in the second row, the volume of gas increases
Notes:
SW—Single width DW—Double width
SI—Single inlet DI—Double inlet
Arrangements 1, 3, 7, and 8 are also available with
bearings mounted on pedestals or base set
independent of the fan housing.
ARR. 1 SWSI—For belt drive or
direct connection. Impeller
overhung. Two bearings on base.
ARR. 2 SWSI–For belt drive or ARR. 3 SWSI—For belt drive or ARR. 3 DWDI—For belt drive or
direct connection. Impeller direct connection. One bearing direct connection. One bearing
overhung. Bearing in bracket on each side and supported by on each side and supported by
supported by fan housing. fan housing. fan housing.
ARR. 4 SWSI—For direct drive. ARR. 7 SWSI—For belt drive or ARR. 7 DWDI—For belt drive or
Impeller overhung on prime mover direct connection. Arrangement 3 direct connection. Arrangement 3
shaft. No bearing on fan. Prime plus base for prime mover. plus base for prime mover.
mover base mounted or integrally
directly connected.
ARR. 8 SWSI—For belt drive or direct ARR. 9 SWSI—For belt drive. ARR. 10 SWSI—For belt drive.
connection. Arrangement 1 plus Impeller overhung, two bearings, Impeller overhung, two bearings,
extended base for prime mover. with prime mover outside base. with prime mover inside base.
FIGURE 13.6
Fan drive arrangements for centrifugal fans AMCA standard 99-2404-03. (Adapted from Robinson Fans, Zelienople, PA.)
to 11,170 ACFM (316 ACMM = actual cubic meters per temperature. This can be seen in the last row that shows
minute) at the higher elevation at standard temperature approximate horse power.
and 11,847 ACFM (336 ACMM) at the same elevation
and higher temperature. Even though the flow rates
are the same on a standard basis, more horsepower is (Discharge pressure
required to push the increased volume corrected to the inches of water column)
higher temperature and, conversely, less horsepower HP = (ACFM) × (13.1)
(6354 × efficiency)
is required to push the decreased volume at the lower
the motor during start-up, causing a breaker trip and

shutting the system down.
The final piece of information that is needed to size a
blower is the discharge pressure to overcome the system
back pressure. Typically, blower manufacturers tend to
use static pressure, that is, the portion of pressure mea-
sured with a pressure gauge. The pressure drop must
be calculated for each fitting in the system. The compo-
nents of the pressure drop are as follows:
Ptotal = Pvelocity + Pstatic (13.3)
where
FIGURE 13.7
Arrangement 4 fan.
ρV 2
Pvelocity = (13.4)
2 gc
The example in this table uses discharge pressure of Ptotal = Pvelocity × K (13.5)
30 in. (760 mm) water column and a mechanical effi-
ciency of 75%. From this it is plain that as the tempera-  L
ture or elevation increases, so does the required horse K = f ×  (13.6)
 D
power. So, this shows the importance of specifying this
information correctly. A fan designed to push 10,000 for a pipe, where L is the length of the pipe, D is the
ACFM (280 ACCM) at sea level will not work adequately inside diameter of the pipe, and f is the friction factor.
at 3000 ft (900 m) for the same application. For additional K values for fittings, see Table 8.4. The
However, Table 13.2 is simply a demonstration of how friction factor is calculated using
standard units will not accurately describe the conditions to
which a fan may be subjected. Blowers are constant volume
64
machines, meaning that the volume of air or gas pushed
f = for laminar flow (13.7)
through the impeller stays the same. But the density change Re
needs to be considered in the HP equation. As the tempera-
ture changes, the density does too (see Table 13.3). 0.3164
f = for turbulent flow in reasonably
Re0.25
(Discharge pressure
smooth piping (13.8)
inches of water column)
HP = (ACFM) ×
(6354 × efficiency)
Re is the Reynolds number and is defined by
 Density 2 
 Density 1  (13.2) ρVD
Re = (13.9)
µ
The motor should be sized based on the correspond-
ing horse power at the maximum flow rate and tempera- where
ture. Additionally, the blower vendor will need to verify ρ is the density
that the motor is correctly sized at the low temperature µ is the dynamic viscosity
end so that too much electrical current is not pulled by V is the average velocity of the fluid in the pipe
TABLE 13.2
Effects of Temperature and Pressure on Volume and Horsepower
68°F at 32°F at 100°F at 68°F at 3,000 100°F at
Sea Level Sea Level Sea Level ft ASL 3,000 ft ASL
SCFM 10,000 10,000 10,000 10,000 10,000
ACFM 10,000 9,318 10,606 11,170 11,847
lb/h 44,940 48,180 42,360 40,200 37,920
HP 63 58 67 70 75
TABLE 13.3
Effects of Density on Horsepower
68°F at 100°F at
68°F at 32°F at Sea 100°F at 3,000 ft 3,000 ft
Sea Level Level Sea Level ASL ASL
ACFM 10,000 10,000 10,000 10,000 10,000
lb/h 44,940 48,180 42,360 40,200 37,920
lb/ft3 0.075 0.080 0.071 0.067 0.063
HP 63 67 60 56 53
For air at standard conditions,

Surge point
Re = 102.3 DV (Imperial) (ft, ft/m) (13.10) Design point
Fan curve
Re = 65, 970.3 DV (SI) (m, m/s) (13.11)
Static pressure
For a full description of calculating pressure drop
through ducting, see Chapter 8.
Proper selection of a blower is not accomplished by a
single straightforward formula. Experience, usage, and
System curve
careful evaluation of each application are necessary to
ensure proper fan selection. It is always best to work
with the blower manufacturers to get the most cost-
effective recommendation. When possible, calculate Volumetric flow
the most accurate pressure drop required, so that the
fan may be appropriately sized. Be sure to account for FIGURE 13.9
future expansion and system change. Basic vane axial fan curve.
With all the information discussed earlier, the blower
vendor will generate a performance curve. The perfor- charts for each model of fan that shows the performance
mance curve plots the flow rate (volumetric) versus the of a particular fan with different size impellers at differ-
discharge pressure (static). As a minimum, there are ent speeds.
two parts to the performance curve, the fan curve and The fan curve represents the performance of a par-
the system curve. See Figures 13.8 and 13.9 for repre- ticular fan. Following the curve all the way to the
sentative fan curves. The fan curve is a graphical rep- right would indicate what is called the free discharge,
resentation of the fan performance for one wheel size. meaning there is no back pressure and the flow rate
Manufacturers also have what are called multi-rating is at its maximum. Following the curve back to the
vertical axis would show the flow rate if the exit of
Surge point the fan were completely blocked off. The more static
discharge pressure required to overcome the pres-
Design point sure losses, the less flow rate the fan can push out
and, conversely, a reduced static discharge pressure
Fan curve results in a greater flow rate. The system curve shows
Static pressure
the possible combinations of discharge pressure and

volumetric flow rates for that particular application
from no flow up to the design point. The surge point,
or peak pressure point, is the location on the curve
where operation to the left may result in unstable per-
System curve formance. The pressure will fluctuate in positive and
negative pressure pulses that result in loss of flow and
greatly increased noise and vibration. Blower surge,
Volumetric flow
is similar in effect to pump cavitation, leading to lost
efficiency, vibration, and improper performance. It is
FIGURE 13.8 strongly recommended to choose operating points to
Basic centrifugal fan curve. the right of the surge point. It is also recommended
Vertical line from

design point down
to HP curve
Static pressure
HP curve
Horizontal line from
intersection of vertical
line and HP curve for
HP value on right axis
FIGURE 13.10
Basic centrifugal fan curve with horsepower.
that the design point should not be within 10% of the the BHP value, so that the motor is operating within
flow of the surge point. Some fan curves will also have its safety factor. The design fan speed should not be
a horsepower curve and the value for horsepower on greater than 85% of the maximum safe fan speed.
the right-hand axis. This value is known as the brake Figures 13.11 and 13.12 are examples of typical fan
horse power (BHP), and it can be found by following curves. See Figures 13.13 and 13.14 for more examples
a vertical line down from the design point until it of fans in the field.
intersects the BHP curve, and then reading along the One final comment about blower curves is that they
BHP axis horizontally (see Figure 13.10). The motor, can be changed by controls (dampers) or by speed.
or driver, chosen should supply more horsepower, The fan laws are a handy approximation tool that
40 800
35 700
30 600
SP
Fan static pressure (In-H2O)
25 500
BHP
20 400
15 300
BHP
10 200
System resistance
5 100
0 0
0 10 20 30 40 50 60 70
Volume (CFM × 1000)
FIGURE 13.11
Forward tip blade operating curve for 1780 RPM, 70°F, and 0.075 lb/ft3 density. (Data provided by Robinson Fans, Zelienople, PA.)
40 320
SP
35 280
30 240
Fan static pressure (In-H2O)
25 200
BHP
20 150
15 120
10 BHP 80
5 40
System resistance
0 0
0 10 20 30 40 50 60 70
Volume (CFM x 1000)
FIGURE 13.12
Backward curved blade operating curve for 1780 RPM 70oF and 0.075 lb/ft3 density. (Data provided by Robinson Fans, Zelienople, PA.)
FIGURE 13.13
One primary and one backup fan in the field with ducting.
BHP varies as a function for cube of the speed (RPM)3
3
 RPM new 
BHPnew =  × BHPold (13.14)
 RPM old 
Changing the speed of the fan actually shifts the

fan curve up or down, depending upon the change.
There will be more discussion on this topic in Section
13.5.1.
Note: Before increasing the speed of the fan, verify
the maximum safe speed with the manufacturer. Also
verify that the existing electrical system can handle the
additional load. Be aware that there are critical speeds
that must be avoided. The critical speed is a point where
the fan operation is close to the natural frequency and
increased vibration will occur.
There are many variations of the fan laws, but the
speed law referenced earlier is the most useful. A sec-
ond useful set is for changing the diameter of the impel-
ler. The size change law is as follows:
FIGURE 13.14
Six-blade vane axial fan in the field. 3
 DIA new 
CFM new =  × CFM old (13.15)
operators can use to change the fan performance if  DIA old 
required. The speed change fan laws state the following:
2
 DIA new 
SPnew =  × SPold (13.16)
• Volumetric flow varies directly as a function of
 DIA old 
speed (RPM).
• Static pressure varies as a function of the square 5
 DIA new 
of speed (RPM2). BHPnew = × BHPold (13.17)
 DIA old 
• BHP varies as a function for cube of the speed
(RPM3).
Again, all three laws must be applied at the same time.
Use this second set with a little caution. There are limits
These correlations give users the flexibility to vary the
as to how big or small a wheel diameter may be changed.
speed to change the system or avoid undesirable oper-
Results will vary with this second set.
ating points. However, all three laws must be applied
Finally, the density change fan laws can be practical
simultaneously. The speed cannot be changed with-
and easy to use. Notice that these are directly propor-
out affecting all three attributes of flow, pressure, and
tional and are given as follows:
horsepower. For example, the flow cannot be increased
without the discharge pressure being increased. If the
CFM new = CFM old (no change) (13.18)
system cannot handle the increased pressure, then
increasing the volume should be avoided.
Mathematically, the speed change fan laws can be  Density new 
stated as follows: Volumetric flow varies directly as a  Density old 
function of speed (RPM)
 Density old 
 RPM new  BHPnew =  × BHPold (13.20)
CFM new = × CFM old (13.12)  Density new 
 RPM old 
Static pressure varies as a function of the square of

speed (RPM2) Example 13.1
2
 RPM new  A chemical manufacturer has a process that
 RPM old 
now has a reduced amount of product going

to it for the rest of the equipment life. The unit
was originally designed for 50,000 CFM of air Example 13.2

at 10 in. WC static pressure with a speed of
A boiler has forced draft burners. But the installed
1800 RPM using a 115 hp motor. The new flow
ducting is a different configuration than was ini-
rate only needs to be 30,000 CFM. The owner
tially designed and takes more pressure drop than
would like to purchase a new motor or VFD to
was expected. In order to generate more static pres-
reduce utility cost and flow rate. What will the
sure, the owner would like to look at increasing the
new motor speed be?
impeller diameter. The flow rate of the gas is 50,000
To begin, use the speed change law for flow
CFM at 10 in. WC. The motor horsepower is 115
rate:
and the impeller diameter is 36 in. What size does
the new impeller need to be to get 10.5 in. WC?
 RPM new  To find a new diameter, use the size change
CFM new = CFM old × 
 RPM old  laws. Again, all three laws have to be applied at
the same time.
First, solve for the new diameter using the
 RPM new  static pressure formula:
30, 000 CFM = 50, 000 CFM × 
 1, 800 RPM  2
 DIA new 
SPnew =  × SPold
 DIA old 

RPM new = 1080 RPM
2
 DIA new 
10.5 in. WC =  × 10 in. WC
 36 in. 

Note that none of the fan laws in each set can
be applied individually. The other two laws in DIA new = 36.9 in.
the speed change must be applied. The static

pressure and horsepower are calculated as The new diameter seems feasible but the owner
follows: should always check with the manufacturer before
making such a change. The manufacturer will
need to verify if the impeller will fit appropriately
2 in the housing. The new flow rate will be as follows:
 RPM new 
SPnew =  × SPold
 RPM old 
 DIA new 
3
CFM new =  × CFM old

 DIA old 
2
 1080 RPM 
SPnew =  × 10 in. WC 3
 1800 RPM   36.9 
CFM new =  × 50, 000 CFM
 36 

SPnew = 3.6 in. WC
CFM new = 53, 845 CFM

3 and the horsepower is
 RPM new 
BHPnew = × BHPold
 RPM old 
5
 DIA new 
BHPnew =  × BHPold
 DIA old 

3
 1080 
BHPnew = × 115 hp
 1800   36.9 
5
BHPnew =  × 115 hp
 36 

BHPnew = 25 hp

BHPnew = 130 hp

From the aforementioned example, reducing The change in diameter has a great impact on the
the speed would probably be very beneficial. horsepower. A 5% change in discharge pressure
Because the pressure drop for the blower to leads to a 13% change in horsepower. Assuming
overcome was less than 10 in., it can be assumed the increase in flow rate is acceptable, the owner
that the new flow rate will simply follow the will also have to decide if he is willing to use the
system curve and the discharge pressure of 3.6 1.15 service factor for his motor. As a minimum, the
in. WC is sufficient to overcome the pipe losses. owner can run in the service factor while waiting
Obviously, 1/4 of the horsepower would be sig- for a new motor to be purchased and installed at a
nificant savings. later time.
13.5.1 Fan Control Damper closing
Most combustion applications require some sort of con-

trol of the fan or the air handling system. Control can
be accomplished in several different ways. However,
Design point
before considering what type of control devices are
required, the user will want to consider what turndown
Static pressure
is required, meaning how far below the design rate will
the system need to operate. Additionally, the user needs
to consider what the daily, or nominal, operating point
Damper opening
will be. With this information at hand, the designer can
develop a control scheme that will fit the system. Fans
are typically controlled by pressure, temperature, flow
rate, and in some cases, an optical device sensitive to
smoke formation.
The simplest and probably the most common form of Volumetric flow
control is using an inlet or an outlet damper. Although
the two devices can certainly function in the same FIGURE 13.16
Inlet damper effects on fan performance.
manner, the results are a little different. The outlet
damper will change the system curve by putting more
or less resistance, or back pressure, on the fan. Figure the same spot regardless of how open or closed the
13.15 shows how the system curve changes as the out- damper is. Effectively, the inlet vanes change the blower
let damper is opened or closed. Closing the damper performance curve so that the horsepower reduction
pushes the system resistance up and decreases the ratio is actually greater than the flow ratio change. The
available flow rate. Note that closing the damper too turndown associated with an inlet damper is around 4
much may result in undesirable operation around the to 1, but varies depending upon the type of fan.
surge point. Dampers can have single blade, opposing blade, or
Opening the damper reduces the system back pressure parallel blade configurations. Opposing blades offer the
and results in a higher flow rate. The turndown associ- best control configuration because the air flow is evenly
ated with an outlet damper alone is approximately 6 to 1. distributed across the damper. Parallel and single blade
Using an inlet damper has a completely different dampers are more cost effective and can work well if the
effect on the fan than the outlet damper. Instead of control of the air does not need to be precise, but these
changing the system curve, the inlet damper changes dampers can create an uneven air flow distribution in
the fan curve, as seen in Figure 13.16. Opening or clos- the ducting. Figure 13.17 shows a fan with both inlet and
ing the inlet damper pivots and changes the shape of the outlet dampers.
fan curve along the system curve, allowing much lower Speed control offers even greater flexibility than
turndown than an outlet damper alone. Note though, either an inlet or an outlet damper. Using a variable
that the left-hand side of the curve is still anchored at
Damper closing
Design point
Static pressure
Damper opening
Volumetric flow

Outlet damper effects on fan performance. Centrifugal fan with inlet and outlet dampers.
Speed reduced
Surge point
Design point
Design point Fan curve
Static pressure
Static pressure
Speed increased System curve
Volumetric flow
Volumetric flow FIGURE 13.19

Variable and controlled pitch change effects on fan performance.
FIGURE 13.18
Speed change effects on fan performance.
frequency drive (VFD) shifts the entire curve up or

down proportionally (see Figure 13.18). Using the VFD
offers around a 10 to 1 turndown, and the efficiency
of the fan stays the same throughout. Note, however,
that it is possible to turn the fan down below the speed
required for the internal motor fan to keep the motor
bearings cool. Consult your blower vendor before
finalizing your design. In case of using a VFD, ask the
vendor to provide performance curves for the mini-
mum rated speed, maximum speed, and three other
speeds equally spaced between the minimum and
maximum speeds.
Other control options may include two speed motors
and multiple blowers, depending upon the application.
A two speed motor simply has two sets of windings
FIGURE 13.20
within that can turn the motor at a high speed and low
Close-up of variable pitch blades on a vane axial fan.
speed (1/2 the RPM of high speed). Depending upon
the application, a two speed may be a good fit. Other
variations may be to use multiple blowers. Bringing changed during operation, but with the lower cost of
blowers on line as required can result in significant VFDs this option is becoming less common.
cost savings.
One final topic of fan control is variable pitch and
13.5.2 Materials of Construction
variable control impellers in vane axial blowers. The
angle of the impeller blades in centrifugal blowers can- The materials of construction of a fan must be suitable for
not be changed without changing the entire impel- the services of the application. For most air applications,
ler assembly. However, some vane axial blowers come carbon steel is adequate. Low temperatures may require
with what is called a variable pitch impeller, meaning stainless steel or low temp alloys. Applications where
if the operator shuts down the unit, the pitch of blades caustic or acidic gases can collect in the fan, such as an
can be changed to alter the fan curve. This has a similar ID fan on a thermal oxidizer, need to have corrosion
effect on the fan curve as an inlet damper (see Figure resistant alloys on all wetted parts. High temperature
13.19), but the pitch of the blades can only be changed applications, above 250°F (120°C), may require the use of
when there is a problem and a shutdown is required. heat flingers, which are heat transfer fins on the shaft,
Figure 13.20 shows a close-up view of these variable to keep the high temperatures from damaging the bear-
pitch blades. Another option is to purchase a controlled ings. Other applications may use a plastic or FRP mate-
pitch impeller, meaning the pitch of the blades can be rial for corrosion resistance. FRP fans should not run at
3600 RPM, but should actually run at the lowest possible motor with a service factor of 1.15 can handle loads up
speed to prevent fatigue cracking and failure. Internal to 15% more for short periods of time. The service factor
linings should, in general, be avoided as they tend to peel is available in case the user needs to operate the unit
off after years of operation. Using alloys in this case may above the original design conditions for a short amount
result in a higher up-front cost, but less in maintenance. of time. The service factor of the motor is not to be used
The cost of an unexpected shutdown is much greater to satisfy the design requirements.
than a little up-front capital cost. Motors for combustion applications should generally
be suitable for a hazardous area. The equipment sup-
plier does not have intimate knowledge of the area the
13.5.3 Motors and Drives
equipment is going into, and must sometimes make
Motors for fans and blowers can be very complex and, assumptions. It is the responsibility of the user of the
in fact, an entire book can be written on the subject. In equipment to inform the vendor of area classification
this section are discussed the highlights and pitfalls to requirements. Most applications will meet the NEC
help in selection, but not the extreme details of motors Class I, Div 2, Group C&D, T3 requirements meaning
and drivers. The most common driver for fans is the that the equipment is suitable for a hazardous area, but
electrical induction motor. The most common voltage that explosive conditions are not normally present. The
for combustion applications is around 460 V, but larger comparable IEC equivalent is Class I, Zone 2, IIB, T3.
voltages can be used to reduce the power supply wiring However, it is important to note that depending upon
size. Because of the inefficiencies of our physical world, the country in which the equipment is installed, the
induction motors do not turn at exactly their rated two may not be interchangeable, meaning where IEC
speed. For instance, a 1800 RPM motor will generally equipment is required NEC may not be an acceptable
turn at 1780 RPM. This is known as slip. Synchronous substitute and vice versa. Area classification needs to be
motors, meaning a motor that does turn at its rated verified by the local permitting agencies. Also note that
speed, can be used, but the additional cost may not be an NEC Class I, Div 2 motor may only require a weath-
justified for small motor sizes. An additional consider- erproof terminal box, while IEC motor may require an
ation is the frequency of the supply voltage. As stated, explosion-proof terminal box.
a 1800 RPM motor will turn close to the rated speed In certain applications where electricity is not available,
using a 60 Hz power supply, but the rated speed will or even when the possibility of the loss of electricity can
be around 1500 RPM using a 50 Hz power supply. This create a problem, other drivers are available. Steam tur-
may be a critical factor in developing proper pressure at bines offer a fairly efficient way to make use of a common
the end of the impeller blades. utility in most facilities. Occasionally, end users will have
Motors for centrifugal fans should have a totally an electric-driven fan and a steam-driven backup in case
enclosed fan cooled (TEFC) enclosure, meaning the of an emergency. Steam turbines tend to be more expen-
motor has a small fan attached to the shaft that blows sive and require a higher level of maintenance compared
cooling air across the motor housing. Because the motor to electric motors. Another option is an internal combus-
for most vane axial fans is mounted within the air stream, tion motor with a gear box. These options may fit some
a totally enclosed air over (TEAO) motor is acceptable. applications a little better than an electric motor.
However, a belt-driven motor mounted on the outside of
a vane axial housing would need to be TEFC.
13.5.4 Couplings and Belts
The motor should be adequately sized for the duty it
will see. As discussed earlier, the fan needs to be sized Couplings and belts are used to transfer the drive energy
for the flow rate at high temperature, but the motor from the motor to the fan impeller. Non-lubricated, flex-
needs to be adequately sized for the start-up amper- ible couplings are common today and join the motor
age at the required flow rate and low temperature. drive shaft to the fan shaft (see Figure 13.21). Other than
Start-up current drawn can be as much as seven times periodic balance checks, they require low maintenance
the full motor current load. Because of all the nuances and are easy to use. The materials must be suitable for
involved in motor selection, let the blower vendors size the application. Additionally, couplings protect the
the motors, but make sure they include all the options motor and the fan shaft by serving as the weak point in
required and that the motor horsepower and torque are the rotating system. A failed coupling is easier to replace
greater than what the fan requires. than a motor or fan shaft and wheel. Belts are not as
Another item to consider is the motor service factor. common in the combustion industry, but they offer a
The motor service factor is important because it tells lower cost alternative to couplings, and exact alignment
the user how much extra horsepower a given motor is not as important as with a coupling (see Figure 13.22).
can output. A motor with a service factor of 1.0 cannot Belts offer a low cost alternative to speed changes with-
exceed its nameplate horsepower without damage, but a out using a VFD. However, belt speed settings can only
FIGURE 13.21
Close-up of a flexible coupling.
FIGURE 13.23
Oil lubricated bearings with reservoir.
bearings are adequate. Grease bearings do not leak and

are easier to maintain. Bearings should be an L-10 grade
with approximately 100,000 h of expected run life. Sleeve
bearings are adequate for many applications up to 250°F
(120°C). Ball and roller bearings can handle temperatures
up to 550°F (290°C). Oil lubrication may be required for
higher speed and higher temperature services. The sim-
plest oil lubrication option is the oil bath, where a pool
of oil sits in the lower portion of the bearing housing
and lubricates the bearings as they pass through the oil.
Oiler systems are available that have an oil reservoir in
which operators can see the oil level and add lubrica-
FIGURE 13.22 tion when required (see Figure 13.23). Another option
Belt-driven centrifugal blower.
is what is known as an oil mist system that uses com-
pressed air and injects a small amount of oil as needed.
be changed when the fan is shut down. Belts need to be The most complicated option is an oil recirculation sys-
routinely checked for tightness and wear. Additionally, tem, which has a pump and reservoir. For critical appli-
belts tend to slip, that is, efficiency loss arises, as time cations, or high temperature applications, an oil cooling
goes on. Both belts and couplings need to have some system may be required. Additionally, heat slingers or
sort of protective guard (OSHA or equivalent) to keep shaft coolers may be used to reduce the temperature
personnel from accidentally contacting moving parts. on the bearings. Oil mist lubrication may be required
in severe service applications where the D × N value
is greater than 200,000 where D is the diameter of the
13.5.5 Bearings and Lubrication
wheel in mm and N is the speed in RPM. Heat slingers
Bearings and the bearing lubrication system are typi- are simply a set of cooling fins that dissipate heat into
cally the most critical components of turbo machinery the atmosphere by radiation and convection.
to design and maintain, and blowers are not an excep- If the fan is in critical service, meaning an emergency
tion. When possible, consider eliminating blower bear- shutdown will cause much delay and loss of revenue,
ings and using an arrangement 4 or 5 fan, meaning the then it is recommended to monitor the vibration and
impeller is attached directly to the motor shaft, for small temperatures of the bearings. Vibration is measured
applications, less than 50 hp, and almost all vane axial by an accelerometer or switch while temperature is
fans. The elimination of the bearings means the sys- measured by resistance temperature detectors (RTD).
tem is less complex and requires less maintenance. The Careful review of the vibration and temperature data
motor bearings are all that would need to be lubricated. can give an indication of potential bearing failure. Refer
Some motors come with sealed bearings that do not to Table 13.4 in the next section. If vibration occurs in
require adding lubrication. For most services, greased the unsatisfactory range, then new bearings should be
TABLE 13.4
Fan Bearing Vibration Limits
Vibration Limits for Fans (Bearing Measurements)
Vibration Severity Displacement (Mils Peak to Peak)
Velocity (in./s) Quality For Various Fan Speeds (RPM)
Peak RMS Rigid Flexible 3600 1800 1200 900 720 600
0.025 0.018 Good Good 0.1 0.3 0.4 0.5 0.7 0.8
↓ ↓
0.040 0.028 0.2 0.4 0.6 0.8 1.1 1.3
0.062 0.044 0.3 0.7 1.0 1.3 1.6 2.0
0.100 0.071 0.5 1.1 1.6 2.1 2.7 3.2
0.16 0.11 Satisfactory 0.8 1.6 2.5 3.3 4.1 5.0
0.26 0.18 ↓ Satisfactory 1.4 2.7 4.1 5.4 6.8 8.1
0.40 0.28 Unsatisfactory ↓ 2.1 4.2 6.3 8.4 10.5 12.6
0.62 0.44 ↓ Unsatisfactory 3.3 6.6 9.9 10.3 16.5 19.8
1.00 0.71 Unacceptable ↓ 5.3 10.6 15.9 21.2 26.5 31.8
1.56 1.10 ↓ Unacceptable 8.3 16 24.8 33.1 41.4 49.7
Source: Jogensen, R., Fan Engineering, Buffalo Forge Company, Buffalo, NY, 1983.
installed on skids, or in critical areas where low noise

is required, should be installed with vibration isola-
tors. The skid may need additional stiffening above
and beyond what would normally be required to sim-
ply support the equipment. A flexible sleeve should
be attached to the blower discharge and, depending
upon the installation, the inlet, so that vibration will
not be carried into the ducting (see Figure 13.26). This
is referred to as vibration isolation. The fan vendor
should balance the wheel to ISO G 6.3 as a minimum
for limit vibration.
Aside from proper blower sizing and overall system
design, the single most important time in the life span
of the blower will be the initial installation and balanc-
ing. An experienced and competent turbo machinery
installation specialist can correctly align the motor,
FIGURE 13.24 coupling, and impeller, and can make the difference
Maintenance of arrangement 8 bearings. between replacing the bearings as soon as 2 years or
the full life time of 8 or 10 years. Vibration readings are
purchased and a shutdown of the blower in question typically taken on the bearings, if they exist, and loca-
planned. If the vibration is in the unacceptable range, tions on the motor. Any velocity reading less than 0.1
the fan should be shut down to avoid possible cata- in./s (2.5 mm/s) is considered good. For other ranges,
strophic failure, which can lead to equipment damage see Table 13.4.
and injury to personnel. Bearing temperatures greater Different kinds of vibration can be indicative of the
than 150°F–170°F (65°C–76°C) indicate that the bear- problem. See the Table 13.5 for more information.
ings need to be regularly monitored to determine if any
maintenance is required. Figure 13.24 shows an arrange- 13.5.7 Shaft Seals
ment 8 fan during maintenance of the bearings.
Because the shaft on centrifugal fans has to pass
through the housing on at least one side of the housing,
shaft seals are required to keep air or gases from escap-
13.5.6 Vibration and Installation
ing through these openings. On combustion air blow-
Blower foundations should be rigid and approxi- ers, some amount of leakage is acceptable and expected.
mately five times the weight of the blower itself. See Felt and graphite seals do the job well and only need to
Figure 13.25 for a typical foundation drawing. Units be replaced when overhauling the fan. On other critical
Fan wheel
Driver
Inlet
box
Outboard
concrete
pedestal Fan
HSG
Foundation material Foundation material (Min. WT = 5X Fan and driver WT)
Soil Soil
Bedrock
Bedrock
FIGURE 13.25
Fan foundation.
FIGURE 13.26
Inlet and outlet expansion joints for vibration isolation of ducting.
TABLE 13.5
Fan Vibration Diagnostic Clues
Amplitude Frequency Phase Possible Cause
Steady-radial 1 × RPM Steady-single Unbalance (see text for clues to
largest reference mark correcting)
Axial largest 1 or 2 × RPM 1,2, or 3 reference Misalignment or bent shaft
marks (check with dial indicators)
Unsteady 1 × RPM Unsteady Resonance (check for impeller or
other flexibility using modal-
analysis techniques)
Unsteady 2 × RPM Two reference Looseness (check bolts, keys, etc.)
marks
Unsteady RPM ÷ 2 Unsteady Oil whip (uncommon for fans)
Low 60 or 120 Hz One or two Electrical
rotating marks
Erratic Many × RPM Erratic Faulty antifriction bearings
Erratic 1 or 2 × belt RPM Unsteady Defective belts (check by freezing
with strobe)
Low Blade-passing Steady Aerodynamic (check cutoff
clearance)
Source: Jogensen, R., Fan Engineering, Buffalo Forge Company, Buffalo, NY, 1983.
applications where escaping gases can cause injury or that the blower is connected to a duct and that the duct
death due to high temperatures, acids and caustics, or will have some noise suppression properties. Applications
toxins, a substantial shaft seal is required. Shaft seals where fresh air is drawn into the system will typically
can be a carbon ring seal, or packing glands. Note that have an unducted inlet. So, the unducted noise numbers
to obtain a seal, the seal material needs to be in contact should be used for noise estimates. Occasionally, lag-
with the shaft. This contact causes friction and as a result ging and noise insulation will have to be installed on the
of this friction, heat builds up in the shaft and bearings. blower and ducting to prevent noise transmission through
Other options may include a purged seal, where nitro- the steel to meet certain noise requirements. Noise is dis-
gen or another inert gas is continuously injected into the cussed in more detail in Chapter 11.
seal to maintain a positive pressure, preventing gases
from escaping the blower. Bellows seals and spring
13.5.9 Filtration
seals are also available depending upon the application.
Bellows seals offer a much greater amount of protection Filtration is rarely used for combustion applications.
but have a higher up-front cost and greatly increased However, if there is a large amount of particulates in
amount of maintenance. the area, such as sand, then a filter may be considered.
Installations in dusty or sandy climates may require fil-
tration. Filtration of particulates can prevent abrasion
13.5.8 Noise Considerations wear on the fan impeller. Filtration can be as simple as a
Blower noise considerations can be a big issue, especially replaceable corrugated filter, much like the air condition-
if the equipment is in an area where personnel are con- ing filter in a house, to the very complex pulse jet bag
tinuously exposed to the rotating machinery in question. filters. Other options are electrostatic dust collectors and
Typically, blower noise should be less than 85 dBA at a cyclonic settling drums. The options on filtration can be
distance of less than 3.28 ft (1 m), in an area where work- very complex and in some cases, the filtration system
ers perform their duties in an 8 h period. OSHA allows may cost more than the blower itself. Carefully consider
for higher levels depending upon the circumstances, but if filtration is required.
85 dBA is considered the industry norm. Inlet silencers
are available for fresh air applications. For applications
that require further noise suppression, a noise insulat-
ing enclosure, sometimes called an attenuator, may be
13.6 Operational Costs
installed, but these can be fairly expensive. The higher
the speed or pressure, the more noise a fan will generate. When considering operational costs, the user must
When considering noise, blower vendors typically dis- understand where the unit will operate the majority
cuss it in terms of ducted and unducted. Ducted means of the time. Never assume that the latest technology is
TABLE 13.6
Control Options Relative to Design Rate 13.7 Inspection and Testing
Normal Operation Control
The level of inspection and testing required will depend
100% of design rate Single speed blower upon three things, the criticality of the blower in service,
75%–100% of design rate Single speed blower with damper the size of the blower, and the level of comfort required
50%–75% of design rate 2 speed blower with damper by the owner. The most basic inspection check would be
25%–50% of design rate VFD a visual and dimensional check. This visual and dimen-
0%–25% of design rate VFD with damper sion check should verify all of the connection flange
dimensions and the anchor bolt hole dimensions. Note
that for off-the-shelf blowers, a dimensional check may
what is best for the application without reviewing other be all that is available. Material certifications may also
possibilities. For instance, take the example of Table 13.6 be required. Discuss the quality control and documen-
showing possible recommended control schemes for a tation requirements with the blower vendor before pro-
flare. The users must understand where the system is curing services. Additionally, the basic inspection may
going to operate most of the time; then it might be ben- require an impeller balancing certificate with reference
eficial to evaluate if an inlet damper or VFD would save to the applicable balancing standard.
money. If the unit was a batch process that was full on Fans that are designed for low or high temperatures
when used and completely off when not, then a single or hazardous and toxic gases may require additional
speed blower would be sufficient. Further, if the nor- inspection of the welds. The type of testing required
mal operating point was only half of the smokeless will depend upon the material of the components to be
capacity, then a two speed motor might be of lower cost welded and the weld material. Radiographic examina-
than a VFD and still function approximately the same. tion uses x-ray exposure to take a “picture” of the weld
However, if normal rates are below 50% then a VFD metal to see if there are any flaws within the weld and
would be justified and substantial operating cost sav- the surrounding area. However, the shape of the piece
ings can be achieved. to be welded must be somewhat flat to use this process.
Table 13.7 shows the potential relative savings using Other methods that are available for testing of welds
different control strategies. The base case assumes no include ultrasonic testing to look for cracks or imper-
damper control, while the others assume the process flow fections below the surface, liquid penetrant examina-
required to be 110,000 CFM (3,100 CMM). All other aspects tion to look for surface cracks, and magnetic particle
are the same and the cost of electricity is $0.15/kWh. The examination is used to find defects in alloy materials
calculation is and welds. Additionally to testing, weld procedures
and other documentation may be required to verify
Annual power cost = HP × 0.746 kW/hp × 365 days/year proper techniques. It is important to discuss inspection
requirements with the fan vendor before purchasing
× 24 h/day × $0.15/kWh
any equipment.
Some applications such as fans handling toxic or haz-
HP is obtained by the HP curves in Figures 13.27 ardous gases may require a pressure test of the housing.
through 13.29. This is done in order to verify that there are no leaks in
Another consideration is the blower efficiency. Use the the housing. Pressure testing may be accomplished using
fan wheel that offers the highest efficiency. Centrifugal water, known as a hydro test, or air, known as a pneu-
blowers should have a mechanical efficiency of 60%– matic test. The test pressure required is usually 1.5 times
70%, while vane axial blowers should be in the 60%– the working pressure of the fan for a minimum of 30 min.
80% range. Fans with 50% efficiency or less should be The mechanical run test is performed to verify if the
re-evaluated. fan and the drive system are correctly balanced and is
TABLE 13.7
Potential Controls Cost Savings
Control Type Flow Rate % Operation hp Cost Savings
Base case 150,000 100 1275 $1,232,167 —

Outlet damper 110,000 73 1095 $1,073,363 $158,799
Inlet damper 110,000 73 880 $862,614 $369,553
VFD 110,000 73 521 $510,707 $721,460
50 System resistance
SP3 curve
Static pressure–inches H2O

40 SP1 = 40˝
Outlet
damper
Parallel
30 resistance
blade
outlet
SP2 = 22.2˝ damper
20
10
Opposed
blade
With outlet outlet
Design
dampering damper
2000 fan BHP
BHP
follows base
HP curve BHP1 = 1257
1000
BHP2 = 1095
50 100 150 200 250

Thousands CFM
FIGURE 13.27
Outlet damper fan curve with horsepower.
50 System resistance
SP–1180 RPM
curve
Static pressure–inches H2O
40 SP1 = 40˝
(T.B.)
30 Parallel bladed
Fan curve with 100%
inlet box
open damper
damper
SP2 = 22.2˝
20 Fan curve (M.C.R.)
with partially
closed
damper
10
2000 Radial
BHP1 = 1257 inlet
BHP
damper
PM
1180 R Prespin inlet
1000 BHP–
damper has reduced
BHP2 = 880 BHP curves
50 100 150 200 250

Thousands CFM
FIGURE 13.28
Inlet damper fan curve with horsepower.
50 SP – 1180 RPM System resistance
Static pressure—inches H2O

curve
40 SP1 = 40˝
Speed control:
30 SP – 880 RPM 1. 2 speed motors
2. DC motor
3. Variable speed AC motor
20 SP2 = 22.2˝ 4. Steam turbines
5. Eddy current
10 6. Fluid drive
Fan laws for RPM change

CFM2 = CFM1 (RPM2/RPM1)
2000 P2 = P1 (RPM2/RPM1)2
BHP2 = BHP1 (RPM2/RPM1)2
0 RPM
BHP
1000 – 118 HP1 = 1257

BHP
H P – 88 0 RPM BHP = 521
B 2
50 100 150 200 250

Thousands CFM
Annual operating costs: (880 hp)(0.746 kW/hp)(365 days/year)(24 hours/day)(0.05 $/kWh)=$275,722/year
FIGURE 13.29
Speed control fan curve with horsepower.
usually performed for a minimum of 2 h. During the 2 h, requirements, the price may be four times as much as a
the bearing vibration and temperature are measured at standard unit, or even more.
regular intervals. Mechanical run tests are generally
not performed on fans with less than 30 hp. Sound tests
may be performed as well to determine the overall noise
level. The sound meter should be calibrated for the A
scale and noise measurement taken at 3.28 ft (1 m) from 13.8 Maintenance and Troubleshooting
the blower and at multiple points around the blower to
find the maximum level. Fan maintenance is extremely important. The most
For critical fans and large fan applications, a perfor- common items in need of attention are the fan bearings.
mance test may be required. A performance test is what The speed of the fan and the shaft size determine the
the fan manufacturer would do in order to develop a size of the bearings and consequently the maintenance
fan curve for a particular fan. The fan is connected to requirements. Grease bearings should be lubricated
a test chamber that has a flow straightener, a pitot tube routinely according to the manufacturer’s instructions.
and manometer set up to measure the total and velocity Lubrication of the bearings is extremely important and
pressure, and a set of dampers to adjust the back pres- should be carried out regularly. Even automatic lubri-
sure on the chamber and fan. cation systems should be routinely checked in order
A final note concerning quality control and fan ven- to catch problems before a failure occurs. During shut-
dors is that it is recommended to discuss in detail any downs, bearings, if not too worn, need to be cleaned
requirements concerning API 560 or 673. These stan- and repacked with proper grease or oil. Bearings that
dards contain a varying degree of design and inspec- need to be replaced should be overhauled or the entire
tion requirements that some vendors are simply not assembly replaced. The wheel, coupling, and motor
prepared to meet. API 560 and 673 should not be should be checked for balance, stress cracking, and
applied to small fans, with less than 30 hp (22 kW), excessive vibration. The fan housing should be checked
unless all parties are in agreement. The documenta- and debris removed.
tion in these standards is stringent. Additionally, they Often fans may develop problems in the field
can require a torsional and lateral analysis of the shaft. despite the best maintenance efforts. Table 13.8 pro-
Applying these standards can greatly increase the vides a general list of common problems and possible
cost of a blower. Depending upon the degree of the solutions.
TABLE 13.8
Blower Trouble Shooting Guide
Problem Probable Cause
Hot bearings Recheck temperature measurement (don’t trust “hand feel”)

Over lubrication
Wrong lubricant (check MFG recommendations)
Improper internal clearances
Misalignment
High ambient temperature, no ventilation around bearings
Sunlight
Excessive thrust loading
Improper shimming
High vibration
Bearing race turning inside housing
Moisture in the bearings
Excessive vibration Accumulated material on impeller
Worn or corroded impeller
Bent shaft
Impeller or sheaves loose on shaft
Motor out of balance
Impeller out of balance
Sheaves eccentric or out of balance
Bearing or drive misalignment
Loose or worn bearings
Weak or resonant foundation
Foundation unlevel
Structure not cross braced
Fan operating in unstable system conditions
Excessive horsepower draw Fan speed higher than design
Gas density higher than design
Impeller rotating in the wrong direction
Fan size or type not appropriate for the application
Insufficient air flow Duct elbows near fan inlet or outlet
Restricted fan inlet or outlet
Impeller rotating in the wrong direction
Fan speed lower than design
System resistance higher than design
Dampers shut
Faulty ductwork
Dirty or clogged filters and/or coils
Inlet or outlet screens clogged
Excessive air flow System resistance less than design
Fan speed too high
Filters not in place
Air registers or grilles not installed
Improper damper adjustment
Inoperative fan Blown fuse
Broken belts
Loose sheave
Motor too small
Wrong voltage
Excessive noise Imperfection in bearing elements
Improper clearance
Internal wear of bearing parts
5. E. A. Avallone and T. Baumeister, III, Marks Standard

Handbook for Mechanical Engineers, 9th edn., McGraw
References Hill, New York, 1987.
1. R. Jogensen, Fan Engineering, Buffalo Forge Company, 6. H. L. Gutzwiller, Fan Performance and Design, Robinson
Buffalo, NY, 1983. Industries, Inc., Zelienople, PA, January 2000.
2. AMCA Publication 200–95 (R2007), Air Systems, Air 7. API Standard 560, Section 11, Fired Heaters for General
Movement and Control Association International, Inc., Refinery Service, 3rd edn., Washington DC, May 2001.
Arlington Heights, IL, 2007. 8. API Standard 673, Centrifugal Fans for Petroleum,
3. AMCA Publication 201–02 (R2007), Fans and Systems, Air Chemical, and Gas Industry Services, 2nd edn., Washington
Movement and Control Association International, Inc., DC, January 2002.
Arlington Heights, IL, 2007.
4. R. W. Fox and A. T. McDonald, Introduction to Fluid
Mechanics, John Wiley & Sons, Inc., New York, 1985.
14
Burners and Combustion Systems for Industrial and Utility Boilers
Vladimir Lifshits
CONTENTS
14.1 Introduction.................................................................................................................................................................... 359
14.2 Burner Design Fundamentals...................................................................................................................................... 360
14.3 Techniques of Reducing NOx Formation.................................................................................................................... 362
14.3.1 Flue Gas Recirculation and Injection of Steam into the Flame................................................................... 363
14.3.2 Air Staging.......................................................................................................................................................... 365
14.3.3 Fuel Staging........................................................................................................................................................ 366
14.3.4 Lean Premixed Combustion............................................................................................................................ 366
14.3.5 Furnace Gas Circulation and Combining Different Techniques................................................................ 367
14.4 Burners for Package Boilers.......................................................................................................................................... 368
14.4.1 Conventional Round Burners and Burners with Air Staging..................................................................... 369
14.4.2 Burners with Fuel Staging................................................................................................................................ 371
14.4.3 Burners with Partial Lean Premix Combustion............................................................................................ 372
14.4.4 Premixed Burners...............................................................................................................................................374
14.4.5 Burners for Low Heating Value Gaseous Fuels and Fuels with Highly Variable Composition............ 376
14.4.6 Oil Atomizers for Package Boilers................................................................................................................... 377
14.5 Burners for Enhanced Oil Recovery Boilers.............................................................................................................. 379
14.6 Burners for Multiple-Burner Process Steam Field-Erected Boilers......................................................................... 382
14.7 Gas and Oil Burners for Utility Boilers...................................................................................................................... 385
14.7.1 Combustion Systems for Oil and Gas Wall-Fired Utility Boilers............................................................... 385
14.7.1.1 NOx Reduction Techniques in Wall-Fired Utility Boilers............................................................. 385
14.7.1.2 Burners for Wall-Fired Utility Boilers.............................................................................................. 388
14.7.2 Combustion Systems for Corner-Fired (T-Fired) Boilers.............................................................................. 389
14.7.3 Oil Atomizers for Utility Burners................................................................................................................... 391
14.8 Specialized Burners....................................................................................................................................................... 391
14.8.1 Warm-Up Burners.............................................................................................................................................. 391
14.8.2 Flue Gas Reheat Burners................................................................................................................................... 392
14.9 Summary......................................................................................................................................................................... 394
References................................................................................................................................................................................. 394
Most modern varieties of boilers with burners as a spe-

cific part of the system were developed predominantly
14.1 Introduction
over the last century, whereas earlier boilers were firing
Burners for industrial and utility boilers form a large fuels such as coal or wood on a grate placed under the
class of combustion devices designed to efficiently con- steam and water carrying tubes in an enclosed space
vert the chemical energy of the fuel into heat within the with some rudimentary means of combustion process
space provided by a boiler radiant section also referred intensification. The best legendary book on boilers that
to as the furnace. Further, heat absorption occurs in the withstood over 40 editions was originally published by
convection section. The combination of the radiant sec- the Babcock & Wilcox Company and is considered a
tion and convection section substantially defines the bible for the boiler industry.1 The book gives an excellent
boiler. Chapter 1 also describes various types of boilers. overview of boilers with an emphasis on utility boilers,
359
but goes very little into the more specific field of burners to the burner components with typical preference for
and especially modern gas- and oil-fired burners that compact flames.
are a focus of this chapter. Another very diverse book on However, this classical approach often results in high
the subject of industrial combustion published in 2001 levels of NOx formation. For example, using gaseous
gives a good overview of combustion technologies of fuels, the oxidation process started from the fuel-rich
boilers at that time.2 limits of flammability, generating high prompt NOx,
Two main factors that have driven the development followed by combustion with high flame temperatures
of new industrial burners over the last few decades close to those observed in intense stoichiometric flames,
were the worldwide increase in the cost of fuel and the generating high thermal NOx. For liquid fuels the situa-
requirement and legislation for cleaner air. These mod- tion was somewhat similar except that the droplet evap-
ern standards, already applied over a substantial period oration phase preceded the bulk of the fuel oxidation.
of time, required close to absolute completeness of the In order to reduce NOx formation without detrimental
combustion process, minimum losses of heat within effects on the performance characteristics, the com-
the exhaust flow, and minimum supporting energy bustion process needed to be substantially modified.
requirements. Now more attention is being focused on These modifications resulted in new low NOx concepts
the effective utilization of byproducts and lower grade applied to the industrial burner design and new low
fuels in addition to the main conventional fuels like NOx concepts of the overall combustion process inside
coal, oil, and natural gas. However, most of all, the need the furnace that, where possible, are applied together
for reducing air pollution resulting from combustion for achieving the desired performance. With some
was the main factor driving the development of new exceptions, these new burners and combustion systems
burners. In many instances, these new requirements overall became more complex and more demanding in
resulted in levels of pollutants (see Chapter 10) and terms of the required mechanical accuracy of all the
specifically NOx emissions entering the atmosphere components and sophistication of the combustion con-
that were approaching close to the theoretical limits of trols that were being developed concurrently with the
a particular process. new burners.
As boilers vary considerably, so must the burners. While low NOx burners and combustion systems
Descriptions of various burner types in this chapter will present a very substantial part of modern combustion
more or less follow the variety of boiler types used by equipment design, the field of industrial boiler burners
the modern industries as well as specific requirements is rather vast and has room for a variety of products not
to their operation. The following will be considered: requiring any special means of NOx emission reduction.
Examples of such equipment are auxiliary burners like
• Burners for package boilers generating process warm-up burners that are not meant for extended con-
steam and high pressure superheated steam tinuous operation, burners firing fuels that naturally
• Burners for field-erected boilers equipped with generate low NOx such as fuels with low heating values
several burners having a substantial quantity of nonoxidizing species
or systems where the focus of NOx control is shifted to
• Burners for wall-fired utility boilers,
post-combustion cleanup. This equipment can be highly
• Burners for tangentially fired (T-fired) boilers specialized and may need to meet unique requirements
• Burners used primarily for boilers warm-up other than the NOx performance parameters and will
require considerable expertise to be properly designed
Combustion equipment for low NOx once through and applied.
steam generators (OTSG) used for enhanced oil recovery This chapter is structured to outline various com-
will be discussed. At the end of the chapter, examples bustion features of different boilers, goes into some
will be given of burners used in black and red liquor details of applicable emission reduction techniques, and
boilers of the paper industry and a new class of burners describes the design of different burners and associated
for reheating the flue gas upstream of selective catalytic performance characteristics.
reduction (SCR) equipment located in the boiler exhaust.
For many years the main goal of a high quality burner
and especially an industrial size burner was to provide
mixing between the fuel and air that will be complete
within the allocated space while providing ignition of
14.2 Burner Design Fundamentals
the fuel/air mixture starting at an early stage in the
mixing process smoothly propagating through the fuel As the capacity, fuels, and performance requirements
air mixture as it converts to products. This primary goal for industrial combustion equipment vary widely, so
was to be accomplished without creating heat damage must the burner designs. The only required common
Burners and Combustion Systems for Industrial and Utility Boilers 361
components of all industrial burners are those parts register draft loss (RDL), selecting a burner with a low
providing the desired air flow shape and structure and RDL can produce a benefit by lowering fan horsepower.
fuel distribution components including multiple fuel However, lower RDL burners can be physically bigger
injectors that may also be the same parts affecting the and may require bulkier fans and larger air ducts and ple-
air flow. Refractory throats or quarls surrounding the nums. Lower RDL burners generally also produce lower
base of the flame, while common for many burners, may intensity flames and have lower turndowns. Depending
not be a necessary attribute for some industrial burn- on the application, the burners are usually designed for a
ers. Many other components like the air plenum around draft loss ranging at high fire from as little as 2 to as high
the burner—the wind box—may be a component of as 12 in. WC (5 to 30 mbar) with the most common range
the combustion system but not a single burner. These being from 4 to 8 in. WC (10 to 20 mbar).
modern burners are also equipped with many common Most industrial sites have similar utilities that can
components, providing a means of flame ignition and be used by the burner. Gaseous fuels like natural gas
monitoring. are typically available with considerable pressure rang-
The process of burner design is a complicated multi- ing from 10 psig (0.7 barg) to about 50 psig (~3.4 barg).
step and usually iterative process requiring deep under- Liquid fuels, if used, are delivered by pumps generating
standing of different fundamental disciplines, as was pressure as necessary for the type of selected atomizers:
explained in Chapter 1 and Refs. [3,4]. As for any burner, 100–200 psig (7–14 barg) when steam– or compressed
the design of an industrial burner starts from consider- air–assisted atomizers are used or 300–1000 psig (20–70
ations of basic fluid dynamics principles applied for siz- barg) for mechanical atomization. The steam for atomi-
ing the main burner components—air and fuel passages, zation is usually available with pressures of 100–250 psig
geometrical characteristics of the flame stabilization (7–17 barg). If compressed air is used, its typical available
components like bluff bodies, and flow swirling registers pressure range is about 15 psig (1.0 barg) or ~100 to 150
or spinners. During this phase, the designer must make psig (7–10 barg). A good burner design would effectively
sure that the necessary quantities of air and fuel can use most of the available mechanical energy of the utili-
be delivered to the combustion zone with the required ties to achieve the desired combustion results instead of
distribution quality and target pressure losses on both excessively wasting energy in the control components.
the fuel and air sides of the burner. As a second step, The process of modern burner design or burner
the designer conceives of the desired mixing patterns selection also requires in-depth understanding of the
between the air and fuel and checks that the means of origination of different pollutant formations and the
achieving them are adequate, and finally checks the pat- available means of its reduction described in this chap-
terns of flame propagation and stabilization. What often ter. Modern modeling tools like computational fluid
simplifies the process is when the designer can start with dynamics (CFD) software packages have reached a high
a known burner design and apply these principles for level of sophistication and allow with some accuracy
modifying the combustion device to the new conditions simulation of some simple combustion problem includ-
with respect, for example, to the changes in the fuel or ing detailed modeling of the burner performance as
design capacity or some other performance parameter. described in Chapter 9. The more advanced CFD pack-
Among the important characteristics to consider here ages also include some amount of chemical kinetics to
would be some bulk characteristics like the swirl num- approximate modeling of the formation of CO and NOx.
ber for the flow and the ratio of momentums between the Some companies have also developed their own codes
air flow and fuel jets as described in several well-known for estimating NOx emissions in a postprocessor using
Refs. [5,6]. For an even deeper understanding of the issues, the modeling results by commercial codes. The suc-
the designer may even need go into the field of chemical cessful use of CFD and additional tools requires know-
kinetics and turbulent mass transfer models and con- ing their limitations and considerable background in
sider criteria such as Damköhler numbers defined as the mechanical engineering and physics, without which the
ratio of characteristic residence time in specific areas of CFD results may deviate greatly from reality or miss
the flame to the characteristic reaction time. important phenomena of the process.3
The combustion air to the burners is usually provided While predominantly operating on gaseous fuels due
by a centrifugal fan (see Chapter 13), which can be posi- to its currently comparatively lower cost, many indus-
tioned locally or remotely to the burner. The combustion trial boiler burners are designed to have oil firing capa-
air is provided with enough pressure and mass flow to bilities as a backup. A few sites still operate on oil where
overcome the sum of the burner air-side pressure loss gas is not available. Liquid fuels are burned in indus-
and the boiler convection section pressure loss, unless trial burners using fuel atomizers described in Chapter
the boiler is of balanced draft type equipped with an 5. The combustion process of atomized fuel is very dif-
induced draft fan. As the fan power usage is a function ferent from gas and difficult to accurately model even
of the burner differential pressure, also sometimes called with modern CFD packages. The droplets are formed in
different sizes and the characteristics of the droplet size parameter when considering different burner designs
distribution are seldom well defined. Even with the use and combustion solutions in this chapter.
of advanced particle size instrumentation, the design of
the atomizer specific to a combustion process contains a
substantial degree of proprietary know-how.
It is also important to understand that as the burner
operates it interacts with several dynamic systems. 14.3 Techniques of Reducing NO x Formation
These interactions take place at several levels. The
Two species nitrogen oxide (NO) and nitrogen dioxide
first one is obvious—the flame behavior depends on
(NO2) together constitute the definition of NOx. During
the configuration of the space available for the flame
the combustion process most of the NOx formed will be
and the temperature of the surrounding walls confin-
NO. In the atmosphere the NO is then oxidized to NO2
ing the flame. The second level is the configuration of
in the presence of sunlight and certain hydrocarbons.
the supply and distribution of combustion air or com-
When computing NOx emissions on a weight basis all
bustion air mixed with some amount of flue gas. The
the NOx is considered as NO2. When measured out of
third level is acoustical interactions with the furnace
a combustion chamber, the instrument should measure
space, the convection part of the system, the exhaust,
both NO2 and NO as parts per million volumetric, dry
as well as the wind box and the air delivery system.
(ppmdv).
These interactions may be rather complicated and dif-
As described in Chapter 10, there are three generally
ficult to anticipate. Overlooking the possibility of these
recognized significant sources of NOx emissions formed
interactions can result in difficulty in solving problems
during the combustion process:
discovered only during the burner commissioning and
should be addressed as a complete system issue during
the design phase of any project. • Thermal NOx, formation of which was first
There are several types of pollutant emissions (see described by Y. B. Zeldovich,7 comes from
Chapter 10) that are affected by the burners and not just oxidation of atmospheric nitrogen (N2) to NO
by the type of fuel or minor species that may be pres- inside the areas of the flame where tempera-
ent in the combustion air such as NOx emissions, car- tures are developed in excess of about 2600°F
bon monoxide (CO) emissions, unburned hydrocarbons (1400°C).
(UHC), and particulate matter (PM). These pollution • Prompt NOx experimentally discovered first by
emissions are discussed in Chapter 10. In general, well- C. P. Fenimore8 is also the result of atmospheric
designed combustion devices generating a flame inside nitrogen oxidation formed predominantly in
an appropriately sized volume can easily achieve negli- the initial parts of the flame via reaction with
gible amounts of CO and UHC emissions. The organic fuel radicals where local stoichiometric condi-
portion of PM related to the incomplete oxidation of tions are fuel rich.
liquid fuels can also be easily reduced to insignificant • NOx originated from oxidation of nitrogen pres-
levels with good quality fuel atomization. The control of ent in the fuel in several forms other than N2
NOx emissions can be difficult and will be an important also called fuel bound nitrogen (FBN).
Formation rate of NO at 2% O2 (vol)

1.0E+05
1.0E+04
NO, ppm/s
1.0E+03
1.0E+02
1.0E+01
1.0E+00
2500 2700 2900 3100 3300 3500
Temperature, °F
FIGURE 14.1
Approximate rate of thermal (Zeldovich) NO formation.
An approximate theoretical relation for thermally with minimum overall excess air and or using air stag-
formed NO as a function of temperature is shown ing explained in Section 14.3.2 reduces the conversion
in Figure 14.1. The chart takes into account only the of FBN to NOx.
forward going reactions of nitrogen oxidation that in Extensive literature exists on the subject of NOx for-
boilers is almost always the case. It shows about 10 mation and control for those who want to study deeper
times increase in the NO formation rate with about into the subject.
220°F (~120°C) increase in temperature. However, All practical ways of reducing NOx formation during
this steep increase in the NO with temperature never combustion that are described in the following sections
directly translates to the relation between the boiler complicate the combustion process and make it more
effluent NOx and peak theoretical flame temperature, susceptible to increases in other emissions (CO, UHC,
also called the adiabatic flame temperature (AFT). PM). So the art of delivering lowest NOx combustion
The explanation of this is that the temperatures close often becomes a balancing act of designing the process
to the AFT occur in the flame in only small areas at to lower NOx with minimum trade-offs to the increase
low residence times due to heat transfer from the of other pollutants.
flame to the surrounding lower temperature gas and
combustion chamber walls and the rate of heat dissi-
14.3.1 Flue Gas Recirculation and Injection
pation from these areas is greatly accelerated with the
of Steam into the Flame
temperature. Furthermore, some experimental NOx
data presented in Section 14.3.4 indicate a much lower The peak flame temperature and thus thermal NOx can
dependence of NO formation on temperature that be effectively reduced when the initial reactants—fuel
is difficult to explain on the basis of known kinetic or combustion air—are mixed with nonreactive gases
reactions. like atmospheric nitrogen, carbon dioxide, or water
If uncontrolled, the concentration of thermally pro- vapor introduced into the combustion zone at tempera-
duced NOx in medium size boilers can reach the level tures substantially below the temperature of the flame.
of a few hundred parts per million (ppm) and higher if As all the aforementioned inert-for-the-combustion
preheated air or oxygen-enriched air is used. In large gases are present in the combustion products, the recir-
boilers the uncontrolled NOx could be much higher, culation of cooled flue gas back to the burner (flue gas
reaching a 500–1200 ppm level and more.9,10 recirculation [FGR]) becomes a powerful way of con-
When burning conventional fossil fuels, the pro- trolling thermal NOx.
duction of prompt NOx is typically considered in The quantity of FGR is usually defined as the ratio of
the range of 10–15 ppm. This number, however, is the mass of recirculating FGR, Mflue gas, coming from the
very approximate as the formation of prompt NOx same combustion process to the mass of combustion
also depends on the peak flame temperature and products exiting the stack, Mfgr:
there is no easy experimental way to separate it from
the thermal NOx and produce data for validation
of theoretical models. The control of prompt NOx M fgr
FGR = (14.1)
is only considered after the thermal NOx is already M flue gas
reduced to a very low level. Utilizing lean premixed
burners accomplishes control of both thermal and
prompt NOx. A system designer needs to be careful in sizing the
FBN is mostly a concern in fuel oils and biofuels. FGR delivery system if the outgoing flue gas is diluted
Gaseous fuels on occasion may contain ammonia or with combustion air (leakages in the air heaters, for
other elemental nitrogen compounds that at medium example) or taken from another process that may pro-
temperatures readily oxidize to NOx. The percent con- duce flue gas with a different composition.
version of FBN to NOx depends on many factors: spe- As a generic heat balance equations will show, the
cific molecules with nitrogen carrying radicals; rate recirculation of a typical flue gas in an amount of 10%
of release of species with nitrogen into the gas phase results in 6%–8% reduction in the theoretical peak flame
and the amount of FBN in the fuel. A conversion rate temperature (AFT). Relations between NOx and FGR
of 100% of FBN present in fuel oil in the amount of 0.1% for several specific cases of firing fuel without FBN are
by weight would result in about 130 ppm of NOx in a shown in Figure 14.2.12
low oxygen exhaust. Number 2 oil typically contains The same effect can also be achieved when steam or
from 0.01% to 0.05% of FBN11 with the rate of conver- finely dispersed water is injected directly into the flame,
sion to NOx of 75%–40%. Heavy oil usually contains or mixed with combustion air or fuel. As FGR is almost
between 0.3% and 0.6% of FBN with a conversion to always available, the monetary penalty of using FGR
NOx in the range of 50%–30%. Operation of the burners for NOx control is mostly associated with the cost of
90% In Case 2, the power increase factor with FGR is much

80%
more substantial:
70% n
PfanFGR   %FGR  Tfgr 
= 0.1 + †1 + 1.08 * 
 100%  * Tamb 
60% (14.3)
Pfan 0
NOx reduction

50%
40% where the power factor n varies between 2.3 and 2.8,
30% Power boilers
depending on the relative hydraulic resistance of the
Package boilers burner to the boiler convection section, with n being
20% lower for boilers with convection section pressure drops
Boilers with high
10% air preheat much lower than the burner resistance. In the aforemen-
tioned equations the temperatures of FGR and air passing
0%
0% 10% 20% 30%
through the fan, Tfgr and Tamb, are absolute temperatures.
The equations given earlier are rough references and
FGR %
approximately account for the increased pressure drop
FIGURE 14.2 across the burner and the boiler convection section as
Approximate relation of NOx reduction with FGR for fuels without well as additional pressure losses for inducing the flue
FBN.
gas. For any particular application the requirements
of the fan characteristics can be accurately determined
mechanically moving FGR back to the burner and is usu- based on fundamental considerations of the pressure
ally insignificant compared to a change in any heat losses drop characteristics of the boiler and selected burner.
with the exhaust due to a slight increase in outgoing flue FGR can also be added to the combustion air using
gas temperature. For retrofit applications the increase in a separate FGR fan. This “forced FGR” can be mixed
stack temperature and thermal losses is on the order of with air with the use of a custom designed mixing
5°F–10°F (3°C–6°C) per 10% of FGR, depending on the device or delivered to the burner exiting cross section
design of the boiler convection section. This rule of thumb through a plenum with ports. In the latter case, the
was determined by observations and calculations of the FGR can be strategically directed to the parts of the
heat transfer inside the furnace and the convection section. flame where it is most effective. The method is also
For comparison, a 10°F (6°C) increase in stack temperature called selective FGR—the term that is contrary to the
reduces boiler thermal efficiency by about 0.25%. forced or induced FGR uniformly (or in bulk) mixed
The extent that thermal NOx can be controlled with FGR with combustion air.
strongly depends on the ability of the burner to remain There are pros and cons with either way of introducing
stable and provide sufficient completeness of combus- the FGR. In retrofits, the forced FGR may provide a bet-
tion with the rate of combustion reactions reduced due to ter chance of reusing the existing combustion air fan, but
reduced temperatures. This issue will be discussed fur- would require more complicated FGR ducting arrange-
ther in this chapter for specific types of burners. ment and more complicated combustion controls (see
The addition of FGR to the combustion air prior to also Chapter 12). Some burners, described later in this
the forced draft fan demands a substantial increase in chapter, are also better suited for operation with forced
the fan capacity. The increase depends on whether the FGR if delivered directly to the burner. Utilizing induced
boiler and the burner were originally designed for oper- FGR, consideration must be given to additional water and
ation with FGR—Case 1, or the FGR is a retrofit addition sulfuric acid condensation and corrosion problems that
to the system—Case 2. are better contained when delivering forced FGR to the
Using fan laws (see Chapter 13), it is easy to estimate burner. The benefit of induced FGR is its simplicity.
factors for the fan power increase prior to making the In some cases the flue gas can be mixed with gaseous
selection of the specific equipment or determining if the fuel. For a given percentage of FGR, this method poten-
existing fans will work. In the first case the power of the tially achieves somewhat better NOx reduction than the
fan for the same boiler capacity needs to be increased at FGR mixed with combustion air. This can be explained
least about linearly with the amount of FGR as defined by concentrating the FGR in the high temperature com-
by the approximate ratio shown in Equation 14.2: bustion zone rather than spending it partially on the
dilution of the excess combustion air initially bypass-
ing the combustion zone. Except for a few specific cases,
PfanFGR the technique is difficult to use. It requires specially
 %FGR  Tfgr
= 1.1 + 1.08 * 
 100%  * Tamb
(14.2) designed burners and also complicates the controls and
Pfan 0
safety system.
A small amount of FGR can be mixed inside the burner 14.3.2 Air Staging
with gaseous fuel delivered at high pressure using the
Thermal NOx may also be reduced by burning fuel
fuel gas mechanical energy to aspirate flue gases from
under local conditions substantially different from stoi-
the boiler exhaust and pass the fuel/FGR blend through
chiometric such as lean and rich, when combustion does
the fuel injectors, which is also called fuel induced recir-
not generate high NOx forming temperatures.
culation (FIR). FIR works if the burner is designed for
Using conventional combustion equipment, the fuel
high fuel supply pressure usually in excess of 20 psig
burning typically starts after engaging only a partial
(1.3 barg) and with burners designed for low pressure
amount of air well below the stoichiometric amount.
injectors of the fuel–flue gas blend into the combus-
The balance of the air is added to the combustion after
tion zone. In other cases, an FGR fan with substantial
the products of combustion lose a substantial amount
static pressure is needed. Levels of NOx reduction with
of heat. Thus, the initial rich section of the flame and
this technique may be quite substantial—up to about
the lean final part of the flame are at lower tempera-
80%–85%.
tures due to prior heat transfer from the rich section.
Contrary to the low operating cost of using flue gas
The result is a reduction in peak flame temperature
to control NOx, the use of steam or water injection for
over the entire flame zone and a reduction in ther-
NOx control carries substantial penalties for the effi-
mal NOx. When the burner is designed to enhance
ciency of the boiler or combustion system. The amount
this described effect, the technique is often called air
of steam that is equivalent to 10% FGR would be equal
staging (see Figure 10.21). For the concept to work, it
to about 4.5% of the boiler steam output—an unbear-
is essential for the products of the initial rich flame to
able loss of the system efficiency as well as the capac-
lose a substantial amount of heat prior to the rest of
ity if the produced steam is used. In some cases a low
the air being introduced.
pressure waste steam is available and can be used for
While firing fuels contain FBN, the use of air stag-
NOx control with minimum capital costs and operating
ing also helps to reduce conversion of FBN to NOx as
cost, as waste steam has minimum impact on the effi-
the release of fuel nitrogen takes place in a low oxy-
ciency. It may also be justifiable to use steam when it is
gen environment. The use of air-staged combustion
reserved for some infrequent occasions of peak power
usually generates flames with increased volume. Air
production in a very tight regulatory environment.
staging implementation and potential effectiveness
Some modern burners can utilize steam much more
depends on the space available for the fuel oxidation
effectively by using selective injection points rather
to be completed. Possible problems include increased
than just injecting it into the air, with an increase of its
CO and UHC emissions and potential flame stability
effectiveness by a factor of two or more. However, this
problems as the burner becomes more susceptible to
method usually involves modifications to the burner,
the deviations from optimum operating conditions.
negating to some degree the advantage of its low capi-
Practically, air staging can be achieved with redis-
tal costs.
tribution of the fuel within a burner by creating a
Water injection into the flame is even more prob-
very coarse mixing pattern using fewer fuel injectors,
lematic than the injection of steam. First of all, water
injecting fuel into the flame zone with low momen-
needs to be very finely atomized to below 100 μm
tum or directing fuel to mix differently in specific flow
particles to be effective, and that requires using com-
streams. Multiple air passages and some means of con-
pressed air and special atomizing nozzles. Secondly,
trolling air distribution to those passages can also be
its injection has to be properly distributed throughout
used to achieve the desired effects of air staging. These
the flame body but in a way not to impact flame stabil-
techniques were the first embraced by early combus-
ity. Lastly, all this water in the flame will be converted
tion technologies for controlling the NOx, as it required
to water vapors again, substantially degrading the
relatively minor changes to the combustion systems.
overall efficiency of the process on the order of ~50%
On the air side, the staging can be achieved by impos-
to 75% of the efficiency loss with the injection of the
ing minimum intensity swirling motion (generating
produced steam.
flows with low swirl numbers as defined by Ref. [5]),
The ultimate ability of mixing air with inert-for-com-
or by introducing a portion of air through ports spaced
bustion gases depends on the burner. Conventional
away from the burner.
burners designed for the use of FGR can utilize up to
Air staging alone can typically reduce the NOx by
20% FGR when firing natural gas and a somewhat lesser
20%–60%. Any deeper reduction may become problem-
amount up to ~15% when firing oil, and achieve any-
atic and not reliable. Air-staged burners are usually poor
where from 40% to 75% reduction in the NOx when
candidates for operation with FGR as the combination of
compared with the baseline (no FGR) level. As will
fuel-rich conditions with reduced oxygen results in mar-
be discussed further in this chapter, advanced burner
ginal flammability for most hydrocarbon fuels.
designs can utilize much more FGR.
14.3.3 Fuel Staging large burners the premixing needs to be accomplished

by the burner. To differentiate large premix burners
Fuel staging (see Figure 10.20) requires injecting a por-
from small burners where premixing may take place in
tion of the fuel away from or completely out of the flow
the duct or a pipe upstream of the burner, large premix
of the combustion air. This portion of the fuel is termed
burners are designed differently and are sometimes
staged fuel or secondary fuel while the other portion is
called simulated premixed combustion burners.
referred to as primary fuel. For the fuel staging tech-
In order to achieve the benefits of thermal NOx reduc-
nique to be effective, the secondary fuel needs to enter
tion by premixed combustion over the alternative ways
the combustion zone at some distance from the burner
of controlling the NOx, the mixture of fuel and air
exit cross section after the flame created by primary fuel
requires substantial excess air and/or inert gas like FGR.
loses a substantial amount of heat by radiation and con-
When firing natural gas, for example, the substantial
vection to the boiler water walls. The staged fuel may
benefits of premixed combustion on the NOx are real-
also be injected in a way to bring into the combustion
ized only if the excess air levels are over ~40%. In com-
zone products of combustion from the furnace volume
bination with bulk mixed air and FGR, the excess air
surrounding the flame. When this volume contains low
may be reduced to “normal” levels of 10%–15% or even
oxygen gas at a temperature much lower than the flame,
less with full benefits of the lean premixed combustion.
the effect of the furnace gas re-entrainment into the
Turbulent premixed flames are prone to generating
flame on the NOx will be somewhat similar to the use
intense combustion instabilities with frequencies defined
of selective FGR.
by the burner and the overall system. Only specially
Burners with fuel staging are quite different from con-
designed burners, with some of them described further
ventional burners, and burners with air staging as fuel-
in this chapter, along with accurate controls allow sat-
staged burners are equipped with fuel injectors external
isfactory operation on a large industrial scale with sig-
to the combustion air passage. Fuel-staged burners, if
nificant turndown. The key for reducing premix flames
improperly designed, may be susceptible to partial loss
instabilities is diffusing—stretching the flame front
of flame and combustion instabilities, as the location
either with increased velocity and turbulence of the flow
of the flame front and ignition areas for the staged fuel
and/or with reduction of the oxidation kinetics with high
may be shifting by the highly turbulent combustion
amounts of the excess air or flue gas. Some other tech-
process. Fuel-staged burners can reduce NOx emissions
niques incorporated into the burner design may help to
by up to 50% and can be designed for accepting moder-
improve operation by desynchronizing the combustion
ate amounts of FGR for a deeper NOx reduction up to
instabilities developed in different parts of the flame.13
~80% overall.
Premixed burners are typically designed for operation
with NOx emissions below 15 ppm (corr. 3% stack O2,
dry) with the most advanced boiler burners currently
14.3.4 Lean Premixed Combustion
operating in selected equipment with sub-5 ppm NOx
Lean premixed combustion applied to natural gas can and low excess air. A typical relation between the rates
reduce prompt NOx to below 2 ppm. With high excess air of FGR, excess air, and NOx for a medium size package
or FGR the technique is capable of reducing thermal NOx boiler equipped with a premixed burner is shown in
almost to the level of prompt NOx formation. Lean pre- Figure 14.313 with a series of iso-NOx curves.
mixing technology was developed to reduce prompt NOx When lean premixed combustion is substantially
and operate with higher levels of FGR at the same time. diluted with FGR or excess air to the level below 10 ppm,
The simple delivery of combustion air to the burner the effect of the NOx formation beyond the flame zone
in excess of the stoichiometric amount alone does not becomes significant. Figure 14.4 shows NOx emissions
make the combustion process fuel lean. As the fuel is measured when firing lean premixed burners with
injected into the air flow, the mixing process always a substantial amount of FGR and about 20% of excess
generates fuel-rich mixtures first, and that is where air recorded during the commissioning of QLA burn-
combustion easily starts with any source of ignition. ers in 1999.14 The fuel in this case was natural gas and
In order to achieve oxidation of fuel under lean con- the combustion air was preheated to 425°F (218°C). The
ditions throughout the flame, the fuel and air need to 60,000 lb/h (27,000 kg/h) CE boiler furnace was sized
be premixed with a fuel-lean blend prior to ignition. for about 1 s residence time and had ~50% refractory
Premixing of the fuel and air upstream of the burner for coverage of the flame surrounding walls. The two burn-
industrial size burners would require large volumes of ers were generating relatively short flames not reaching
flammable mixture that would be very difficult to han- the furnace back wall. NOx, CO, O2, and combustibles
dle with adequate safety and would present significant emissions were measured in proximity to the boiler
issues with respect to controlling any leakages. So, in back wall at locations about 10 ft (3 m) opposite to the
Theoretical flammability limit also recorded when testing lean premixed burners at the
60 company test facility in a furnace with minimum refrac-
Ultra low NOx burner stability limit tory coverage. This level of post-flame NOx production is
50 impossible to explain based only on the Zeldovich NOx
rate approximately shown by Figure 14.1. One explana-
ULN burner operating range tion to the observed effect suggested by S. Londerville is
40 with NOx < 9 ppm (corr. 3% O2)
and CO < 50 ppm the conversion of some amount of N2O in the post-flame
2 zone to NO. The formation of N2O during the combus-
FGR (%)
5
30 7 tion tests is seldom measured, as this species is not part
9
12
of the NOx that is regulated.
20 15 No practical concepts exist yet for creating the equiv-
alent of premixed combustion for fuels delivered to a
10 burner in a liquid form.
Numbers at the curves indicate NOx.
ppm corr. to 3% O2
0
0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
14.3.5 Furnace Gas Circulation and
Combining Different Techniques
FIGURE 14.3
NOx performance of a typical Coen premix burner firing natural gas
This technique of NOx reduction uses burner-created
in a package boiler. Numbers on the chart are NOx emissions at 3% O2 aerodynamic effects to return within the furnace a por-
(dry). (From Lifshits, V., Ultra low NOx burner with widened stability tion of combustion products that already lost some heat to
limits, AFRC/JFRC International Symposium, Maui, HI, 1998.) the walls back into the combustion zone. The staged-fuel
jets as described earlier can create this effect. The desired
burners and at the stack using well-calibrated and accu- patterns of furnace gas circulation (FGC) can also be pro-
rate instruments. The NOx was measured using a che- moted by the combustion air momentum. The technique
miluminescent NOx analyzer. An interesting fact here is is always used in combination with fuel staging, achiev-
that almost 50% of the NOx ranging from 3 to 5 ppm was ing NOx reduction in some cases up to 50%–70% when
produced in the post-flame volume of the furnace after firing natural gas and somewhat lesser amount of ~20%
the combustion was complete as proven by very low CO when firing light oil with low nitrogen content. The dif-
readings at the back of the furnace. Similar data of post- ficulty of applying the FGC technique is the dependence
flame production of NOx on the order of 1.5–2 ppm were of its effectiveness on the furnace geometry. Deeper NOx
14
NOx at stack NOx at boiler back wall
12
10
NOx, ppm corr. to 3% O2
0
2450 2500 2550 2600 2650 2700 2750 2800
Adiabatic flame temperature without dissociation, °F
FIGURE 14.4
Coen QLA burner performance at high fire.
reductions can be achieved when burner technology Package boilers are also widely used in marine applica-
effectively utilizes a combination of fuel staging, FGC, tions (see Chapter 17). Marine package boilers are usu-
FGR, and elements of premixed combustion together ally more compact than stationary boilers, fire-only oil,
approaching or exceeding the best performance of the and require very compact flames. These considerations
premixed combustion devices. are important when determining characteristic residence
time of species inside the flame affecting the maximum
allowable size of the evaporating fuel droplets.
Knowing the residence time and furnace pressure is
also important for estimating the kinetic/mixing energy
necessary to be transferred by the burner to the reacting
14.4 Burners for Package Boilers
gases entering the combustion zone.
Package boilers by their definition are substantially pre- Knowing the type of furnace walls, membrane or tan-
fabricated and shop assembled systems with a maximum gent type, has some relevance to the type of burner to be
size, allowing their shipment by some common means selected and regime of its operation. Tangent wall boilers
of transportation and with a minimum field installation are prone to short circuiting—the effects of a small por-
cost. With rare exceptions these boilers are designed for a tion of furnace gas surrounding the flame to pass between
single burner. The thermal capacity of burners varies from the tangent tubes into the convection section carrying
about 30 × 106 Btu/h to 350 × 106 Btu/h (~9 to 100 MW), with it a small amount of unburned fuel (UHC) and CO.
which corresponds to about 25,000–250,000 lb/h (11– Controlling the UHC and CO emissions from tangent tube
113 m-ton/h) of steam production. There are several pack- boilers often requires operation with elevated levels of
age boiler design configurations with respect to the layout excess air, helping to reduce the concentration of UHC and
of the drums, tubes, and radiant and convection section CO reaching the tangent tube walls and selection of burn-
of the boiler that were described earlier in Chapter 1. ers generating more compact flames—typically not very
Package boilers may produce a relatively low pressure low NOx burners unless of the premixed combustion type.
saturated steam at 150–250 psig (10–17 barg) pressure, or Firing into the relatively limited space of a package
high pressure superheated steam of about 600–1500 psig boiler often raises concerns regarding actual or per-
(~40 to 100 barg) for driving steam turbines. ceived flame impingement on the boiler walls. The issue
The boilers are typically designed for burning different is important here because it is somewhat subjective and
gaseous fuels, oil, and on occasion some fine coal often a source of debate when something goes wrong. The
or biomass. The radiant part of the boiler has a close-to- impingement is really defined only by its consequences.
parallelepiped shape with the width varying from 5.5 ft The main concern of flame impingement is creating
to about 12 ft (1.5 to 3 m), height varying from 7 to 12 ft such a high heat flux onto the tube surface that the boil-
(2 to 3.5 m), and the length from 15 to 40 ft (4.5 to 12 m). ing process inside the tube starts developing a film of
In modern boilers, the furnace walls and their tops are steam on the internal tube surface, impeding tube cool-
usually formed by sheets of tubes, each with two fins ing with otherwise two phase steam–water flow. In boil-
welded together forming continuous membrane walls. ers with properly designed water circulation and good
Older types of boilers were designed with walls formed water quality, however, soft impingement of visible
by the tubes placed side by side without gaps and not flame tails onto the tubes cannot create a strong enough
connected by welding—tangent tubes walls. heat flux that could cause any problems.
The furnace space heat release at capacity varies from The other and more often occurring result of the flame
~40,000 to ~120,000 Btu/(h-ft3) (400 to 1200 kW/m3), with being in contact with the tubes is soot deposition on the
the lower numbers being more characteristic for small walls. The soot can be formed when products of incom-
boilers. These numbers can be approximately converted to plete combustion containing carbon atoms or aggregates
average residence time of species passing through the fur- get in contact with the relatively low temperature tubes.
nace in the range from 0.5 to 1.0 s when fired at capacity. The buildup of soot on the surfaces is usually limited, as
However, the actual residence time of a substantial por- soot has a low thermal conductivity; so the conduction
tion of the flow moving through the furnace will be much heat transfer is low and the surface of a thick soot layer
higher than average velocities resulting in lower residence acquires a higher temperature that causes the soot to
times for this portion. The heat release per unit of furnace oxidize. The effect, however, is still unwanted as it will
cross section varies from 0.65 to 3.0 × 106 Btu/(h-ft2) (about interfere with the overall heat transfer from the flame to
2 to 6 MW/m2). Boiler furnaces usually operate at positive the water tubes and may create an accumulation of soot
pressure defined by the hydraulic resistance of the boiler that will fall off onto the furnace floor.
convection section, including the economizer. The range The situation is different when tubes of the superheater
of furnace pressure at peak loads varies from as little as experience impingement. The superheated steam does
3 in. WC (0.75 kPa) to as high as 1 psig (7 kPa). not have the same cooling intensity as the two phase flow
even when it moves with high velocity. The superheater To accomplish all of these goals, the design resulted in
tubes may actually get overheated even without flame burners that mix air and fuel and ignite the mixture as
impingement by fully oxidized high temperature combus- rapidly as possible to create a highly stable and compact
tion products that still emit substantial thermal radiation flame. A strong swirling motion imposed on the flow
that visually can be mistaken as flame impingement. More facilitated the mixing and compactness of the flame. For
often burners making longer flames or burners operating operation on gaseous fuels, the fuel could be injected
with FGR (higher mass flow flames) will generate higher through a gas ring surrounding the burner flow exit-
temperatures in the area of the superheater at the back of ing cross section at the entrance to the flared refractory
the furnace. Burner suppliers need to be aware if there are throat. The gas rings were drilled with a series of holes
some specific requirements for the temperature in the area grouped in small clusters injecting fuel in the direc-
of the superheater and whether the boiler is equipped with tion of about 45°–60° to the burner centerline. Other
the proper means of controlling it. Problems like this usu- gas injection methods were also used. For oil firing, the
ally occur when an old burner needs to be replaced with burner was equipped with an atomizer placed through
a new low NOx burner using FGR, or when a new burner the center of the flame-stabilizing shield.
generates a more transparent (lower luminosity) flame. Register-type swirling devices and some spinners
often had a tendency to produce an overly excessive
swirling motion at the burner centerline and starve this
14.4.1 Conventional Round Burners
region on air or create backflows. This effect was unde-
and Burners with Air Staging
sirable, especially for oil firing when fuel was injected
Round burners are the most common burners histori- from the point close to the burner centerline causing
cally used in package boilers. The term refers to the oil gun coking and other problems. To avoid the effects
burners with a predominantly round shape of the com- of over-swirling, some burners were designed with a
bustion air passage or several concentric air passages small air zone surrounding the atomizer that did not
for combustion air through the burner. The burners are have a swirling motion.
typically equipped with one or two sets of adjustable The overall swirling and burning fuel–air mixture
or fixed register louvers placed at the air inlet into the typically exits the burner through a conical refractory
burner for imposing a swirling motion to the combustion throat with a length of about 50% of the burner diam-
air and a bluff body or another swirler/spinner placed at eter. This divergent throat helps to further facilitate the
the burner centerline close to its exiting plane. With all development of a large size recirculation zone in the
these features, the air flow pattern through the burner wake of the spinner delivering hot combustion prod-
can be described predominantly as two-dimensional ucts to the area of ignition at a location close to the fuel
with all velocity vectors being a function of only radial injection. The glowing parts of the throat also provided
and axial coordinates, but not the angular/circumference a source of ignition for the peripheral parts of the flow.
coordinate. The presence of the louvers or spinner blades, Coen DAZ™—dual air zones burner schematically
or slots in the flame-stabilizing shields creates some shown in Figure 14.5 is an example of such a burner. The
patterns of periodic disturbance over the flow circumfer- burner has two sets of adjustable register-type louvers
ence, but these patterns are usually weak, fast dissipat- typically set for spinning the flow in opposite direc-
ing, and not always essential for the burner operation. tions, achieving very intense stable flames of the nec-
The very first round burner designs had a predomi- essary shape, operating with a wide range of air–fuel
nantly uniform distribution of fuel over the burner cir- ratios, making it a perfect burner at the time for many
cumference and were designed for high stability and applications including marine and package boilers.
flame intensity. The burners trace their lineage back to When the burner was not in use, both sets of louvers
the days of ship-mounted boilers, where boilers had to could be closed to reduce the draft through the burner.
be kept as small and light as possible, thus making the Applying the burner to stationary boilers firing natu-
requirement for short flames very important. Without ral gas with reduced NOx requirements was found to
these boilers in operation, a ship could be left dead in be difficult as the burner generates relatively high NOx
the water, suggesting the need for the burners to reliably emissions of about 70–160 ppm (ref. 3% O2 dry) on natu-
operate regardless of changes to the fuel or air supply. ral gas and its ability to tolerate the FGR was very poor.
The maneuvering demands of the ship also require that The problem stemmed from the pattern of gas injection
these burners be able to operate across a wide range of from the periphery of the air flow.
firing rates and be able to operate within a wide range of A new class of low NOx round burners appeared
air to fuel ratios essential for maintaining good flame then as a first response to the demand of simplify-
stability even with the very crude setting of single point ing the burner design that also allowed operation
positioning combustion controls and deviations in fuel with reduced NOx emissions on gas and oil in boil-
pressure etc. explained in Chapter 12. ers. These burners were equipped with multiple fuel
Throat
Gas ring
Pilot
Flat blade stabilizer and

oil atomizer
FIGURE 14.5
Coen DAZ™ burner.
gas injectors positioned around the spinner or a bluff Figure 14.7 shows another air-staged burner with
body providing patterns of fuel gas injection ben- both types of fuel injectors placed around the spinner—
eficial for operation with FGR, or patterns delivering DFL® burner. On the air side, the burner design includes
delayed mixing with air—air staging. The following multiple concentric passages, some with fixed turning
are a few examples of modern air-staged low NOx vanes, and a concentric bluff body filled with ceramic.
burners for package boilers designed to fire both gas- When properly applied, the burner generates low base-
eous and liquid fuels. line NOx and is capable of operating with low excess air
A Variflame™ burner shown in Figure 14.6 features and low CO emissions.
a venturi-shaped passage for the air flow, providing Figure 14.8 shows a picture of one of the most versa-
a well-balanced air flow pattern just upstream of the tile air-staged burners for a package boiler—the Coen
combustion zone, and two sets of gas fuel injectors DAF™ burner. The burner is mechanically more com-
connected to the same fuel plenum. The first smaller plicated than the Variflame burner. It has two major
set delivers fuel directly into the wake of a large spin- air passages and the ability for external adjustments
ner, developing where the overall flame is stabilized. to the air distribution between the passages. It has a
The second main set placed adjacent to the spinner set of externally adjustable louvers—register that con-
provides fast initial mixing of fuel and air at each trols the swirling motion imposed onto the air flow in
injection point and delayed mixing further in the main either the clockwise or counter clockwise direction. A
body of the flame. The number of main fuel injec- set of fuel injectors around the spinner is connected to
tors varies from 3 to 8, depending on the application. the external or internal gas ring header. If the burner
Another feature of this burner is that a small portion needs to fire multiple gas fuels separately or in com-
of combustion air can be supplied around the venturi bination, additional sets of fuel injectors and ring gas
passage. The amount of this flow can be adjusted. This headers are added. The fuel injectors deliver a small
flow diversion around the venturi has a strong effect portion of fuel into the wake of the spinner while most
on the burner draft loss and the overall shape of the of the fuel is delivered in several lobes to gradually mix
flame. The burner is simple and efficient and can be with the combustion air. The drilling pattern of all fuel
optimized for operation with FGR. injectors is usually customized for each application.
FIGURE 14.6
Coen Variflame™ burner. (From Baukal, C.E., Ed., The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
the FGR rates on natural gas up to 20% with the NOx

Burner design emissions reduced to 25–30 ppm level. When firing oil,
the maximum amount of FGR is limited to 15% with a
Division corresponding reduction in thermal NOx of about 65%.
rolling Main gas
nozzle
14.4.2 Burners with Fuel Staging
Low NOx
baffles Figures 14.9 and 14.10 provide examples of fuel-
Pilot gas
nozzle staged burners: the Coen Delta NOx™ burner and the
Swirler Hamworthy ECOjet® burner. In both burners, the com-
bustion air is delivered through a single venturi-shaped
passage and the staged fuel is injected into the flame
from a series of fuel injectors placed outside of the
diverging refractory throat.
DFL burner
In the Delta NOx burner, the primary fuel is deliv-
ered through a series of small gas headers—spuds—
FIGURE 14.7 placed in radial directions around a centrally located
Hamworthy DFL® burner. spinner or a combination of a bluff body and a spinner.
Fuel injection from each spud is distributed along the
The available substantial adjustments to the air flow radius. The Delta NOx burner ignition points are sta-
allow effective additional control over the flame pat- bilized by the action of the spinner or the bluff body
tern as well as the ability to de-tune the operation from and by the fuel jets arrangement that generates flame
some unwanted acoustical interactions with the system. patterns sufficient to ignite the staged fuel at some dis-
The burner can also operate with good stability with tance from the burner refractory throat. When firing
FIGURE 14.8
Coen DAF™ burner.
In the ECOjet burner, all of the primary fuel is deliv-

ered through a center-fired gas gun. The primary fuel
forms a highly stable primary flame zone. The flame
is stabilized by the action of the spinner. The primary
zone is large enough to provide ignition of the staged-
fuel combustion zone. Figure 14.11 shows the burner
flame when operating on natural gas without the FGR.
Both burner designs are very simple and are usually
designed with an external means of adjusting the air
distribution between the zones. The NOx performance
of the ECOjet burner is slightly better than that of the
Delta NOx burner but less forgiving to deviations in the
excess air ratio. In some cases the ECOjet burner perfor-
mance on natural gas can achieve 10–12 ppm NOx.
14.4.3 Burners with Partial Lean Premix Combustion

A description of burners utilizing partial lean premixed
combustion is given here using the Coen QLN™ burner
as a classic example. Partial lean premixed combustion
FIGURE 14.9
Coen Delta NOx™ burner.
is fundamentally different than partial premixed fuel-
rich combustion, the most common example of which
natural gas without FGR, the burner delivers NOx per- is utilized in gas burners for all cooking ranges. The
formance in the range of 45–60 ppm, depending on burner is schematically shown in Figure 14.12. The
the relative furnace to the heat input size. With about burner was the first in the industry to generate simu-
10% the FGR emissions can be typically reduced by lated lean premixed combustion in a substantial part
about 45%. of the flame. The burner was developed in the early
FIGURE 14.10
Hamworthy ECOjet® gas-only burners.
FIGURE 14.11
Hamworthy ECOjet® flame. Natural gas firing at 30 MW (100 × 106 Btu/h) at Hamworthy Test Facility.
1990s15–17 with a goal to efficiently fire natural gas in uniform. Another departure from convention at the
package boilers with less than 30 ppm NOx without time was the injection of fuel into the middle of the air
using FGR. It had several new revolutionary features. flow streams at a substantial distance from the possible
Visually, the main unique feature is an air distribu- area of fuel ignition to form a uniform very fuel-lean
tion plate or a number of bluff bodies shaping the mixture prior to its possible sources of ignition. It was
combustion air flow in a series of radially aligned pas- known at the time that such premixing would result in
sages separated by bluff bodies and spaced uniformly a high intensity acoustic response; however, the design
around another centrally located bluff body, whereas of the QLN burner solved the problem by desynchro-
in all prior burners for package boilers the air flow exit- nizing these instabilities and reducing them to well
ing the burner was predominantly circumferentially below the would-be critical levels.
Premixed
gas
Core
gas
Throat
Staged gas
FIGURE 14.12
Schematic of a Coen QLN™ burner.
The burner has three zones of fuel delivery: the flame FIGURE 14.13
stabilization zone that burns a small portion of fuel in Coen QLN burner flame with 20 ppm NOx -firing natural gas with-
the wake of the bluff bodies, the premixed zone, and out FGR.
the staged-fuel zone that gets the majority of its fuel
from injectors placed through the ports of the short staged-fuel injectors. A typical appearance of a QLN
straight cylindrical refractory throat. Oxidation of the flame is shown in Figure 14.13.
premixed fuel and staged fuel takes place gradually QLN burners can also fire oil. The NOx emissions on
as premixed fuel and air mix with hot combustion oil are also reduced due to the effects of FGC enhanced
products from the flame stabilization zone. Oxidation by the star-shaped pattern of air injection. However,
of the staged fuel also takes place very gradually as it NOx reduction with the atomizer design by creating the
is first mixed with medium temperature low oxygen air staging effects is reduced as only a narrow range of
combustion products around the flame (FGC effect atomizer designs can be applied with the burner.
described in Section 14.3.5) and then mixes with the Over almost 20 years of its use to date, the QLN was
remaining oxygen of the products of premixed com- successfully adapted for applications with multiple
bustion. In the QLN flame, most of the fuel burns at fuels firing simultaneously or one at a time proving
locally fuel lean or close to stoichiometric conditions, to be very versatile and highly tunable to the specifics
minimizing the production of prompt NOx. The ther- of the application. The product always delivered the
mal NOx is also substantially reduced due to very expected low NOx performance. QLN burners utiliz-
fuel-lean conditions for burning the premixed fuel ing small amounts of forced FGR can operate with NOx
and FGC effect created by the staged fuel. emissions down to a 15 ppm level in small or medium
The NOx emissions with QLN burners depend on the size package boilers and in larger package boilers with
furnace size and the burner peak heat input. In small up to 250,000 lb/h (113,000 kg/h) steam production if
boilers rated to less than 75,000 lb/h (34,000 kg/h) of the furnace of the boiler is slightly oversized.
steam production (~32 ton/h or 29 m-ton/hr) firing natu-
ral gas QLN burners achieve typically less than 25 ppm
14.4.4 Premixed Burners
of NOx with low excess air and no FGR. QLN burners
in larger boilers with a capacity of up to 100,000 lb/h Burners with all premixed and predominantly pre-
(45 ton/h or 45,000 kg/h) of steam emit higher NOx up to mixed combustion were developed in the middle of
about 40–45 ppm. the 1990s. The first burner of this kind utilized rapid
A small amount of induced FGR can be used in QLN mixing of filtered fuel gas passing through hundreds
burners to reduce NOx emissions by about 20%–25%. of tiny laser punched orifices uniformly distributed
Better results, down to 15 ppm NOx, can be achieved through the flow of combustion air mixed with flue
with forced FGR delivered to the plenum around gas. The burner name “the RMB™ burner” is an abbre-
the burner and injected into the furnace around the viation for rapid mix burner. The fuel distribution in the
FIGURE 14.14
Coen RMB™ burner equipped with an air isolation sliding barrel damper.
RMB was accomplished using a multitude of radially

placed airfoil-shaped miniature fuel gas risers impos-
ing a swirling action onto the combined flow of air fuel
and flue gas.18 The flame was stabilized by a strong recir-
culating flow of hot combustion products developed as
the result of swirling flow in combination with a diverg-
ing refractory throat and another refractory bluff body
at the burner centerline. Figure 14.14 shows an overall
view of the burner. Figure 14.15 shows multiple fuel ris-
ers. Figure 14.16 shows enhanced photographs of the
flame.
The burner produces a relatively compact and short
flame. The burner operation can be tuned for low
excess air and NOx as low as 5 ppm. With low excess
air the burner uses substantial amounts of FGR.
During normal operation, the burner does not gener-
ate any appreciable CO. The burner is also capable of
firing liquid fuels.
With the design described earlier, the burner was
developed in several sizes ranging in thermal capacity
from 50 to 100 × 106 Btu/h (15 to 30 MW). For higher
capacities up to 300 × 106 Btu/h (~88 MW) the RMB con-
cept was expanded with a second similarly premixed FIGURE 14.15
fuel zone placed around the primary RMB burner Fuel risers of a Coen RMB burner. (From Baukal, C.E., The John Zink
together making D-RMB™—dual rapid mix burner— Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Figure 14.17. In order to limit the flame diameter, the
second zone was designed for a very low swirl number. motion of the combined fuel, air, and FGR blend.13,19
The outer premix zone is ignited from the primary RMB This burner was designed with a variable geometry fea-
burner. The D-RMB burners were able to control the ture—the area of the burner where the premixed blend
NOx emissions down to 7–9 ppm. The turndown of both enters the combustion zone is modulated with the fir-
burners was in the range of up to about 6–1. Figure 14.18 ing rate. The primary means of flame stabilization is
shows an example of a large D-RMB performance.2 a massive bluff body with a specially designed spin-
Another premixed burner, the QLA burner, utilizes ner on its periphery that desynchronizes combustion
primarily fuel gas jets to generate a weak swirling instabilities. The burner relies on an increased mixing
typically deliver NOx performance down to 6–8 ppm

NOx with capacities as high as 320 × 106 Btu/h (~94 MW).
The burner can operate with preheated air making NOx
emissions as low as 5 ppm14 and operation with ambient
temperature or preheated air. Design modifications of
the burner were incorporated to fire light oil as a backup
fuel. Figures 14.19 through 14.22 show a schematic of the
burner, pictures of the burner assembly, and a close-up
photograph of the flame.
14.4.5 Burners for Low Heating Value Gaseous Fuels

and Fuels with Highly Variable Composition
Firing fuels with low heating value (low Btu gases) pres-
ents some difficulties due to the high percentage of inerts
in the combustion gases, high moisture content, etc.,
making the gas mixture poorly flammable or not flam-
mable at all. These fuels are usually byproducts of some
processes and come to the burner with very low pressure
ranging from a few inches of water column to a few psig
(10–100 mbarg) and vary in composition. Such fuels can
be effectively utilized in boilers with the help of small
amounts of more flammable fuels like natural gas or oil
and a scroll burner. Figure 14.23 shows a picture of one of
such burners designed for firing blast furnace gas (BFG).
Low Btu gases naturally produce very low NOx emis-
sions. Firing low Btu gases that may change in com-
position to a more flammable gas presents additional
challenges that when combined with low NOx emis-
sions requirements are difficult to resolve. When such
fuels are introduced into conventional burners with
fixed fuel nozzles, the changing composition would
not create consistent mixing patterns. This along with
inconsistent flammability characteristics of the mixture
can easily result in unreliable flame stabilization and
combustion instabilities. The problem can be solved
with burners mixing fuel with flue gas in quantities,
making the overall blend of fuel and FGR more con-
sistent from the combustion stand point. Figure 14.24
shows a picture of such a burner. In this particular case,
FIGURE 14.16
the burners were even more complicated to accommo-
Enhanced images of gas-fired RMB flame. (From Baukal, C.E., date two separate fuel streams varying in proportion to
The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) each other. The burner was able to reliably operate with
less than 15 ppm NOx (~31 mg/Nm3), referenced to 3%
distance between the fuel jets and air, making it possi- O2 (vol, dry) firing a wide variety of fuels.
ble to keep minimum fuel orifice sizes relatively large, Mixing FGR with fuel is usually more effective than
in excess of 1/8 in. (~3 mm) in diameter. At reduced mixing FGR with air in terms of the NOx reduction for a
loads, the action of the spinner enhances as a higher given amount of FGR, especially at lower FGR levels as
portion of the flow passes through the spinner blades shown in Figure 14.25.
due to the variable geometry. The variable geometry The higher effectiveness of fuel mixed with FGR of
and presence of smaller auxiliary fuel zones, one in the about 20% can be explained by the fact that the full
wake of the bluff body and one in a staged-fuel zone, amount of flue gas is engaged in the combustion at the
allowed better control over the combustion instabili- very moment of fuel ignition rather than being spread
ties at reduced loads and overall better turndown of over the entire flow including the excess air. The rest
up to 10–1 and higher in some cases. The burner can of the difference is due to a more favorable fuel and
FIGURE 14.17
Coen D-RMB® burner mounted inside the windbox. (From Baukal, C.E., The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
230,000 LB/HR nebraska “A” type package boiler

40
35
30
FGR, NOx, CO, and O2
25 NOx, ppm
CO, ppm
20
FGR, %
15 O2, %
10
0
20 30 40 50 60 70 80 90
Load %
FIGURE 14.18
Coen D-RMB™ burner performance in a large package boiler. (From Baukal, C.E., The John Zink Combustion Handbook, CRC Press, Boca Raton,
FL, 2001.)
air mixing pattern with fuel mixed FGR achieved with 14.4.6 Oil Atomizers for Package Boilers
a specific burner for lower NOx control. For another
For firing liquid fuels, burners for package boilers are
burner, a further increase in the FGR rate may cause the
usually equipped with internal mix steam-assisted
mixing pattern to deviate from the optimum and slow
atomizers generating fine droplets with a Sauter
down the NOx reduction with fuel mixed FGR. This is
mean diameter (SMD) of 40–120 μm. In the industry,
also shown in Figure 14.25 (see data taken when firing a
oil atomizers are also called oil guns. The definition
small Coen CLN burner).
oil taking place in multiple passages where steam

and oil combine. An example of the first type is the
Coen MVI atomizer. An example of the second type
is the Y-jet atomizer. For many package boiler appli-
cations the preferred atomizer type is similar to an
MVI®. The selection of this type of atomizer for many
boilers is preferred as it allows substantial latitude in
making the desired pattern of fuel injection as well
as easy modifications to the flame pattern and per-
formance by replacing a relatively inexpensive atom-
izer tip. Often problems that can be resolved with the
cap drilling are carbon buildup on the caps that may
interfere with the oil spray, collapsing of the flame in
the middle of the oil spray preventing its adequate
mixing with the air, oil impingement on the burner
throat causing deterioration of the throat refractory
and impingement on the boiler sidewalls causing car-
FIGURE 14.19 bon buildup on the boiler tubes.
Coen QLA burner schematic. The amount of atomizing steam and the atomizer
design details need to be accurately tailored to achieve
the desired atomization quality while minimizing the
of SMD, the description of the internal mix atom- usage of these utilities. At high fire operation, the atom-
izer concepts, and examples of industrial atomizer izers in package boilers usually consume steam in an
designs were reviewed earlier in Chapter 5. The same amount of 10%–18% of the mass of fuel oil. Internal
chapter also describes two types of internally mixed mix atomizers typically require 100–200 psig (7–14 bar)
atomizers: one with a common internal chamber for pressure of oil at high fire. The steam pressure require-
the primary stage of atomization and the other type ments are usually slightly higher than those of oil by
with a single stage atomization between steam and about 15 psi (1 bar). Having the steam pressure higher
FIGURE 14.20
Assembled Coen QLA burner (side).
FIGURE 14.21
Assembled Coen QLA burner (front).
FIGURE 14.23
Peabody LVC™ burner for firing BFG.
change in the drilling angles and patterns creates a

substantial change in the flame appearance and the
burner/atomizer performance.
14.5 Burners for Enhanced Oil

Recovery Boilers
Extraction of heavy oil in some oil fields requires in situ
heating of bituminous oil deposits at the source deep
underground in order to make the oil flow. The heat-
FIGURE 14.22 ing is done by injection of high pressure wet steam pro-
Coen QLA burner flame with 7 ppm NOx (natural gas firing). duced by special boilers. The amount of steam used for
enhancing the oil recovery is massive, ranging from to 2
than the oil pressure assures that oil cannot penetrate to 3.5 times the amount of oil produced. This is equiva-
into the steam passages even in cases of passages plug- lent to using 10%–18% of the thermal energy of natu-
ging. Increasing the amount of atomizing steam reduces ral gas for oil reservoirs heating relative to the thermal
droplet SMDs and shortens the flame. energy of mined oil.
As the steam adds substantially to the combined These oil fields usually do not have large supplies of
momentum of steam-fuel jets, varying the amount fresh water, making it necessary to recycle all the water
of steam becomes a powerful tool for optimizing the that comes back out with the oil. This return water car-
mixing patterns between fuel and air and the overall ries a high concentration of dissolved solids that is not
burner operation and flame shape. Often, however, practical to completely remove before using it again as
there are trade-offs between the fineness of atomi- boiler feedwater.
zation and NOx emissions as increased fineness of The steam for enhanced oil recovery is mostly gen-
atomization may reduce the air-staging effects and erated by specially designed OTSGs, drum-less boilers,
increase the NOx. When designing internal mix atom- that are better suited for operating with such water.
izers, it is important to understand effects of interac- These generators are also often called steam flood gen-
tions between different jets in an overall usually very erators. As two phase steam–water flow develops inside
complex pattern. In many cases, even a few degrees the generator tubes, the concentration of solids in the
FIGURE 14.24
Modified Coen LCF burner for simultaneous low NOx firing of multiple fuels of variable composition.
60%
50%
FGR mixed
40%
NOx reduction with FGR
with air
(generic
30% burner)
Fuel induced
FGR, John
20% zink LCF
burner [2]
10% Fuel induced
FGR, Coen
CLN burner
0%
0% 5% 10% 15%
–10%
FGR
FIGURE 14.25
NOx reduction with FGR mixed with combustion air or fuel.
water phase increases in reversed proportion to the to transfer heat from the flame to the tubes with very
remaining percentage of water phase, making it impor- good uniformity and relatively low peak heat fluxes.
tant to closely control this parameter. For process effec- Excessive heat flux on a tube already carrying only a
tiveness, the produced steam quality (steam content in small amount of liquid phase may cause local deposi-
a two phase flow) needs to be rather high—typically tion of solids on the internal tube surface, subsequently
in the range of 75%–80%. Furthermore, due to the causing tube overheating and failure. With this danger
multi-water pass design of these boilers it is necessary in mind, the radiant section of the generators is sized
substantially larger than package boilers with compa-

rable heat input. The OTSGs are used by the industry in
several regions throughout the world. The vast majority
of units not coupled with gas turbines for electric power
generation have a thermal capacity of 50–350 × 106 Btu/h
(15–100 MW) with corresponding furnace sizes varying
from 7 to 21 ft (2.1 to 6.4 m) in diameter and 25 to 65 ft
(7.5 to 20 m) in length.
A massive amount of fuel consumed by these genera-
tors as well as generator placement in environmentally
sensitive areas has prompted regulatory requirements
to operate the boilers with low and ultra-low NOx emis-
sions. Among the number of low NOx technologies
described earlier in this chapter are burners with par-
tial premixed combustion and fuel staging that have
demonstrated the best performance and necessary
reliability for operation in mostly unattended environ-
ments. For example, natural gas–fired QLN-burners
operating without the use of FGR and with low excess
air can be tuned to produce NOx emissions in the range
of 14–15 ppm to 20–25 ppm NOx, depending on the
generator size and the level of combustion air preheat
used in cold climates. For lower NOx emission, down to
9–15 ppm, the same burners can be set to operate with
some FGR.
Many burners in the field are designed to operate with
natural gas or a combination of natural gas blended with
“produced” gas. The produced gas is essentially a portion
of recycled natural gas that is injected into the oil reser-
voir for being absorbed by the oil to reduce its viscos- FIGURE 14.26
ity. The produced gas carries some moderate amounts of Large Coen QLN burners mounted inside windboxes.
CO2 (~30%–35% vol) and small amounts of other species.
It also contains trace amounts of H2S, making it more
difficult to apply FGR due to the corrosion problems. So, QLN-II burners. The versatility of the burner allows
QLN burners delivering low NOx emissions without the addition of secondary fuel streams of different calo-
use of FGR are especially attractive to industry. Figure rific value that are sometimes available as byproducts
14.26 shows the largest QLN burners built to date for of the oil recovery process. Figure 14.29 shows the rela-
OTSGs. As a reference, the round wind box on the pho- tion between NOx emissions and FGR for QLN-II burn-
tograph is 12 ft (3.6 m) in diameter. ers sized for 50–100 × 106 Btu/h (15–29 MW) heat input
In the most environmentally sensitive areas, regula- operating with low excess air. With high FGR and NOx
tions further require operation of a steam flood genera- of about 3 ppm, the fuel oxidation visually becomes
tor with even lower NOx emissions. For this market, an substantially flameless with very complete combustion
ultra-low NOx burner, QLN™-II burner, was developed (low CO).
that is capable of operating with sub-5 ppm NOx.20,21 In some areas, the steam generators are powered by
The burner concept is an extension of the original QLN syngas produced in the process of bitumen oil mining.
burner technology modified specifically for operation This syngas is mostly composed of CO (~60% vol) and
with FGR. The burner requires FGR for operation. The H2 (~35% vol) and hydrocarbon gases, primarily CH4,
amount of FGR depends on the desired NOx level. For as the balance. For low NOx, burning of this syngas
sub-5 ppm NOx on natural gas, less than 30% FGR gas along with natural gas, a version of a DAF burner
is needed. Figure 14.27 shows a view of the burner operating with FGR is used. Figure 14.30 shows pho-
from inside the furnace. With some modifications, the tographs of the discharge end of the burner. Figure
burner can be configured to fire light oil without any 14.31 shows photographs of the burner flames when
impact on its performance on gas. Figure 14.28 shows a firing natural gas and syngas inside a large steam
row of operating steam flood units, all equipped with generator.
FIGURE 14.27
Coen QLN-II burner.
FIGURE 14.28
Row of boilers equipped with Coen QLN-II™ burners rated to 63 × 106 and 90 × 106 Btu/h (18.5–26 MWt).
designed at the time when larger burners were not

available. Process steam boilers with multiple burners
14.6 Burners for Multiple-Burner Process
are typically equipped with two, four, or six burners
Steam Field-Erected Boilers placed on one of the furnace walls. For higher effi-
When the capacity of a boiler needs to exceed the capac- ciency, many field-erected boilers are equipped with
ity of available package boilers or when the available combustion air heaters delivering air with 300°F–650°F
footprint cannot accommodate a large package boiler, (150°C–340°C) temperature.
the boiler is built for firing with multiple burners. Typically, all burners in service are controlled by a sin-
There are also older existing multiple-burner boilers gle fuel control valve and therefore operate at the same
20
18
16
NOx, ppm ref. 3% O2 (dry)

14
12
10
0
5% 10% 15% 20% 25% 30% 35% 40% 45%
FGR
FIGURE 14.29
Typical NOx performance of QLN-II™ burner with FGR.
FIGURE 14.30
Large 350 × 106 Btu/h (100 MW) Coen DAF™ burner for firing syngas and natural gas.
(a) (b)
FIGURE 14.31
Flames of Coen DAF™ burner firing natural gas (a) and syngas (b) (baking soda was added to the air flow for better visualization of syngas flame).
Heavy duty
Air damper
rollers
cylinder
Gas
spuds
Circular gas
header
Insulated front
plate
Air
Spinner damper
FIGURE 14.32
Schematic of Coen Delta Power™ burner.
heat input and with a common air supply. For multiple- Due to typically lower overall volumetric heat release
burner applications, each burner can be brought in and in field-erected boilers than in package boilers, the
out of service independently, allowing greater flexibil- flame shaping requirements to the burners are less
ity in operating turndown and availability of the boiler stringent. This makes application of a simple Variflame
when any of the burners needs to be serviced. From the burner described earlier in the chapter, or another
burner operation standpoint, the online change in the burner like Delta Power™ burner shown schematically
number of burners in service translates into the require- in Figure 14.32, a common choice. Both burners are sin-
ment to reliably operate during disturbances in the air gle air zone venturi type designed for reduced NOx (air
to fuel ratio created during the changes and the ability staging) and low excess air.
to light off the burners at substantial load and with some Delta Power burners are also designed for a low pres-
flow of FGR if used. Burners for multi-burner boilers are sure drop, making them well suited for retrofits with
usually equipped with air isolation dampers of a slid- higher air preheat temperature or the addition of FGR.
ing barrel type designed to pass only a small amount of The burner is also equipped with an integrated air flow
combustion air in the closed position for cooling. measuring device, allowing troubleshooting and adjust-
When selecting burners for multiple-burner boilers, it ments of the operation in case of some problems with
is important to consider the possibility of heat damage the air distribution. The burner is equipped with an
to the burners out of service (BOOS) from the heat gener- external gas header with cane-type gas spuds (gas tips).
ated by adjacent burners. Other aspects include the need The external gas header makes it possible to replace or
to deliver fuel gas to the burner through a single line so service spuds without dismantling the burner.
that fuel to the individual burner can be easily cut off. More sophisticated lower NOx QLN burners have also
From a performance standpoint, it is important been successfully applied on many field-erected boilers
to know that emissions generated by the boiler will operating with and without preheated air firing natural
depend on the number of burners in operation due to gas, refinery fuels, and oil (as a backup). When using QLN
the heat transfer between the adjacent flames typically burners, some special procedures in the combustion con-
increasing the NOx. The effect is strongly dependent on trols were implemented to reduce the disturbances when
the burner spacing. changing the number of burners in service.
There are a few field-erected boilers equipped with turbine cycle, it is important to maintain the design tem-
two premixed burners. The two burners in these boilers perature of the steam coming out of the superheater and
operate as one burner firing only simultaneously. Two reheater. The excessive temperature in the superheater
burners placed close together on the furnace wall may can be controlled with water injection—attemperation.
form a cell with the added effect of enhancing flame However, attemperation in the reheater is very undesir-
stability of each other due to the flame interactions. able as it changes the ratio of flows in different stages of
This produces lower NOx emissions than with a single the turbine and reduces the cycle efficiency. Conversion
burner—as low as 5 ppm in some cases. of utility boiler operation to gas shifts the heat distribu-
tion to the superheater and the reheater, requiring more
attemperation and loss of efficiency or the inability to
operate at full load due to tube overheating.
The use of forced FGR recirculation back into the fur-
nace through a hopper on the bottom is also a common
14.7 Gas and Oil Burners for Utility Boilers
feature of the boiler serving to control the loss of steam
Steam cycles of fossil fuel–fired utilities have much temperature coming out of the superheater and reheater
lower efficiency of converting thermal power to elec- at partial boiler loads.
tric than combined cycle plants using large stationary When evaluating a multi-burner combustion system,
gas turbines. Practically all oil and gas utility boilers it is important to understand that burners are only one
are either older boilers or coal-fired boilers converted of many elements affecting the system performance.
to gas and oil. The conversions of coal-fired boilers to The overall performance of the combustion system sub-
gas were prompted by concerns of global warming, stantially depends on the design of the fuel distribution
more stringent emission requirements and the increas- and air distribution to the burners. The difficulties in
ing availability of natural gas. Oil firing capabilities designing uniform air and fuel distribution system are
are still considered as an important backup alternative. quite substantial. The relative magnitude of the buoy-
The two most common types of utility boilers are wall- ancy effect on the air distribution and hydrostatic effects
fired boilers and T-fired boilers that use two quite dif- on the pressure of liquid fuel delivered to the burners
ferent combustion concepts. vary with load and often are not properly accounted
for. Addressing potential air distribution problems by
compartmentalizing the windbox and measuring and
14.7.1 Combustion Systems for Oil and
controlling the air flow to individual burners may look
Gas Wall-Fired Utility Boilers
attractive at a glance. However, it often creates more
The wall-fired utility boilers are usually equipped with problems than it solves as the complexity of the system
a number of burners ranging from 6 to 36 that are placed is increased and the reliability reduced. The potential
on one or two opposing walls of the boiler. The furnaces severity of oil maldistribution between the burners
are generously sized with a significant part of the fur- depends on the piping design and on the type of atom-
nace volume above the burner levels. A typical average izers and available margins of the supply pressure.
residence time of the combustion products in the fur- Typically during retrofits, the windbox is modeled and,
nace at high fire is about 1–1.5 s. All boilers operate with if necessary, modified for uniform air distribution.
air preheated to 450°F–650°F (230°C–340°C). Gas- and oil-fired utility burners are usually sized for
Operation of utility boilers is continuously closely the available draft loss typically in the range from 4 to 6 in.
monitored by the operators overseeing the perfor- WC (10–15 mbar). Sizing burners for lower differential
mance. When necessary, the operators can add biases air pressure is helpful for reducing NOx emissions and
to setpoints of different controlled parameters, one of creating more luminous flames, but may also result in
which is excess oxygen at the stack that is normally set longer flames, impingement problems, as well as addi-
in the range of 1%–2%. The boiler controls also moni- tional difficulties in achieving the desired quality of the
tor multiple parameters and alarm operators if anything air distribution.
deviates outside of the normal range.
Most of the wall-fired utility boilers were designed
14.7.1.1 NOx Reduction Techniques
originally for firing coal—a fuel producing high temper-
in Wall-Fired Utility Boilers
ature and very luminous flames. Firing faster burning
fuels like natural gas creates a different temperatures Air staging and FGR are the main NOx reduction tech-
pattern and more transparent flames with a typically niques used in wall-fired utility boilers. The selection
higher relative rate of heat transfer to the steam tubes in of the means of NOx reduction must be considered in
the area of the superheater and reheater and less in the combination with an evaluation of its effect on the boiler
saturated steam tubes zone. For the efficiency of the steam operation in addition to NOx reduction. The air staging
techniques tend to increase the luminosity of the flame simple techniques are called fuel biasing and BOOS.
and may reduce the mostly unwanted shift in the heat The effectiveness of BOOS on the NOx reduction is
distribution toward the superheater when converting very boiler dependent. The optimal pattern of burn-
boilers to fire gas. The addition of FGR may increase this ers with shut-off fuel is usually established experi-
shift toward the superheater. FGR introduced through mentally. The optimum pattern may also depend on
the burners for the purpose of NOx control has a strong the boiler firing rate and availability of the burners
effect on the thermal NOx and a strong effect on the heat and the BOOS effect on the temperature at the super-
transfer in the furnace and superheater. By comparison, heater. The BOOS technique can be applied in com-
FGR introduced through the boiler hopper has a rela- bination with staged burners. The effectiveness of the
tively small effect on the effluent NOx but practically the described NOx reduction techniques is illustrated in
same effect on the superheater. The quantitative evalu- Figures 14.33 and 14.34.
ation of the effects, checking adequacy of the available The cumulative NOx reduction from the uncontrolled
attemperating flows, etc., often requires creation of com- level will always be less than the sum of reductions
plex CFD models (see Chapter 9) of the boiler. of each technique applied separately, as illustrated
Air staging using low NOx utility burners can be in Figure 14.33. The NOx emissions data shown in
done to a larger extent than in package and small field- Figure 14.33c are shown in relation to the burner zone
erected boilers. This is due to the substantially higher heat release rate (BZHR). The parameter is defined as
residence time in utility furnaces, higher combustion the total heat input coming into the furnace from fuel
air temperatures, and more intense mixing taking place relative to the surface area of a parallelepiped match-
in utility furnaces. The techniques are effective for both ing the boiler cross-sectional dimensions and having a
oil- and gas-fired boilers. height from the slopes of the hopper to 4 ft (1.2 m) above
The leading modern technique of air staging when the upper level of firing burners. The BZHR at high fire
firing heavy oil was originally developed by a series of operation is a boiler design parameter to which many
Electric Power Research Institute (EPRI) funded proj- different boiler emission data are correlated with rea-
ects in the early 1990s. The technology is called REACH. sonable statistical accuracy.
The first early reports on REACH are summarized in The initial success of these techniques for moderate
Ref. [22]. The essence of the refined REACH technology NOx control in the 1970s and 1980s prompted changes
resulting from those studies is the creation of a two- to the boiler designs with the addition of a row of ports
lobed pattern of fuel injection by the atomizer in con- above the upper burner level for injection in the furnace
junction with a simple venturi-type burner equipped of a portion of combustion air that otherwise would flow
with a refined medium size spinner. The exact pattern through the burners. The technique was called over-fired
of fuel injection and atomizer components are tailored air (OFA). NOx control with OFA is more effective than
to the specifics of the application. For heavy oil firing, a with BOOS, but requires substantial boiler modifications
typical level of NOx reduction with air staged burners as the OFA ports are not part of the original boiler. In
is ~40%. Similar staging techniques were developed for some cases, the whole row of upper burners was con-
firing gas with 25%–50% effective NOx reduction from verted to OFA ports. Typically, up to 20% of the total
the original uncontrolled levels (see Figure 14.33). combustion air can be diverted to the OFA ports, leaving
When firing large arrays of burners, the effects of air uniformly sub-stoichiometric amount of air to the burn-
staging can also be achieved or enhanced with uneven fuel ers. OFA should not be combined with NOx reduction
distribution between the burner levels, with some levels with BOOS as it is always more effective to increase the
firing fuel rich and other levels firing fuel lean. Obviously OFA than to shut off fuel to some of the burners.
such shifts can be done only up to the extent of maintain- For maximum effectiveness, the OFA ports need to be
ing combustion stability at each burner. A field study on properly designed to achieve adequate mixing between
fuel biasing performed on a 250 MWe (850 × 106 Btu/h) the OFA air and the upcoming combustion products and
boiler with 16 burners arranged in four rows23 illustrates leave sufficient time to complete the fuel oxidation pro-
the effectiveness of fuel biasing when firing natural gas. cess before the furnace gas gets cooled in the superheater
With 13% higher than average gas flow to the lower level and the convection pass to temperatures when the oxi-
and similar reduction in the gas flow to the upper level of dation of combustibles—primarily the CO—is stopped.
burners, the NOx was reduced by 30% in comparison with OFA is usually controlled separately from the rest of the
the unbiased operation (see Figure 14.34). The biasing also combustion air. Each port can have an isolation damper
helped to reduce attemperation in the superheater and to and resembles a smaller burner without the fuel parts.
operate the boiler up to its full capacity. When properly designed, OFA allows lower excess air
For stronger staging effects, the boiler may also oper- operation than with BOOS. CFD modeling of the boiler
ate with the fuel to some burners cut off but without combustion process is often used for designing the OFA
the burner isolation on the air side. These relatively system.
0.9
Original burner
0.8 Gas REACH
Gas REACH + 2 BOOS

0.7
Gas REACH + 3 BOOS
0.6
Gas REACH + IFGR
NOx, lb/MBtu
0.5
0.4
0.3
0.2
0.1
0
0 20 40 60 80 100 120 140 160 180
(a) Load, MW
0.5 700
NOx ranges adjusted
600
0.4 for 0.3% nitrogen
content in #6 Oil
NOx, ppm, 3% O2 dry
500
Pre-retrofit Uncontrolled
NOx, Ib/MBtu
0.3 LNB 400

LNB + BOOS
LNB + BOOS + IFGR 300
0.2 Staged
LNB + IFGR + STEAM
200
0.1
100
0.06 NOx guarantee
0 0
0 10 20 30 40 50 60 150 200 250 300 350 400
(b) Boiler generation, MW (c) BZHR, kBtu/ft2 h

FIGURE 14.33
Examples of NOx reduction with air staging and FGR in utility boilers when firing natural gas (various boilers). (a) and (b) illustrate NOx reduc-
tion firing natural gas; (c) illustrates NOx reduction when firing #6 oil in different boilers.
1.1
Further reduction in NOx emissions can be achieved
1
with FGR injected into the air stream downside of the
air heater. This is usually done in lieu of or in paral-
Relative NOx
0.9 lel with the injection of the FGR through the furnace
hopper. When the boiler is equipped with OFA, the
0.8 FGR is injected only into the portion of the combustion
air going to the burners. The effect of FGR on the NOx
0.7 reduction is very strong in gas-fired boilers operating
0.6
with highly preheated air. An addition of 10% FGR
0.6 0.7 0.8 0.9 1 1.1 typically reduces NOx by 50%–55% as illustrated in
Ratio of gas flows between lower and upper burners Figure 14.35.
One of the curves in Figure 14.33a shows another
FIGURE 14.34 example of the FGR effect on NOx. The sharp increase
Effect of fuel biasing on the NOx. 250 MWe CE boiler with 16 wall-
fired burner; natural gas firing. (From Lifshits, V. and Crovato, G.,
in the NOx shown by the curve at high fire was due to
Experience with high efficiency, low emission burners to improve limited fan capacity of the system causing a substantial
plant operation, Latin America Power’98, Conference Papers.) drop in the FGR at high fire.
90%
80%
70%
Percent thermal NOx reduction

Natural gas
60%
50%
40%
30% #6 Oil
20%
10%
0%
0% 5% 10% 15% 20%
Flue gas recirculation
FIGURE 14.35
Effect of FGR on thermal portion of NOx in different utility boilers.
The FGR effect on the NOx is much lower when fir- flames with FGR dissipate less heat from the fuel-
ing #6 oil, as a substantial amount of the formed NOx rich combustion zone. Furthermore, FGR shifts the
comes from fuel nitrogen. With a high nitrogen content fuel-rich flammability limit closer to stoichiometric
of ~0.5% to 0.6%, the effect of the FGR may not be signifi- conditions, making it more difficult to apply deep air
cant if the thermal NOx component is low, as illustrated staging NOx reduction techniques. When consider-
by the test data in Figure 14.36. These old data were ing the addition of FGR, one also needs to consider
taken when firing #6 oil with 0.54% FBN with preheated that FGR is usually the only NOx reduction technique
air to ~360°F (180°C) in an 8 ft (2.4 m) dia furnace. The that may cause an increase in the heat transfer in the
heat input was ~27 × 106 Btu/h (8 MW). The burner was superheater. Corrosion issues are also a problem when
not a low NOx design. firing with FGR when applied to fuels like #6 oil con-
The reduction of thermal NOx with FGR when fir- taining sulfur.
ing #6 oil is somewhat substantial in boilers with high The combined effects of multiple NOx reduction tech-
BZHR not equipped with low NOx burners. In combi- niques are always less than the simple multiplication
nation with low NOx burners and air staging, the use of different NOx reduction factors. Applying additional
of FGR becomes less effective. The likely explanation NOx reduction techniques often requires easing or tak-
of this effect is that lower temperature less luminous ing away other means of NOx reduction in order to
maintain flame stability.
460
14.7.1.2 Burners for Wall-Fired Utility Boilers
440
The type of burners applied to wall-fired utility fur-
NOx, ppmvd, ref. 3% O2, dry
420
naces is similar to those used in field-erected boil-
400 ers. The available round opening between the tightly
packed tubes is always all that is available for mount-
380
ing the burner. The burners for utilities are built heavy
360 duty with thick, often stainless materials. For reliable
operation and light off at any load level, the burners
340
are equipped with high capacity pilots and flame scan-
320 ners with good discrimination between the monitored
300
burner flame and the flame of adjacent or opposite wall
0% 2% 4% 6% 8% 10% 12% 14% 16% burners. The air isolation dampers are usually part
of the burner, except for rare cases when each burner
FIGURE 14.36
NOx emissions firing #6 oil with 0.54% FBN (COEN test data with CPF
has a dedicated windbox with an air isolation damper
burner; not a low NOx design). being part of the windbox. Other requirements for the
wall-fired utility burners stem from the NOx reduction wall-fired furnaces. This reduces the burner differen-
techniques utilized by the system. tial air pressure at high fire to about 3–4 in. WC (7.5–10
The array of burners that operate on oil and gas in mbar). However, stratification of species concentration
large wall-fired utilities and incorporate the necessary in the radial direction usually makes it necessary to
fuel staging techniques includes the already described operate the boilers overall with excess air slightly higher
Delta Power and Variflame burners. Another product for than for wall-fired boilers. For a typical T-fired boiler,
wall-fired utilities is a Dynaswirl™ burner that is a ver- the stack O2 is maintained in the 0.8%–1.5% range when
sion of the Variflame burner that permits fitting a larger firing gas and 1.5%–3% when firing oil.
capacity burner into the limited opening between the A potential problem that needs to be addressed with
steam–water tubes. The utility versions of these burn- the burner design is the ability to operate throughout
ers are built to withstand high levels of air preheat and the load range without generating combustion instabili-
thermal radiation from the furnace and usually include ties while making the emission guarantees. The insta-
air flow measurement devices added to the venturi bilities occur when some of the flames get detached by
passages. a substantial distance from the fuel injection points and
Fuel-staged burners, like the QLN, Delta NOx, or get reattached. The conditions are more likely to occur
ECOjet burners, are not suitable for wall-fired utility at partial loads and during warm-ups when the overall
retrofit applications due to the high costs of making firebox temperature is relatively low.
additional ports in tube walls for placing staged-fuel An oil and gas or gas-only burner for a T-fired boiler
injectors and likely insufficient benefits over the air- may be defined as a vertically stretched part of one corner
staging techniques. containing the following three major components: a few
air distribution buckets with a common tilting mechanism
if the burner is designed for tilting, a flame-stabilizing
14.7.2 Combustion Systems for Corner- Fired
spinner or shield, and a few fuel nozzles positioned in the
(T-Fired) Boilers
vicinity of the flame stabilizer. This definition is similar to
The combustion concept of a T-fired boiler is based on cre- what originally constituted a coal firing burner.
ating a single large vertical flame vortex in the middle of When a T-fired boiler is retrofitted with new lower NOx
an elongated parallelepiped-shaped furnace as explained burners or burners firing a new fuel, the common scope
in Chapter 1. Originally, the very crudely distributed mix of burner replacement usually includes modifications to
of some of the combustion air and pulverized coal was the openings from the corners into the furnace and some
delivered into the vortex through a series of corner com- stationary turning vanes for improved air distribution
partments placed at several levels. Additional preheated through the burner. Extensive modeling of the air distri-
combustion air was delivered through vertically elongated bution to the burners and through the burners usually
compartments above and below the fuel injection points. provides the basis for the design. When the new operation
The process of fuel oxidation in the vortex is very needs tilting burners—usually the case when the original
heavily air staged. Most of the flame stabilization is boiler was designed with that feature, some components
done effectively by the vortex. Reduced temperature of the existing tilting system can be preserved. The buck-
combustion products from the areas of proximity to the ets are usually replaced and to some extent modified. The
furnace walls and the corners readily re-enter the vortex fuel injectors and flame stabilizers are always new and
by the aspirating effects of flat air jets coming from the custom-engineered for the application.
corners. All these features of the combustion process The new burners for T-fired boilers usually come with
deliver lower NOx emissions than wall-fired boilers. stringent emission requirements especially for NOx.
As the boiler load modulates, the relative intensity of The key NOx reduction concepts used in T-fired boilers
the heat transfer in the superheater changes. In the T-fired are similar to those used in wall-fired boilers: low NOx
boilers, control of the superheater and reheater tempera- burners; levels out of service—not firing levels with only
tures is done usually by means of “tilting buckets,” adding the air flowing through the buckets; OFA—upper level
up or down vertical component to the direction of air flow of buckets with the fuel, but not air cutoff; and FGR.
and fuel injection coming from the corners. FGR injection The OFA in T-fired boilers can be of closed coupled type
into the furnace through the hopper is also used for the (CCOFA)—when an upper part of the air compartment
same purpose. The FGR that can be added for additional above the upper point of fuel injection passes more air
control of NOx emissions is also a factor to be considered than the part below the fuel injection level—or of sepa-
in the boiler superheater and reheater designs. rated type (SOFA)—when the upper level of air com-
As the mixing and combustion substantially take partments is substantially remote from the upper level
place in the vortex, the requirement for uniform air dis- of fired burners. The effects of FGR on the NOx reduc-
tribution to individual corners and in the vertical direction in T-fired boilers are quite similar to the wall-fired
tion in T-fired boilers is more relaxed as compared with boilers, as illustrated in Figure 14.35. One of the main
burner design goals is the need to operate throughout

the boiler load range without generating combustion
instabilities while making the emission guarantees. The
instabilities occur when some of the flames get detached
by a substantial distance from the fuel injection points
and reattached again. Insufficient means of flame sta-
bilization incorporated into the burner design, exces-
sive or improper staging, and the use of FGR—most of
the same techniques that make the system low NOx —
create difficulties in achieving operation without exces-
sive instabilities. The best burner designs successfully
navigate through these difficulties and deliver NOx FIGURE 14.38
emissions in the range of 40–60 ppm when firing natu- Main components of a fixed geometry burner for a gas and oil T-fired
boiler.
ral gas and 150–230 ppm NOx when firing #6 oil. The
actual numbers depend on factors such as preheated air
temperature, boiler size, and nitrogen content in the oil.
Figure 14.37 shows natural gas flame of a scaled down
prototype of a T-fired burner operating over a wide
range of excess air with a very stable flame anchored
at the spinner. The flame in this example was quiet and
without any signs of possible instabilities.
There are two types of burners currently offered for
T-fired boiler retrofits: burners without tilting—fixed
geometry burners—and burners designed to be inte-
grated with the tilting buckets. The fixed geometry
burners are usually custom designed and include air
flow conditioning plates and turning vanes. Other parts
include a spinner with an oil gun in the middle and a
series of lance-type gas fuel injectors placed around the
spinner. Some areas of the air passages where air enters
the furnace may be blocked with refractory-lined plates.
The pattern of fuel injection by the lances and the oil
FIGURE 14.39
gun is customized to the remaining air of the air flow Spinners for tilting burners.
passages and the needs to achieve the desired air stag-
ing and flame stabilization effects. the middle of the burner and means to change direc-
The design of tilting burners is more complex. It tionality of the gas injection. Figures 14.38 through 14.41
includes flexible oil atomizers placed in the middle of show overall and close-up views of different burners
tilting spinners or shields connected to the bucket in and some of their parts. The photograph in Figure 14.41
shows some distortion of the bucket structure caused by
the furnace heat sustained during operation with insuf-
ficient air flow through some compartments.
Flame monitoring in T-fired boilers involves monitor-
ing the overall flame in the vortex and monitoring the
attachment of the flames to the burners. The logic of
generating alarms and burner shutdowns in a T-fired
boiler is different from that of the wall-fired boilers.
At high loads, any flame detachment from the burners
will not likely generate instability problems so that the
operation of the boiler and the burner can continue as
long as the main flame vortex is present. This mode of
operation may be necessary for operation with some
additional FGR needed to control the NOx at high fire.
FIGURE 14.37
This mode of operation is called Fireball Logic. When the
Low NOx natural gas flame. T-fired burner prototype operating boiler load reduces along with a reduction in the FGR
with FGR. rates, the flame can become reattached to the burners
(a) (b)
FIGURE 14.40
Flame stabilizers and buckets of tilting (a) gas- and (b) oil-fired burner.
atomizers usually get replaced with low steam con-

sumption internal mixed atomizers producing sprays
with SMD of about 80–120 μm.
Both types of internal mixed atomizers described in
Chapter 5 are common in utility boilers. Single stage
Y-jet atomizers may have a very slight advantage with
respect to the amount of steam they use; however, the
two stage internal mix atomizers like MVI allow much
larger latitude in the design of the drilling patterns and
are able to achieve low NOx emissions. Steam consump-
tion of these atomizers at high fire varies from 5% to
10% of the amount of oil. More details on the principles
of fuel atomization can be found in Chapter 5.
FIGURE 14.41
Corner of a T-fired boiler with tilting burners with some heat damage.
by itself or with the help of burner pilots that can be 14.8 Specialized Burners
brought back into operation for the transition.
14.8.1 Warm-Up Burners
Many boiler applications require dedicated warm-up
14.7.3 Oil Atomizers for Utility Burners
burners. One group of such boilers is black and red
Due to very high heat inputs in utility boilers, on the liquor recovery boilers in the paper industry. Black and
order of several GWt, utility boiler operation is much red liquor boilers are used for the recovery and recycling
more concerned with minimizing the amount of steam of chemical components as well as recovery of thermal
used for liquid fuel atomization. energy contained in byproducts of paper production.
Mechanical atomizers do not use steam and from that The liquors are sprayed on the boiler walls covered
standpoint are the most efficient. The mechanical atom- with a thin refractory layer. Volatile organic compounds
izers need higher pressure pumps, delivering fuel with evaporate and burn, engaging oxygen in the furnace vol-
up to ~1000 psig (70 barg) pressure to the burners. They ume. The heaviest residue of the liquor flows along the
may have low turndown and are more susceptible to walls to the bottom of the boiler burning in the process
erosion. However, the main drawbacks are symmetrical using furnace oxygen. The whole process generates a
fuel injection, poor control over the atomization quality, substantial amount of PM. For the whole process to start,
and minimal ability to make changes to the fuel injec- the recovery boilers need warm-up burners that use
tion pattern. For those reasons, existing mechanical conventional fuels like gas or oil. The warm-up burners
may also be tasked to carry a substantial load to comple- combustion. All warm-up burners have minimally flared
ment or substitute the firing of liquors for the purpose of throats or not flared at all and cantilevered center-fired
power generation. The main requirement for the warm- guns that also support flame stabilizers. The stabilizer
up burners is to have small openings into the furnace as and fuel guns are often equipped with retraction mecha-
smaller openings would require a minimum amount of nisms. Flame scanning of the warm-up burner flames
purge air to fend off burner contamination with PM that often presents a challenge due to substantial blocking of
can easily form aggregates plugging the throats. The sec- the flame view by the flame stabilizer and high dust load
ond requirement for the burners is to make them tolerant in the flame area.
of the intense thermal radiation from the refractory-
lined furnace walls and flame. The third requirement
14.8.2 Flue Gas Reheat Burners
is simplicity and reliability of operation. The combus-
tion air going to the burners is initially not preheated Reheating boiler flue gas may be required for tem-
but then reaches a temperature of up to 600°F (320°C). At perature conditioning purposes upstream of the
high fire operation, the air is delivered to the burners in emission reducing catalysts. An example may be an
the amounts of 0.7–0.8 times of stoichiometric. The rest SCR process of NOx conversion to atmospheric nitro-
of the air for combustion comes from tramp air in the gen requiring temperature of flue gas in the range of
furnace. There are typically no strict NOx requirements about 650°F–750°F (340°C–400°C) for the most efficient
to the warm-up burner performance as liquids contain operation.
substantial amount of nitrogen that easily converts to a As the boiler load modulates, so does the temperature
NOx. So these boilers rely on post-combustion cleanup of of the exhaust, making it necessary during some or all of
the exhaust including the removal of NOx. the regimes to add heat to the flow. A convenient way of
Another group of boilers that uses warm-up burn- adding heat would be with special duct burners that gen-
ers with similar requirements is fluidized bed boilers. erate flames using their own source of combustion air as
Warm-up burners for these boilers are installed 5–7 ft the oxygen content in the surrounding flue gas flow is
(1.5–2 m) above the fluidized bed of finely dispersed typically very low. The difficulty with this approach is
coal or biomass at an angle toward the bed to allow flame quenching by the flue gas that may cause incom-
intense heat transfer from the flame to the bed material. plete combustion or even flame blowout. Using external
Using compact burners here is beneficial for creating burners with combustion chambers outside, on the other
higher momentum intense flames capable of delivering hand, is also inconvenient due to difficulties in han-
heat to the bed without being quenched/extinguished dling high temperature combustion products and dis-
by the bed particles. The available air pressure drop for tributing them across the flow of flue gas. Applications
the burners is usually high—over 8 in. WC (20 mbar) as of this type need to be carefully custom-engineered. For
the combustion air comes from the same source as the minimizing the stack losses, it is important that reheat
air for bed fluidization that is substantially higher. systems operate with minimum amounts of combustion
For both types of warm-up applications, the same air and uniform temperatures downstream. To achieve
type burners can be applied. Figure 14.42 shows exam- these goals, CFD modeling is usually utilized. Figure
ples of warm-up burners designed for high intensity 14.43 shows one of the CFD modeling plots of the reheat
FIGURE 14.42
Coen warm-up gas burners.
Flame shape—2500°F/5000 ppm CO iso-surface
3.60e+03
3.42e+03
3.24e+03
3.06e+03
2.88e+03
2.70e+03
2.52e+03
2.34e+03
2.16e+03
1.98e+03
1.80e+03
1.62e+03
1.44e+03
1.26e+03
1.08e+03
9.00e+02
7.20e+02
5.40e+02
3.60e+02 Y
1.80e+02
Z X
–6.66e–05
Contours of static temperature (°F) Nov 13, 2009

FLUENT 6.3 (3d, pbns, spe, nbe)
FIGURE 14.43
Conceptual design of low CO flue gas reheat system for refinery gas firing.
process, showing an example of such a solution with Another burner concept that can be applied for such
side-fired burners and refractory-lined flame shields application would be a ProLine™ burner originally
positioned across the duct designed for very low CO designed as a forced air for low NOx and low CO duct
emissions generated by the reheat burners. The key fea- burner for air heating applications. The burner uti-
ture of the design is a sufficient delay in mixing of the lizes simulated premixed lean combustion concepts.
burner combustion p roducts with surrounding flue gas Figure 14.44 shows the appearance of the burner flame
under different operating scenarios. when firing natural gas.
FIGURE 14.44
Coen ProLine™ burner flames at low and high fire operation.
their histories, the companies will continue to advance

innovation in industrial burner and controls technology
14.9 Summary
to meet the new and anticipated industry needs.
The variety of boiler burners described in this chapter is
still a small part of the burners used in industrial boil-
ers. The diversity of the equipment, fuels, geographical
regions, and environmental standards make the need
to custom-engineer the majority of combustion systems
References
even when previously developed burner concepts are
applied. The customization starts with selecting the 1. S.C. Stultz and J.B. Kitto (Ed.), Steam: Its Generation
most suitable burner type based on the available fuels, and Use, 40th Edn., Babcock & Wilcox—A McDermott
capacity demand, and environmental requirements. Company, Barberton, OH, 1992.
The next typical step is balancing between the econom- 2. C.E. Baukal (Ed.), The John Zink Combustion Handbook,
ics and performance. During that process, some com-
3. C.E. Baukal, V.Y. Gershtein, and X. Li (Eds.), Computational
promises to the equipment specification can be made. In
Fluid Dynamics in Industrial Combustion, CRC Press, Boca
retrofit applications, the variety of initial options with Raton, FL, 2001.
respect to what to reuse, refurbish, modify, or replace 4. C.E. Baukal, Industrial Combustion Testing, CRC Press,
often includes both steps to jointly become an itera- Boca Raton, FL, 2011.
tive optimization process. In some cases, none of the 5. J.M. Beer and N.A. Chigier, Combustion Aerodynamics,
prior developed burner types are sufficiently suited Robert E. Krigier Publishing Company, Malabar, FL,
for the application, prompting development of the new 1983.
concepts. 6. A.K. Gupta, Swirl Flows, Abacus Press, Kent, England, 1984.
From this perspective, the issues discussed in this 7. Y.B. Zeldovich, P.Y. Sadovnikov, and D.A. Frank-
chapter and also in Chapter 12 on the combustion con- Kamenetskii, Oxidation of Nitrogen in Combustion,
Academy of Science, USSR, Institute of Chemical
trols should help end users and boiler designers to bet-
Physics, Moscow-Leningrad, Russia, 1947.
ter understand the issues and options that combustion
8. C.P. Fenimore, Formation of nitric oxide in premixed
engineers face when selecting and designing burners hydrocarbon flames, Thirteenth Symposium on Combustion,
and combustion systems. The days have long passed The Combustion Institute, Pittsburgh, PA, pp. 373–380,
when a burner and combustion system were simple and 1971.
separate. 9. W. Bartok, A.R. Crawford, H.J. Hall, E.H. Manny, and
Changes in the burner designs and system applica- A. Skopp, Stationary sources and control of nitrogen
tions have been driven primarily by emissions since oxide emissions, Proceedings, 2nd International Clean Air
the passage of the Clean Air Act in the early 1970s. Of Congress, Academic Press, Washington, DC, pp. 80–90,
necessity, the burner designs have become more com- 1971.
plex, requiring a much deeper technical understanding 10. Electric Power Research Institute, NOx controls for utility
boilers, Proceedings, July 7–9, Cambridge, MA, 1992.
of the system interactions and required controls.
11. D.W. Turner, R.L. Andrews, and Cg.W. Siegmund, Influence
NOx emissions are typically about 10 times lower and
of combustion modifications and fuel nitrogen content on
CO emissions are on the order of 100 times lower than in nitrogen oxide emissions from fuel oil combustion, AIChE
the 1970s. Each application of a burner has to be evalu- Symp. Ser., 68, 55 (1971).
ated based on the details of a particular furnace design. 12. A.H. Rawdon and S.A. Johnson, Application of NOx
The performance of boilers and especially their super- control technology to power boilers, Presented at the
heaters and reheaters is heavily impacted by the uti- American Power Conference, May 10, 1973.
lized low NOx techniques. Chapter 1 and this chapter 13. V. Lifshits and S. Drennan, Development of an Ultra Low
discuss custom-engineered solutions. In many ways, NOx Burner with widened Stability Limits, American Flame
every application of a burner is a custom-engineered Research Committee, Maui, Hawaii, 1998.
solution. The only question will be the degree of custom 14. Ultra low NOx gas-fired burner with air preheat, CARB
Contract Number 94–354, Final Report, prepared by
engineering that is required. Every burner type will
Coen Company and Arthur D. Little for the California Air
perform differently in different furnaces. Each applica-
Resources Board, California Environmental Protection
tion has technical and economic limitations, especially Agency, November 2000.
retrofit projects. 15. V. Lifshits and S. Londerville, Vibration resistant low
In August 2012 Coen Company celebrated 100 years NOx burner, U.S. Patent No. 5,310,337.
since becoming a corporation, and Hamworthy corpo- 16. V. Lifshits and S. Londerville, Vibration resistant low
ration is just 2 years younger. In the second century of NOx burner, U.S. Patent No. 5,460,512.
17. V. Lifshits, Development of a high performance versa- 21. V. Lifshits, Energy efficient ultra low NOx burner with
tile low NOx burner, Presented at AFRC International reduced flue gas recirculation, P16th IFRF Members’
Symposium, Baltimore, MD, September 1996. Conference Symposium, Boston, MA, June 2009.
18. S.J. Bortz, Apparatus and method for reducing NOx, CO 22. D. Giovanni, Atomization and swirler design techniques
and hydrocarbon emissions when burning gaseous fuels, for improved NOx emissions, NOx controls for utility
U.S. Patent No: 5,407,347. boilers, EPRI Meeting, Cambridge, MA, July 7–9, 1992.
19. V. Lifshits, Low NOx fuel gas burner, U.S. Patent No: 23. V. Lifshits and G. Crovato, Experience with high effi-
6,027,330. ciency, low emission burners to improve plant operation,
20. V. Lifshits, Energy efficient low NOx burner and method Latin America Power’98, Conference Papers, Buenos Aires,
of operating same, U.S. Patent No: 7,422,427. Argentina, pp. 592–601, 1998.
15
Duct Burners
Peter F. Barry, Stephen L. Somers, Stephen B. Londerville, Kenneth Ahn, and Kevin Anderson
CONTENTS
15.1 Introduction.................................................................................................................................................................... 398
15.2 Applications.................................................................................................................................................................... 398
15.2.1 Cogeneration....................................................................................................................................................... 398
15.2.2 Combined Cycle................................................................................................................................................. 398
15.2.3 Air Heating......................................................................................................................................................... 398
15.2.4 Fume Incineration.............................................................................................................................................. 400
15.2.5 Stack Gas Reheat................................................................................................................................................ 400
15.3 Burner Technology........................................................................................................................................................ 400
15.3.1 In-Duct or Inline Configuration...................................................................................................................... 400
15.3.2 Grid Configuration (Gas Firing)...................................................................................................................... 402
15.3.3 Grid Configuration (Liquid Firing)................................................................................................................. 403
15.4 Fuels................................................................................................................................................................................. 404
15.4.1 Natural Gas......................................................................................................................................................... 404
15.4.1.1 Refinery/Chemical Plant Fuels......................................................................................................... 404
15.4.1.2 Low Heating Value............................................................................................................................. 404
15.4.1.3 Liquid Fuels......................................................................................................................................... 404
15.5 Combustion Air and Turbine Exhaust Gas................................................................................................................ 404
15.5.1 Temperature and Composition........................................................................................................................ 404
15.5.2 Turbine Power Augmentation.......................................................................................................................... 405
15.5.3 Velocity and Distribution................................................................................................................................. 405
15.5.4 Ambient Air Firing (Air-Only Systems and HRSG Backup)....................................................................... 406
15.5.5 Augmenting Air................................................................................................................................................. 406
15.5.6 Equipment Configuration and TEG/Combustion Airflow Straightening................................................ 406
15.6 Physical Modeling......................................................................................................................................................... 407
15.6.1 CFD Modeling.................................................................................................................................................... 408
15.6.1.1 Wing Geometry: Variations............................................................................................................... 408
15.7 Emissions........................................................................................................................................................................ 409
15.7.1 Visible Plumes.................................................................................................................................................... 410
15.7.2 NOx and NO versus NO2.................................................................................................................................. 410
15.7.3 CO, UBHC, SOx, and Particulates.....................................................................................................................411
15.7.3.1 Carbon Monoxide................................................................................................................................411
15.7.3.2 Unburned Hydrocarbons................................................................................................................... 412
15.7.3.3 Sulfur Dioxide..................................................................................................................................... 412
15.7.3.4 Particulate Matter............................................................................................................................... 412
15.8 Maintenance................................................................................................................................................................... 413
15.8.1 Accessories...........................................................................................................................................................416
15.8.1.1 Burner Management System..............................................................................................................416
15.8.1.2 Fuel Train..............................................................................................................................................416
397
15.9 Design Guidelines and Codes...................................................................................................................................... 418

15.9.1 NFPA 8506 (National Fire Protection Association)....................................................................................... 418
15.9.2 Factory Mutual................................................................................................................................................... 418
15.9.3 Underwriters’ Laboratories.............................................................................................................................. 418
15.9.4 ANSI B31.1 and B31.3 (American National Standards Institute)................................................................ 419
15.9.5 Others.................................................................................................................................................................. 419
References................................................................................................................................................................................. 419
costs, but produce relatively high levels of pollutants.

Turbine engines are used in both small and large systems
15.1 Introduction
(3 MW = 10 × 106 Btu/h and above) and, although more
Linear and in-duct burners were used for many years expensive, generally emit lower levels of air pollutants.
to heat air in drying operations before their general use Fossil fuels used in cogeneration systems can consist
in cogeneration systems. Some of the earliest systems of almost any liquid or gaseous hydrocarbon, although
premixed fuel and air in an often-complicated configu- natural gas and various commercial-grade fuel oils are
ration that fired into a recirculating process airstream. most commonly used. Mixtures of hydrocarbon gases
The first use was in high-temperature, depleted oxygen and hydrogen found in plant fuel systems are often
streams downstream of gas turbines, in the early 1960s, used in refining and petrochemical applications. Duct
to provide additional steam for process use in industrial burners are capable of firing all fuels suitable for the
applications and for electrical peaking plants operating engine/turbine, as well as many that are not, including
steam turbines. As gas turbines have become larger and heavy oils and waste gases.
more efficient, duct burner supplemental heat input has Heat recovery for large systems is usually accom-
increased correspondingly. plished by convective heat transfer in a boiler (commonly
Linear burners are applied where it is desired to referred to as a heat recovery steam generator, also known
spread heat uniformly across a duct, whether in ambi- by the acronym HRSG). Smaller systems utilize either a
ent air or oxygen-depleted streams. In-duct designs are steam or hot water boiler or, alternatively, some type of
more commonly used in fluidized bed boilers and small air-to-air heat exchanger for direct transfer to a process.
cogeneration systems. Supplementary firing is often incorporated into the
boiler/HRSG design as it allows increased production of
steam as demanded by the process. The device that pro-
vides the supplementary firing is a duct burner, so called
because it is installed in the duct connecting the engine/
15.2 Applications turbine exhaust to the heat recovery device, or just down-
stream of a section of the HRSG superheater (see Figures
15.2.1 Cogeneration 15.4 and 15.5). Oxygen required for the combustion pro-
Cogeneration implies simultaneous production of two cess is provided by the turbine exhaust gas (TEG).
or more forms of energy, most commonly electrical
(electric power), thermal (steam, heat transfer fluid, or 15.2.2 Combined Cycle
hot water), and pressure (compressor). The basic process
involves combustion of fossil fuel in an engine (recipro- Combined cycle systems incorporate all components
cating or turbine) that drives an electric generator, cou- of the simple cycle configuration with the addition of
pled with a recovery device that converts heat from the a steam turbine/generator set powered by the HRSG.
engine exhaust into a usable energy form. Production This arrangement is attractive when the plant cannot be
of recovered energy can be increased independently of located near an economically viable steam user. Also,
the engine through supplementary firing provided by when used in conjunction with a duct burner, the steam
a special type of burner known as a duct burner. Most turbine/generator can provide additional power during
modern systems will also include flue gas emission periods of high or “peak” demand.
control devices. A typical plant schematic is shown in
Figure 15.1. Aerial views of typical combined cycle elec- 15.2.3 Air Heating
tric power plants are shown in Figures 15.2 and 15.3.
Reciprocating engines (typically diesel cycle) are used Duct burners are suitable for a wide variety of direct-
in smaller systems (10 MW = 34 × 106 Btu/h and lower) fired air heating applications where the physical
and offer the advantage of lower capital and maintenance arrangement requires mounting inside a duct, and
Duct Burners 399
Transform
60 MW of electricity
Steam from the HRSG can
available for sale or use
Natural 40 power the steam turbine
gas to produce additional 20
Gas turbine Generator electricity Steam
turbine Generator
Inlet
Natural gas is burned in Extraction
control
a gas turbine couple steam
valve Condensing
to a generator to
produce electricity steam
District heating
Pressure
reducing
Exhaust Heat valve
recovery Steam Steam
Exhaust gases from steam
the gas turbine are generator
directed to a HRSG to (HRSG) Steam can be
produce steam directed for
production
of more electricity
Water
Process industry
Pump
Condenser
High-quality
Water Water Condensate drinking water
treatment system for home
Exhaust steam from and industry
the steam turbine is
cooled, turning back Reverse
Pump into water (condensate) osmosis
which is return to HRSG
Cooling Sea or
tower brackish
water
FIGURE 15.1
Typical plant schematic. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
particularly for processes where the combustion air is at cross exchange with hot stack gas increases the
an elevated temperature and/or contains less than 21% air temperature and velocity. Burners are shut
oxygen. Examples include off when the desired air preheat is reached and
the bed can sustain combustion unaided.
• Fluidized bed boilers (see Figure 15.6): where • Combustion air blower inlet preheaters: where
burners are installed in combustion air ducts burners are mounted upstream of a blower inlet
and used only to provide heat to the bed dur- to protect against thermal shock caused by ambi-
ing startup. At cold conditions, the burner is ent air in extremely cold climates (−40°F/°C and
fired at maximum capacity with fresh ambi- below). This arrangement is only suitable when
ent air; but as combustion develops in the bed, the air will be used in a combustion process as
FIGURE 15.2
Cogeneration at Teesside, England. (Courtesy of Nooter/Eriksen, Fenton, MO.)
it will contain combustion products from the 15.2.5 Stack Gas Reheat
duct burner.
Mounted at or near the base of a stack, heat added by a
• Drying applications: where isolation of combus- duct burner will increase natural draft, possibly elim-
tion products from the work material is not inating the need for induced draft or eductor fans. In
required, such as certain paper and wallboard streams containing a large concentration of water vapor,
manufacturing operations. the additional heat can also reduce or eliminate poten-
tially corrosive condensation inside the stack. A source
of ambient augmenting combustion air is often added if
15.2.4 Fume Incineration the stack gas oxygen concentration is low. This arrange-
ment may also provide a corollary emissions reduction
Burners are mounted inside ducts or stacks carry- benefit (see Section 15.7). A discussion on testing duct
ing exhaust streams primarily composed of air with burner performance is given in Ref. [1].
varying concentrations of organic contaminants.
Undesirable components are destroyed, both by an
increase in the gas stream bulk temperature and
through contact with localized high temperatures cre-
ated in the flame envelope. Particular advantages of 15.3 Burner Technology
the duct burner include higher thermal efficiency as no
outside air is used, lower operating cost as no blower is 15.3.1 In-Duct or Inline Configuration
required, and improved destruction efficiency result- Register or axial flow burner designs are adapted for
ing from distribution of the flame across the duct sec- installation inside a duct. The burner head is oriented
tion with grid-type design. such that the flame will be parallel to and co-flow with
Duct Burners 401
FIGURE 15.3
Combination (oil and gas) fired duct burners at Dahbol, India. (Courtesy of Enron, Houston, TX.)
Integral deaerator
Silencer Uses low temperature heat
Attenuates noise level to deaerate feed-water for
to meet government improved thermal efficiency
and site requirements CO catalyst
Reduces carbon
monoxide in the
flue gas
SCR catalyst
Reduces nitrous
oxides in the
flue gas
Duct burner
Provides supplementary firing
of turbine exhaust to increase
Diverter valve unfired steam production
Modulates steam production
in the bypass systems
FIGURE 15.4
Typical location of duct burners in an HRSG. (Courtesy of Hamon Deltak.)
Teesside power limited Steam conditions: psig °F

Teesside, England HP 1250 psig 900°F
Customer: Enron power IP 250 psig 450°F
LP Saturated
Start-up: April 1993 Special features: Supplemental firing
Plant capacity: 1875 MW silencing
Fuel: Natural gas
Gas turbines: W 701 DA HP steam drum
No. of units: 8 IP steam drum
LP steam drum
HP economizer no: 3
HP evaporator 19΄–0˝ dia.
HP superheater no. 1
IP superheater/IP evaporator
HP superheater no. 2 HP/IP economizer no. 2
Duct burner LP evaporator
HP/IP/LP economizer no. 1
Silencer baffles
115΄–3˝
FIGURE 15.5
Schematic of HRSG at Teesside, England. (Courtesy of Nooter/Eriksen, Fenton, MO.)
distribution than grid-type. On the other hand, they are

more flexible in burning liquid fuels, can be more easily
modified to incorporate augmenting air, and sometimes
represent a less expensive option for high firing rates in
small ducts without sufficient room for grid elements.
15.3.2 Grid Configuration (Gas Firing)

A series of linear burner elements that span the duct
width are spaced at vertical intervals to form a grid.
Each element is comprised of a fuel manifold pipe
fitted with a series of flame holders (or wings) along
its length. Fuel is fed into one end of the manifold
pipe and discharged through discrete multiport tips
FIGURE 15.6 attached at intervals along its length, or through holes
Fluidized bed startup duct burner. (From Baukal, C.E. (ed.), The John
drilled directly into the pipe. Gas ports are positioned
such that fuel is injected in co-flow with the TEG. The
wings meter the TEG or airflow into the flame zone,
the air or TEG stream, and the fuel supply piping is fed thus developing eddy currents that anchor ignition.
through the duct sidewall, turning 90° as it enters the They also shield the flame in order to maintain suit-
burner (see Figure 15.7). Depending on the total firing ably high flame temperatures, thereby preventing
rate and duct size, one burner may be sufficient, or sev- excessive flame cooling that might cause high emis-
eral may be arrayed across the duct cross section. Inline sions. Parts exposed to TEG and the flame zone are
burners typically require more air/TEG pressure drop, typically of high-temperature alloy construction (see
produce longer flames, and offer a less uniform heat Figures 15.8 and 15.9).
Duct Burners 403
FIGURE 15.7
Inline burner. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
15.3.3 Grid Configuration (Liquid Firing)

As with the gas-fired arrangement, a series of linear
burner elements comprised of a pipe and flame hold-
Fuel ers (wings) span the duct width. However, instead of
TEG flow + injector multiple discharge points along the pipe length, liquid
spud fuel is injected downstream of the element through the
duct sidewall, and directed parallel to the flame holders
Fuel Flame (cross flow to the TEG). This configuration utilizes the
supply holder duct cross section for containment of the flame length,
runner thus allowing a shorter distance between the burner and
downstream boiler tubes (see Figure 15.10). The injec-
FIGURE 15.8 tion device, referred to as a side-fired oil gun, utilizes
Linear burner elements. (From Baukal, C.E. (ed.), The John Zink
a mechanical nozzle supplemented by low-pressure air

Gas flame from a grid burner. (From Baukal, C.E. (ed.), The John Zink Oil flame from a side-fired oil gun. (From Baukal, C.E. (ed.), The John
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
(2–8 psi) (14–55 kPa) to break the liquid fuel into small heating values (range of 50–500 Btu/scf = 1.9–19 MJ/m3).
droplets (atomization) that will vaporize and readily These fuels burn more slowly and at lower temperatures
burn. Although most commonly used for light fuels, this than conventional fuels, and thus require special design
arrangement is also suitable for some heavier fuels where considerations. Fuel pressure is reduced to match its
the viscosity can be lowered by heating. In some cases, velocity to flame speed, and some form of shield or
high-pressure steam may be required, instead of low- “canister” is employed to provide a protected flame
pressure air, for adequate atomization of heavy fuels. zone with sufficient residence time to promote complete
combustion before the flame is exposed to the quench-
ing effects of TEG.
Other considerations that must be taken into account
are moisture content and particulate loading. High
moisture concentration results in condensation within
15.4 Fuels
the fuel supply system, which in turn produces corro-
15.4.1 Natural Gas sion and plugging. Pilots and igniters are particularly
susceptible to the effects of moisture because of small
Natural gas is, by far, the most commonly used
fuel port sizes, small igniter gap tolerance, and the insu-
fuel because it is readily available in large volumes
lation integrity required to prevent “shorting” of electri-
throughout much of the industrialized world. Because
cal components. A well-designed system might include
of its ubiquity, its combustion characteristics are well
a knock-out drum to remove liquids and solids, insula-
understood, and most burner designs are developed
tion and heat-tracing of piping to prevent or minimize
for this fuel.
condensation, and low-point drains to remove con-
densed liquids. Problems are usually most evident after
15.4.1.1 Refinery/Chemical Plant Fuels a prolonged period of shutdown.
Solid particulates can cause plugging in gas tip ports
Refineries and chemical plants are large consumers of or other fuel system components and should therefore
both electrical and steam power, which makes them be removed to the maximum practical extent. In gen-
ideal candidates for cogeneration. In addition, these eral, particle size should be no greater than 25% of the
plants maintain extensive fuel systems to supply the smallest port, and overall loading should be no greater
various direct- and indirect-fired processes as well as than 5 ppm by volume (weight).
to make the most economical use of residual products.
This latter purpose presents special challenges for duct
burners because the available fuels often contain high 15.4.1.3 Liquid Fuels
concentrations of unsaturated hydrocarbons with a ten-
In cogeneration applications, duct burners are com-
dency to condense and/or decompose inside burner
monly fired with the same fuel as the turbine, which
piping. The location of burner elements inside the TEG
is typically limited to light oils such as No. 2 or naph-
duct, surrounded by high-temperature gases, exacer-
tha. For other applications, specially modified side-
bates the problem. Plugging and failure of injection noz-
fired guns or an inline design can be employed to burn
zles can occur, with a corresponding decrease in online
heavier oils such as No. 6 and some waste fuels.
availability and an increase in maintenance costs.
With appropriate modifications, however, duct
burners can function reliably with most hydrocarbon-
based gaseous fuels. Design techniques include insu-
lation of burner element manifolds, insulation and
heat-tracing of external headers and pipe trains, and 15.5 Combustion Air and
fuel/steam blending. Steam can also be used to peri- Turbine Exhaust Gas
odically purge the burner elements of solid deposits
before plugging occurs. 15.5.1 Temperature and Composition
When used for supplementary firing in HRSG cogen-
eration applications, the oxygen required for the com-
15.4.1.2 Low Heating Value
bustion reaction is provided by the residual in the TEG
By-product gases produced in various industrial pro- instead of a new, external source of air. Because this
cesses, such as blast furnaces, coke ovens, and flexi- gas is already at an elevated temperature, duct burner
cokers, or from mature landfills, contain combustible thermal efficiency can exceed 90% as very little heat is
compounds along with significant concentrations of required to raise the combustion products temperature
inert components, thus resulting in relatively low to the final fired temperature. TEG contains less oxygen
Duct Burners 405
17 the flame from anchoring to the burner. Grid-type con-

Depends on:
figurations can generally operate at velocities ranging
Fuel composition
from 20 to 90 ft/s or 6 to 27 m/s and pressure drops of less
TEG velocity
than 0.5 in. water column. Inline or register burners typi-
TEG oxygen, % (vol., wet)
cally require velocities of 100–150 ft/s (31–46 m/s) with a

No augmenting air required pressure drop of 2–6 in. water column (5–15 mbar).
Grid burners are designed to distribute heat uniformly
across the HRSG or boiler tube bank, and thus require a
reasonably uniform distribution of TEG or air to supply
the fuel with oxygen. Inadequate distribution causes local-
Augmenting air required
ized areas of low velocity, resulting in poor flame defini-
tion along with high emissions of CO and UHCs. Turbine
11 exhaust flow patterns, combined with rapidly diverging
500 1100 downstream duct geometry, will almost always produce
TEG temperature, °F
an unsatisfactory result that must be corrected by means
of a straightening device. Likewise, the manner in which
FIGURE 15.11
ambient air is introduced into a duct can also result in
Approximate requirement for augmenting air. (From Baukal, C.E. (ed.),
The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.) flow maldistribution, requiring some level of correction.
Selection and design of flow-straightening devices are dis-
cussed elsewhere in this chapter (see Figure 15.12).
than fresh air, typically between 11% and 16% by volume, In instances where bulk TEG or air velocity is lower
which, in conjunction with the TEG temperature, will than required for proper burner operation, flow straight-
have a significant effect on the combustion process. As ening alone is not sufficient and it becomes necessary
the oxygen concentration and TEG temperature become to restrict a portion of the duct cross section at or near
lower, emissions of CO and unburned hydrocarbons the plane of the burner elements, thereby increasing the
(UHCs) occur more readily (see Chapter 10), eventually “local” velocity across flame holders. This restriction, also
progressing to combustion instability. The effect of low referred to as blockage, commonly consists of unfired
oxygen concentration can be partially offset by higher runners or similar shapes uniformly distributed between
temperatures; conversely, higher oxygen concentrations the firing runners to reduce the open flow area.
will partially offset the detrimental effects of low TEG Inline, or register, burners inject fuel in only a few
temperatures. This relationship is depicted graphically (or possibly only one) positions inside the duct, and
in Figure 15.11. Duct burner emissions are discussed in can therefore be positioned in an area of favorable flow
more detail elsewhere in this chapter. conditions, assuming the flow profile is known. On the
other hand, downstream heat distribution is less uni-
form than with grid designs, and flames may be longer.
15.5.2 Turbine Power Augmentation
During periods of high electrical demand, various tech-
niques are employed to increase power output, and most
will increase the concentration of water vapor in TEG. The
corresponding effect is a reduction in TEG oxygen concen-
tration and temperature with consequent effects on duct
burner combustion. Depending on the amount of water
vapor used, CO emissions may simply rise, or in extreme
cases the flame may become unstable. The former effect
can be addressed with an allowance in the facility operat-
ing permit or by increasing the amount of CO catalyst in
systems so equipped. The latter requires air augmentation,
a process whereby fresh air is injected at a rate sufficient to
raise the TEG oxygen concentration to a suitable level.
15.5.3 Velocity and Distribution

Regardless of whether TEG or fresh air is used, velocity FIGURE 15.12
across flame stabilizers must be sufficient to promote mix- Drawing of a duct burner arrangement. (From Baukal, C.E. (ed.), The
ing of the fuel and oxygen, but not so great as to prevent John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
As with grid-type burners, in some cases, it may be nec- drop in the TEG temperature and oxygen concentration
essary to block portions of the duct at or just upstream occurs because of dilution. The TEG temperature is also
of the burners to force a sufficient quantity of TEG or air reduced in installations where the HRSG manufacturer
through the burner. splits the steam superheater and places tubes upstream
of the duct burner.
With their research and development facilities, manu-
15.5.4 Ambient Air Firing (Air-Only
facturers have defined the oxygen requirement with
Systems and HRSG Backup)
respect to TEG temperature and fuel composition,
Velocity and distribution requirements for air systems and are able to quantify the amount of augmenting air
are similar to those for TEG, although inlet temperature required under most conditions likely to be encoun-
is not a concern because of the relatively higher oxygen tered. It is usually not practical to add enough air to the
concentration. As with TEG applications, the burner ele- turbine exhaust to increase the oxygen content to an
ments are exposed to the products of combustion, so adequate level. Specially designed runners are therefore
material selection must take into account the maximum used to increase the local oxygen concentration. In cases
expected fired temperature. where augmenting air is required, the flow may be sub-
Ambient (or fresh) air backup for HRSGs presents stantial: from 30% to 100% of the theoretical air required
special design challenges. Because of the temperature for the supplemental fuel.
difference between ambient air and TEG, designing The augmenting air runner of one manufacturer con-
for the same mass flow and fired temperature will sists of a graduated air delivery tube designed to ensure
result in velocity across the burner approximately one- a constant velocity across the length of the tube. Equal
third that of the TEG case. If the cold condition veloc- distribution of augmenting air across the face of the
ity is outside the acceptable range, it will be necessary tube is imperative. The augmenting air is discharged
to add blockage, as described earlier. Fuel input capac- from the tube into a plenum and passes through a sec-
ity must also be increased to provide heat required to ond distribution grid to further equalize flow. The air
raise the air from ambient to the design firing tempera- passes through perforations in the flame holder, where
ture. By far, the most difficult challenge is related to it is intimately mixed with the fuel in the primary com-
flow distribution. Regardless of the manner in which bustion zone. This intimate mixing ensures correspond-
backup air is fed into the duct, a flow profile differing low CO and UHC emissions under most conditions
ent from that produced by the TEG is virtually cer- likely to be encountered. Once the decision has been
tain. Flow-straightening devices can therefore not be made to supply augmenting air to a burner, it is an inev-
optimized for either case, but instead require a com- itable result of the design that the augmenting air will
promise design that provides acceptable results for be part of the normal operating regime of the combus-
both. If the two flow patterns are radically different, tion runner.
it may ultimately be necessary to alter the air injection
arrangement independently of the TEG duct-straight-
15.5.6 Equipment Configuration and
ening device.
TEG/Combustion Airflow Straightening
The TEG/combustion air velocity profile at the duct
15.5.5 Augmenting Air
burner plane must be within certain limits to ensure
As turbines have become more efficient and more good combustion efficiency; in cogeneration applica-
work is extracted in the form of, for example, electric- tions, this is rarely achieved without flow-straightening
ity, the oxygen level available in the TEG continues to devices. Even in non-fired configurations, it may be nec-
get lower. To some extent, a correspondingly higher essary to alter the velocity distribution to make efficient
TEG temperature provides some relief for duct burner use of the boiler heat transfer surface. Figure 15.13 shows
operation. a comparison of flow variation with and without flow
In some applications, however, an additional oxygen straightening.
source may be required to augment that available in the Duct burners are commonly mounted in the TEG duct
TEG when the oxygen content in the TEG is not suffi- upstream of the first bank of heat transfer tubes, or they
cient for combustion at the available TEG temperature. may be nested in the boiler superheater between banks
If the mixture adiabatic flame temperature is not high of tubes. In the former case, a straightening device
enough to sustain a robust flame in the highly turbulent would be mounted just upstream of the burner, while
stream, the flame may become unstable. in the latter it is mounted either upstream of the first
The problem can be exacerbated when the turbine tube bank or between the first tube bank and (upstream
manufacturer adds large quantities of steam or water for of) the burner. Although not very common, some HRSG
NOx control and power augmentation. A corresponding design configurations utilize two stages of duct burners
Duct Burners 407
Comparison of flow variation Perforated plates that extend across the entire duct
9 cross section are most commonly used for flow straight-
8 No flow
distribution ening because experience has shown that they are less
7
6 devices prone to mechanical failure than vane-type devices, even
5 though they require a relatively high pressure drop. The
4
pattern and size of perforations can be varied to achieve
Relative elevation
3
2 the desired distribution. Vanes can produce comparable
1 results with significantly less pressure loss but require
9
8
substantial structural reinforcement to withstand the
With flow
7 distribution flow-induced vibration inherent in HRSG systems.
6 grid Regardless of the method used, flow pattern complexity—
5
4
particularly in TEG applications—usually dictates the use
3 of either physical or computational fluid dynamic (CFD)
2 modeling (see Chapter 9) for design optimization.
1
50 75 100 125 150
Percent flow relative to mean
FIGURE 15.13
Comparison of flow variation with and without straightening device. 15.6 Physical Modeling
(From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
Press, Boca Raton, FL, 2001.) TEG/airflow patterns are determined by inlet flow char-
acteristics and duct geometry, and are subject to both
with heat transfer tube banks in between, and a flow- position and time variation. Design of an efficient (low
straightening device upstream of the first burner. Such pressure loss) flow-straightening device is therefore not a
an arrangement is, however, problematic because the trivial exercise, and manual computational methods are
TEG downstream of the first-stage burner may not have impractical. For this reason, physical models, commonly
the required combination of oxygen and temperature 1:6 or 1:10 scale, are constructed, and flow characteristics
properties required for proper operation of the second- are analyzed by flowing air with smoke tracers or water
stage burner. with polymer beads through the model (see Figure 15.14).
FIGURE 15.14
Physical model of duct burner system. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Although this method produces reliable results, tests

conducted at ambient conditions (known as “cold flow”)
are not capable of simulating the buoyant effects that may
occur at elevated temperatures.
15.6.1 CFD Modeling

Flow modeling with CFD (see Chapter 9), using a com-
puter-generated drawing of the inlet duct geometry, is
capable of predicting flow pattern and pressure drop in
the turbine exhaust flow path. The model can account
for swirl flow in three dimensions, accurately predict
pressure drop, and subsequently help design a suitable
device to provide uniform flow. The CFD model must Contours of velocity magnitude (ft/s) Feb 29, 2000
be quite detailed to calculate flow patterns incident and Through center of duct Fluent 5.3 (3d, segregated, rke)
through a perforated grid or tube bank while also keep-
ing the overall model solution within reasonable com- FIGURE 15.15
Sample result of CFD modeling performed on an HRSG inlet duct.
putation time. Combustion effects can be included in the
calculations at the cost of increased computation time.
CFD simulation has the capability to provide com- flame zone in the required ratio to create a combustible
plete information, provided the aforementioned is true. mixture and ensure that the combustion products do not
The issue of validity has been a hot topic for years. escape before the reactions are complete. In response to
A Department of Energy report2 has cited CFD to be new turbine and HRSG design requirements, each duct
capable of burner manufacturer has proprietary designs devel-
oped to provide the desired results.
1. Predicting catastrophic failure 15.6.1.1.2 Basic Flame Holder
2. Qualitative trends and parametric analysis In its basic form, a fuel injection system and a zone for
3. Visualization mixing with oxidant are all that is required for combus-
4. Predicting nonreacting gaseous flows tion. For application to supplemental firing, the simple
5. Quantitative analysis of gas velocity and tem- design shown in Figure 15.16 consists of an internal
perature patterns
6. Qualitative analysis of radiation heat transfer
7. Flame dynamics and shape
8. Effecting geometry changes Flame holder
9. Models of temperature and heat release pat-

terns and qualitative trends associated with
major species
10. Integration of detailed burner codes with heat-
ing process Fuel
TEG flow supply Drilled pipe
runner
For combustion systems, CFD is the only general-
purpose simulation model capable of modeling reacting
flows in order to predict emissions, heat transfer, and
other furnace parameters. Figure 15.15 shows a sample
result of CFD modeling performed on a HRSG inlet duct.
Flame holder
15.6.1.1 Wing Geometry: Variations

15.6.1.1.1 Flame Holders
Design of the flame stabilizer, or flame holder, is critical FIGURE 15.16
to the success of supplementary firing. Effective emis- Drilled pipe duct burner. (From Baukal, C.E. (ed.), The John Zink
sion control requires that the TEG be metered into the Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Duct Burners 409
FIGURE 15.17
Low-emission duct burner. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
manifold or “runner,” usually an alloy pipe with fuel the cooling needed to protect the metal from oxidation.
injection orifices spaced along the length. A bluff body Alternately, fuels subject to cracking, such as propyl-
plate, with or without perforations, is attached to the ene, may not have the oxygen needed to minimize coke
pipe to protect the flame zone from the turbulence in buildup.
the exhaust gas duct. The low-pressure zone pulls the Another manufacturer supplies custom designs to
flame back onto the manifold. This low-cost runner may accommodate velocity extremes, while maintaining low
overheat the manifold, causing distortion of the metal- emissions. In the design shown in Figure 15.17, the flame
lic parts. Emissions are unpredictable with changing holder is optimized with CFD and research experimen-
geometry and CO is usually much higher than the cur- tation to enhance mixing and recirculation rate. Special
rent typically permitted levels of under 0.1 lb/MMBtu. construction materials are easily accommodated. This
supplier also uses removable fuel tips with multiple ori-
15.6.1.1.3 Low-Emissions Design fices, which can be customized to counteract any unex-
Modifications to the design for lower emission perfor- pected TEG flow distribution discovered after commercial
mance generally have a larger cross section in the plane operation. Figure 15.18 depicts the flow patterns of air/
normal to the exhaust flow. The increased blocked area TEG and fuel in relation to the duct burner flame holder.
protects the fuel injection zone and increases residence
time. The NOx is reduced by the oxygen-depleted TEG
and the CO/UHC is reduced by the delayed quenching.
The correct flow rate of TEG is metered through the ori-
fices in the flame holder, and the fuel injection velocity
15.7 Emissions
and direction are designed to enhance combustion effi-
ciency. The flame zone is pushed away from the inter- Duct burner systems can either increase or reduce emis-
nal manifold (“runner” pipe), creating space for cooling sions (see Chapter 10) from the generally large volume
TEG to bathe the runner and flame holder and enhance of mass flow at the input. Generally this flow includes
equipment life. particulates, NOx, CO, and a variety of HCs includ-
Each manufacturer approaches the geometry some- ing a subset of HCs defined as VOCs (volatile organic
what differently. One manufacturer uses cast alloy compounds). VOCs are defined by EPA (40 CFR 51.100,
pieces welded together to provide the required block- February 3, 1992) as “any compound of carbon, excluding
age. These standard pieces often add significant weight carbon monoxide, carbon dioxide, carbonic acid, metallic
and are difficult to customize to specific applications. carbides or ammonium carbonate, which participates in
Hot burning fuels, such as hydrogen, may not receive atmospheric chemical reaction.” Other compounds are
15.7.2 NOx and NO versus NO2

Formation of NO and NO2 is the subject of ongoing
research to understand the complex reactions (see
Chapter 3). Potentially, several oxides of nitrogen (NOx)
can be formed during the combustion process, but only
nitric oxide (NO) and nitrogen dioxide (NO2) occur in
significant quantities.
In the elevated temperatures found in the flame zone
in a typical HRSG turbine exhaust duct, NO forma-
tion is favored almost exclusively over NO2 formation.
Turbine exhaust NOx is typically 95% NO and 5% NO2.
In the high-temperature zone, NO2 dissociates to NO by
the mechanism of
NO 2 + O + Heat → NO + O 2
However, after the TEG exits the hot zone and enters
the cooling zone at the boiler tubes, reaction slows down
and the NO2 is essentially fixed. At the stack outlet, the
entrained NO is slowly oxidized to NO2 through a com-
FIGURE 15.18 plex photochemical reaction with atmospheric oxygen.
Flow patterns around flame stabilizer. (From Baukal, C.E. (ed.), The The plume will be colorless unless the NO2 increases to
John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
about 15 ppm, at which time a yellowish tint is visible.
also exempt such as methane, ethane, methylene chlo- Care must be taken in duct burner design because NO
ride, methyl chloroform, and other minor chemicals. can also be oxidized to NO2 in the immediate post-flame
In order to accurately predict emission, kinetic equa- region by reactions with hydroperoxyl radicals:
tions are created using first-order equations for oxida-
tion in the general form of NO + HO 2 → NO 2 + OH
d ( Chemical )
= − K [ O 2 ][ Chemical ] (15.1) if the flame is rapidly quenched. This quenching can
dt occur because of the large quantity of excess TEG com-
monly present in duct burner applications. Conversion
where to NO2 may be even higher at fuel turndown conditions
where the flame is smaller and colder. NO2 formed in
K = Ae[ − E/RT ] (15.2) this manner can contribute to “brown plume” problems
and may even convert some of the turbine exhaust NO
and to NO2.
A is the pre-exponential factor/frequency factor in There are two principal mechanisms through which
appropriate units nitrogen oxides are formed:
R is the universal gas constant in appropriate units
T is the absolute temperature in Kelvin 1.
Thermal NOx: The primary method is thermal
E is the activation energy, usually listed in kcal/mol oxidation of atmospheric nitrogen in the TEG.
NOx formed in this way is called thermal NOx.
As the temperature increases in the combustion
15.7.1 Visible Plumes zone and surrounding environment, increased
Stack plumes are caused by moisture and impurities in amounts of N2 from the TEG are converted to
the exhaust. Emitted NO is colorless and odorless, and NO. Thermal NOx formation is most predomi-
NO2 is brownish in color. If the NO2 level in the flue nant in the peak temperature zones of the flame.
gas exceeds about 15–20 ppm, the plume will take on 2.
Fuel-bound nitrogen NOx: The secondary method
a brownish haze. NOx also reacts with water vapor to utilized to form NOx is the reaction of oxygen
form nitrous and nitric acids. Sulfur in the fuel may oxi- with chemically bound nitrogen compounds
dize to SO3 and condense in the stack effluent, causing a contained in the fuel. NOx formed in this man-
more persistent white plume. ner is called fuel NOx. Large amounts of NOx
Duct Burners 411
can be formed by fuels that contain molecularly 1600°F (870°C)—than process burners or fired boilers.
bound nitrogen (e.g., amines and mercaptans). The high-temperature zones in the duct burner flames
If a gaseous fuel such as natural gas contains are smaller due to large amounts of flame quenching
diluent N2, it simply behaves as atmospheric by the excess TEG. Finally, mixing is rapid and there-
nitrogen and will form NOx only if it disassoci- fore retention time in the high-temperature zone is
ates in the high-temperature areas. However, if very brief.
the gaseous fuel contains, for example, ammo- The same duct burner, when used to heat atmospheric
nia (NH3), this nitrogen is considered bound. air, is no longer considered “low NOx,” because the peak
In the low concentrations typically found in flame temperature approaches the adiabatic flame tem-
gaseous fuels, the conversion to NOx is close perature in air.
to 100% and can have a major impact on NOx Clearly, operating conditions have a major impact on
emissions. NO formation during combustion. To properly assess
NOx production levels, the overall operating regime
Bound nitrogen in liquid fuel is contained in the long must be considered, including TEG composition, fuel
carbon chain molecules. Distillate oil is the most com- composition, duct firing temperature, and TEG flow
mon oil fired in duct burners as a liquid fuel. The fuel- distribution.
bound nitrogen content is usually low, in the range of
0.05 weight percent. Conversion to NOx is believed to be
80%–90%. For No. 6 oil, containing 0.30 weight percent 15.7.3 CO, UBHC, SOx, and Particulates
nitrogen, the conversion rate to NOx would be about 15.7.3.1 Carbon Monoxide
50%. Other heavy waste oils or waste gases with high
concentrations of various nitrogen compounds may add Carbon monoxide (CO), a product of incomplete com-
relatively high emissions. Consequently, fuel NOx can bustion, has become a major permitting concern in gas
be a major source of nitrogen oxides and may predomi- turbine-based cogeneration plants. Generally, CO emis-
nate over thermal NOx. sions from modern industrial and aero-derivative gas
The impact of temperature on NOx production in turbines are very low, in the range of a few parts per
duct burners is not as pronounced as in, for example, million (ppm). There are occasional situations in which
fired heaters or package boilers. One reason is that CO emissions from the turbine increase due to high
both the bulk fired temperature and the adiabatic rates of water injection for NOx control or operation at
flame temperature are lower than in fired process partial load, but the primary concern is the sometimes
equipment. large CO contribution from supplementary firing. The
In the formation of NOx, the equations are similar to same low-temperature combustion environment that
formation of thermal NOx and are presented as follows: suppresses NOx formation is obviously unfavorable for
complete oxidation of CO to CO2. Increased CO is pro-
duced when fuels are combusted under fuel-rich con-
d ( NO ) ditions or when a flame is quenched before complete
= 2 Ae( − E/RT ) ( O 2 )eq ( N 2 ) (15.3)
dt burnout. These conditions (see Figure 15.19) can occur
and 1200
k0
(O 2 )eq = (O 2 )0eq.5 (15.4) CO emissions are
(RT )0.5
CO emissions, ppm
depressed by higher
oxygen content in the
TEG and with lower
One generally accepted practice is to assume (O2) in (25–75 fps) TEG velocities
equilibrium with (O) and (O2) concentration using the
Westenburg3 results for k0 for (O2) equilibrium and
Zeldovich constants, A, E, as measured by Bowman.4
When used to provide supplementary firing of tur-
bine exhaust, duct burners are generally considered 0
500 1100
to be “low NOx” burners. Because the turbine exhaust TEG temperature, °F
contains reduced oxygen, the peak flame temperature
is reduced and the reaction speed for O2 and N+ to form FIGURE 15.19
NOx is thus lowered. The burners also fire into much Effect of conditions on CO formation. (From Baukal, C.E. (ed.), The
lower average bulk temperatures—usually less than John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
if there is poor distribution of TEG to the duct burner, always be present in trace quantities, regardless of how
which causes some burner elements to fire fuel-rich and the HRSG system is operated.
others to fire fuel-lean, depending on the efficiency of For HC and VOC incineration several sources are
the TEG distribution device. The factors affecting CO available such as Barnes.6
emissions include
In general,
• TEG distribution
d(C aHb ) moles
• Low TEG approach temperature = −5.52(108 )P −0.815Te12 , 200/T (C aHb )0.5 (O 2 )
dt cm 3 s
• Low TEG oxygen content
(15.6)
• Flame quench on “cold” screen tubes
• Improperly designed flame holders that allow
flame quench by relatively cold TEG
15.7.3.3 Sulfur Dioxide
• Steam or water injection
Sulfur dioxide (SO2) is a colorless gas that has a charac-
For utilization, and performance prediction, kinetic teristic smell in concentrations as low as 1 ppm. SO2 is
data can be utilized from the literature. For instance, formed when sulfur (S) in the fuel combines with oxygen
for CO destruction, several kinetic data are available (O2) in the TEG. If oxygen is present (from excess of com-
such as5 bustion) and the temperature is correct, the sulfur will
further combine and be converted to sulfur trioxide (SO3).
These oxides of sulfur are collectively known as SOx.
2
d[CO]  25, 000  0.5  P 
Except for sulfur compounds present in the incoming
= −1.8107 e  −  (CO)(O 2 ) (H 2O) 
0.5
 particulate matter (PM), all of the sulfur contained in
dt  RT RT 
the fuel is converted to SO2 or SO3. Sulfur dioxide will
(15.5) pass through the boiler system to eventually form the
familiar “acid rain” unless a gas-side scrubbing plant
Most published CO rates involve H2O because CO is installed. Sulfur trioxide can, in the cooler stages of
destruction requires the (OH)−1 radical to produce the the gas path, combine with moisture in the exhaust gas
reaction. to form sulfuric acid (H2SO4), which is highly corrosive
and will be deposited in ducts and the economizer if the
exhaust gas is below condensing temperatures. Natural
15.7.3.2 Unburned Hydrocarbons gas fuels are fortunately very low in sulfur and do not
In the same fashion as carbon monoxide generation, usually cause a problem. However, some oil fuels and
UHCs are formed in the exhaust gas when fuel is burned plant gases can be troublesome in this respect.
without sufficient oxygen, or if the flame is quenched
before combustion is complete. UHCs can consist of 15.7.3.4 Particulate Matter
hydrocarbons (defined as any carbon–hydrogen mol-
ecule) of one carbon or multiple carbon atoms. The Particulate emissions are formed from three main
multiple carbon molecules are often referred to as long- sources: ash contained in liquid fuels, unburned carbon
chain hydrocarbons. UHCs are generally classified in in gas or oil, and SO3. The total amount of particulate
two groups: is often called TSP (total suspended particulate). There
is concern for the smaller sized portion of the TSP, as
1. UHCs as methane this stays suspended in air for a longer period of time.
The PM-10 is the portion of the total PM that is less
2. Non-methane hydrocarbons or VOCs
than 10 μm (1 × 10−6 m) in size. Particles smaller than
PM-10 are on the order of smoke. Typical NOx and CO
The reason for the distinction and greater concern for emissions for various fuels are shown in Table 15.1.
VOCs is that longer chain hydrocarbons play a greater For particulate oxidation, an equation can be devel-
role in the formation of photochemical smog. VOCs are oped from fundamental principles, utilizing a combi-
usually defined as molecules of two carbons or greater, nation of diffusion of oxygen and surface reactivity as
and are sometimes considered to be three carbons or follows:
greater. These definitions are set by local air quality con-
trol boards and vary across the United States.
dm (12Cog Ap )
UHCs can only be eliminated by correct combustion = (15.7)
of the fuel. However, hydrocarbon compounds will dt (1/K m + 1/K r )
Duct Burners 413
TABLE 15.1 dn
Typical NOx and CO Emissions from Duct Burners ∑ dt ρ∆v = nρ(V ⋅ a) (15.9)
Gas NOx (lb/106 Btu Fired) CO (lb/106 Btu Fired)
Natural gas 0.1 0.08
This method can be very useful for fully mixed down-
Hydrogen gas 0.15 0.00
stream products even with coarse volumes. But one
Refinery gas 0.1–0.15 0.03–0.08 must be careful with coarse volumes to be sure that the
Plant gas 0.11 0.04–0.01 temperature and concentrations are uniform.
Flexicoker gas 0.08 0.01
Blast furnace gas 0.03–0.05 0.12
Producer gas 0.05–0.1 0.08
Syn fuels 0.08–0.12 0.08
Propane 0.14 0.14 15.8 Maintenance
Butane 0.14 0.14 1.
Normal wear and tear: If nothing has been
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC replaced in the past 5 years and the burner (or
Press, Boca Raton, FL, 2001. turbine/HRSG set) is operated fairly continu-
Note: NOx emissions from butane and propane can be modified by ously, it is likely that some tips and wings may
direct steam injection into a gas or burner flame. CO emissions are
require replacement.
highly dependent on TEG approach temperature and HRSG fired
temperature. 2.
Damage due to misuse, system upsets, or poor main-
tenance practices: Older systems designed with-
where out sufficient safety interlocks (TEG trip, high
m is the mass of particle temperature) sometimes expose parts to exces-
t is the time sively high temperatures, which results in wing
C is the molar density warpage and oxidation failure.
A is the surface area 3.
Fuel quality/composition: Some refinery fuels or
Km is the diffusion coefficient of oxygen in nitrogen waste fuels contain unsaturated components
Kr is the reaction coefficient of the form Ae−E/RT, A is and/or liquid carryover. Eventually, these com-
the frequency factor, E is the activation energy, R pounds will form solids in the runner pipes or
is the universal gas constant, T is the temperature directly in tips, which results in plugging.
The equation can be integrated for constant density par- The following are some items to consider when opera-
ticles and using particle tracking in time steps with con- tional problems are encountered:
stant or varying oxygen and temperature. An excellent
source of char rate data is available by Smith and Smoot.7 • Plugged gas ports: Which are evidenced by gaps
Then, in all cases, one can post-process thermal map in the flame or high fuel pressure: Gas ports
data in some discrete volume form and/or insert into a may simply consist of holes drilled into the
CFD code using the Rayleigh flux theorem as follows: element manifold pipe, or they may be located
in individual removable tips. Designs of the
cv cs
∂ former type may be redrilled or else the entire
∂t ∫ ∫
n ρ dv = nρ( V ·da) (15.8) manifold pipe must be replaced. Discrete tips
can be replaced individually as required.
where • Warped flame holders (wings): Some warping is nor-
n is the chemical in mass units mal and will not affect flame quality, but excessive
t is the time deformation such as “curling” around the gas ports
ρ is the density will degrade the combustion and emission per-
v is the volume formance. Most grid-type burner designs permit
a is the area replacement of individual flame holder segments.
V is the velocity vector • Oxidation of flame holders (wings) or portions
of flame holders: If more than one-third of the
where described in words, the formation of (n) through flame holder is missing, it is a good candidate
the volume surface is equal to the integrated rate of for- for replacement. Fabricated and cast designs
mation over the control volume. are equally prone to oxidation over time. Most
It is a simple extrapolation to extend this concept for grid-type burner designs permit replacement of
even coarse volumes as follows: individual flame holder segments.
Vent to Vent to
atmosphere atmosphere
V1 V4
PI PI PSL PSH PI
Gas To main
FM
supply burner
V1 V2 V3 V3 V5
ST
To lgnition
system
(see Figure 15.22)
FM = Flowmeter V1 = Manual shutoff valve
PI = Pressure gauge V2 = Pressure regulator (optional)
PSH = High pressure interlock V3 = Main burner safety shutoff valve
PSL = Low pressure interlock V4 = Main burner shutoff atmospheric vent valve
ST = Cleaner or strainer V5 = Main flow control valve
FIGURE 15.20
Typical main gas fuel train: single element or multiple elements firing simultaneously. (From Baukal, C.E. (ed.), The John Zink Combustion
Vent to Vent to
V4 V8
PSH To
main
(Optional burner
location) PSL V7
PI PSL
PI
Gas
FM
supply
To
V1 V2 V3 V3 V5 other
main
burners
To ignition
system
(see Figure 15.23)
V6
FM = Flowmeter V4 = Main burner header shutoff atmospheric vent valve

PI = Pressure gauge V5 = Main flow control valve
PSH = High pressure interlock V6 = Main flow bypass control valve (optional)
PSL = Low pressure interlock V7 = Individual burner safety shutoff valve
V1 = Manual shutoff valve V8 = Main burner header charging atmospheric vent valve
V2 = Pressure regulator (optional) (optional)
V3 = Main safety shutoff valve
FIGURE 15.21
Typical main gas fuel train: multiple elements with individual firing capability. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook,
Duct Burners 415
Vent to
atmosphere
V4
PI
Gas
To igniter
supply
V1 V2 V3 V3
PI = Pressure gauge
V1 = Manual shutoff valve
V2 = Igniter flow control valve
V3 = Igniter safety shutoff valve
V4 = Igniter shutoff atmospheric vent valve
FIGURE 15.22
Typical pilot gas train: single element or multiple elements firing simultaneously. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook,
Vent to Vent to
V4 V8
To
(Optional PI igniter
location) (typical)
V7
PSL
Gas
supply
V1 V2 V3 V3 To
other
igniters
PSH
(permanently
installed)
PSL
PI = Pressure gauge V3 = Igniter header safety shutoff valve

PSH = High pressure interlock V4 = Igniter supply atmospheric vent valve
PSL = Low pressure interlock V7 = Individual igniter safety shutoff valve
V1 = Manual shutoff valve V8 = Igniter header atmospheric vent valve (optional)
FIGURE 15.23
Typical pilot gas train: multiple elements with individual firing capability. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
• Severe sagging of runner pipes (grid design only): (DCS). At a minimum, the duct burner management
If the manifold pipe is no longer supported system should include the following:
at both ends, it should be replaced. Beyond
that relatively extreme condition, sagging at • Flame supervision for each burner element
midspan in excess of approximately 2–3 in. • Proof of completed purge and TEG/combustion
(5–7 cm) should be corrected by runner airflow before ignition can be initiated
replacement and/or installation of an auxil- • Proof of pilot flame before main fuel can be
iary support. activated
• Automatic fuel cutoff upon detection of flame
failure, loss of TEG/combustion air, and high or
15.8.1 Accessories
low fuel pressure
15.8.1.1 Burner Management System
Other interlocks designed to protect downstream equip-
All fuel-burning systems should incorporate controls ment can also be included, such as high boiler tube tem-
that provide for safe manual light-off and shutdown, perature or loss of feed water.
as well as automatic emergency shutdown upon detec-
tion of critical failures (see Chapter 12). Control logic
15.8.1.2 Fuel Train
may reside in a packaged flame safeguard module, a
series of electromechanical relays, a programmable Fuel flow to the burners is controlled by a series of valves,
logic controller (PLC), or a distributed control system safety devices, and interconnecting piping mounted on
Oil
return
V13
PI PSL TI TSL PI
ST
Oil
supply FM
V1 V5 V3 V3 V9
Scavenging
To
medium
main
V9 V10
burner
PSL PDS PI
Atomizing ST
medium
supply
V1 TR V12 V9 V11
FM = Flowmeter V1 = Manual shutoff valve

PI = Pressure gauge V3 = Main burner safety shutoff valve
PDS = Differential pressure alarm and trip interlock V5 = Main flow control valve
PSH = High pressure interlock V9 = Check valve
PSL = Low pressure interlock V10 = Scavenging valve
TI = Temperature gauge V11 = Atomizing medium individual burner
(optional for unheated oil) shutoff valve, automatic
TSL = Low temperature or high viscosity alarm V12 = Differential pressure control valve
(optional for unheated oil) V13 = Re-circulating valve
ST = Cleaner or strainer (optional for unheated oil)
TR = Trap
FIGURE 15.24
Typical main oil fuel train: single element. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
Duct Burners 417
Oil
return
V13
(Optional
location)
PI PSL TSL PSL PI
ST
Oil
supply FM
V1 V3 V5 V7 V9
Scavenging
To other medium To
V3a V6 main V9 V10 main
burners burner
(typical)
PDS PI
Atomizing
medium
supply Steam or air header
TR V12 V11a V9 V11
FM = Flowmeter V5 = Main flow control valve

PI = Pressure gauge V6 = Main flow by-pass control valve (optional)
PDS = Differential pressure alarm and trip interlock V7 = Individual burner safety shutoff valve
PSL = Low pressure interlock V9 = Check valve
TSL = Low temperature or high viscosity alarm V10 = Scavenging valve
(optional for unheated oil) V11 = Atomizing medium individual burner shutoff
ST = Cleaner or strainer valve, automatic
TR = Trap V11a = Atomizing medium header shutoff valve,
V1 = Manual shutoff valve automatic (alternate to “V11”)
V3 = Main safety shutoff valve V12 = Differential pressure control valve
V3a = Circulating valve (optional for unheated oil) V13 = Re-circulating valve (optional for unheated oil)
FIGURE 15.25
Typical main oil fuel train: multiple elements. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
PSL PI
Light ST
oil
supply
V1 V6 V3 V7 V9
To
Scavenging
igniter
To other medium
(typical)
igniters V9 V10
(permanently
installed)
PI = Pressure gauge
PSL = Low pressure interlock
ST = Cleaner or strainer
V1 = Manual shutoff valve
V3 = Igniter safety shutoff valve
V7 = Individual igniter safety shutoff valve
V9 = Check valve
V10 = Scavenging valve
FIGURE 15.26
Typical pilot oil train: single element. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
PSL PI
Light ST
oil
supply
V1 V6 V3 V7 V9
To
Scavenging
igniter
medium
To other (typical)
igniters V9 V10
(permanently
installed)
PDS
Steam
or air
V12 V9
PI = Pressure gauge V1 = Manual shutoff valve

PDS = Differential pressure alarm and trip interlock V3 = Igniter safety shutoff valve
PSL = Low pressure interlock V6 = Igniter flow control valve
ST = Cleaner or strainer V7 = Individual igniter safety shutoff valve
V9 = Check valve
V10 = Scavenging valve
V12 = Differential pressure control valve
FIGURE 15.27
Typical pilot oil train: multiple elements. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
a structural steel rack or skid. A properly designed fuel national standards. Specific requirements for burner
train will include, at a minimum, the following: safety systems are included, but as stated in the foreword,
NFPA 8506 does not encompass specific hardware appli-
• At least one manual block valve cations, nor should it be considered a “cookbook” for the
• Two automatic block valves in series design of a safe system. Prior to the issuance of NFPA
• One vent valve between the automatic block 8506, designers often adapted NFPA boiler standards to
valves (gas firing only) HRSGs, which resulted in design inconsistencies.
• Flow control valve 15.9.2 Factory Mutual
• High and low fuel pressure switches
An insurance underwriter that publishes guidelines
• Two pressure gages, one each at the fuel inlet on combustion system design; Factory Mutual (FM)
and outlet also “approves” specific components such as valves,
pressure switches, and flame safeguard equipment
Depending on the custom and operating require-
that meet specific design and performance standards.
ments at a particular plant, pressure regulation, flow
Manufacturers are given permission to display the FM
measurement devices, and pressure transmitters can
symbol on approved devices. Although FM approval
also be incorporated. See Figures 15.20 through 15.27 for
may be required for an entire combustion control sys-
typical duct burner fuel system piping arrangements.
tem, it is more common for designers to simply specify
the use of FM-approved components.
15.9.3 Underwriters’ Laboratories

15.9 Design Guidelines and Codes Well known in the United States for its certification
of a broad range of consumer and industrial electri-
15.9.1 NFPA 8506 (National Fire
cal devices, underwriters’ laboratories (UL) authorizes
Protection Association)
manufacturers to display their label on specific items
First issued in 1995, this standard has become the de that have demonstrated compliance with UL stan-
facto guideline for HRSGs in the United States and dards. Combustion system designers will frequently
many other countries that have not developed their own require the use of UL-approved components in burner
Duct Burners 419
management systems and fuel trains. Approval can

also be obtained for custom-designed control systems,
References
although this requirement generally applies only to a
few large cities and a few regions in the United States. 1. S. Londerville, Performance prediction of duct burner
systems via modeling and testing, Chapter 26 in Industrial
Combustion Testing, C.E. Baukal (ed.), CRC Press, Boca
15.9.4 ANSI B31.1 and B31.3 (American Raton, FL, 2011.
National Standards Institute) 2. Department of Energy, Improving industrial burner
These codes address piping design and construction. designs with computational fluid dynamic tools:
Progress, Needs and R & D priorities, Workshop Report,
B31.1 is incorporated in the NFPA 8506 guideline, while
September 2002.
B31.3 is generally used only for refining/petrochemical
3. A.E. Westenberg, Turbulence modeling for CFD, Combust.
applications. Sci. Technol., 4, 59–67 (1971).
4. C.T. Bowman, Kinetics of pollution formation and
15.9.5 Others destruction in combustion, Prog. Energy Combust. Sci., 1,
33–45 (1975).
The following may also apply to duct burner system 5. G.C. Williams, H.C. Hottel, and A.C. Morgan, The com-
designs, depending on the country where equipment bustion of methane in a jet-mixed reactor, 12th Symposium
will be operated: (International) on Combustion, The Combustion Institute,
Pittsburgh, PA, p. 913, 1969.
• National Electrical Code (NEC) 6. R.H. Barnes, M.H. Saxton, R.E Barrett, and A. Levy,
• Canadian Standards Association (CSA) Chemical Aspects of Afterburner Systems, April 1979,
EPA report EPA-600/7-79-096, NTIS PB298465
• International Electrotechnical Commission (IEC) Page 21.
• European Committee for Electrotechnical 7. D.L. Smoot and P. Smith, Coal Combustion and
Standardization (CENELEC) Gasification, Plenum Press, New York, pp. 81–88, 1985.
16
Air Heaters
Nigel Webley
CONTENTS
16.1 Introduction.................................................................................................................................................................. 421
16.2 Applications.................................................................................................................................................................. 421
16.3 Design Principles......................................................................................................................................................... 421
16.4 Types.............................................................................................................................................................................. 423
16.4.1 Twin Shell, Twin Inlet Air Heater................................................................................................................. 423
16.4.2 Twin Shell, Twin Inlet, Reverse Flow Combustion Air.............................................................................. 424
16.4.3 Twin Shell, Single Inlet, Reverse Flow.......................................................................................................... 424
16.4.4 All-Metal Air Heater....................................................................................................................................... 424
16.4.5 Air Heaters with Low Heating Value Fuels................................................................................................. 425
16.4.6 Pressurized Air Heaters.................................................................................................................................. 426
16.5 Duct Burner as Air Heater.......................................................................................................................................... 427
16.6 Indirect-Fired Air Heaters.......................................................................................................................................... 428
16.7 Solid Fuel–Fired Air Heaters...................................................................................................................................... 428
16.8 Instrumentation and Control Systems...................................................................................................................... 428
16.9 Refractory Issues.......................................................................................................................................................... 429
16.10 Emissions...................................................................................................................................................................... 430
to be heated. Typical applications include the drying

of detergents, minerals, fertilizers, animal feeds, and
16.1 Introduction
coal. Air heaters are also used as an indirect source of
Air heaters are used in a wide range of industries—for heat for drying of foodstuffs such as milk powder and
preheating of process heaters in the petrochemical and dried potatoes, roasting of coffee beans, and production
refining industries; for detergent, coal, or fertilizer dry- of food-grade chemical additives. Heaters are supplied
ing; and for other heating and drying applications, for with spray dryers, rotary dryers and kilns, and fluid-
example, in the chemical, soap, paper, food, and cement ized bed dryers. Pressurized air heaters are used as a
industries. The capacity range is from 1 to 100 MW source of heat for preheating process equipment such as
(3 to 340 MM Btu/h) with outlet temperatures typically fluid catalytic cracker units (FCCU) in oil refining appli-
from 200°C to 1000°C (392°F to 1800°F). Direct-fired air cations. For combustion of some gases and liquids, an
heaters are suitable for use with various fuels includ- air heater is sometimes used as a precombustor before
ing residual fuel oil and diesel oil, a wide range of a boiler.
gases including natural gas, liquefied petroleum gas
(LPG), refinery gas, coke oven gas, and blast furnace
gas, and coal.
16.3 Design Principles

The simplest and least expensive form of an air heater
is a refractory-lined combustion chamber with a burner
16.2 Applications
and dilution air entry at one end and a hot gas exit at the
Air heaters are used mainly for high-temperature dry- other. This type of system is generally suitable for appli-
ing processes where the products of fuel combustion cations requiring a high outlet temperature (<800°C ≈
diluted by air are contacted directly with the product 1500°F). The refractory thickness has to be sufficient
421
Combustion air inlet
Burner
Hot
gases
outlet
Fuel
Dilution air inlet
FIGURE 16.1
Peabody twin shell air heater.
for the anticipated heat losses and to maintain a suit- the dryer exhaust temperature. For a given flow rate,
able temperature at the outside surface. The main disad- if the dryer outlet temperature starts to fall, it means
vantage of this type of system is that the dilution air is that insufficient heat is being provided to the process,
introduced close to the burner and can have a quench- so the air heater burner firing rate should be increased
ing effect on the flame. accordingly.
The Peabody twin shell air heater (see Figure 16.1) is If the burner firing rate increases, the outlet tempera-
an inherently efficient design with an air-cooled com- ture at the air heater will increase and this will result
bustion chamber that results in a low shell tempera- in an increase in the dilution airflow rate to maintain
ture. The twin shell air heater typically consists of a a constant temperature at the outlet. In effect, the mass
refractory-lined combustion chamber with an axially flow rate of hot gases to the drying process is increased,
mounted burner. Products of combustion from the but maintained at a constant temperature.
burner are mixed with dilution air and supplied to the The design basis for any air heater is the mass flow
process at a constant temperature. The most important rate of “hot air” to be supplied at a given temperature.
consideration is that the dilution air is introduced sepa- The temperature rise of the required airflow is calcu-
rately, well-downstream of the burner and therefore lated, and with the specific heat of air over the tempera-
cannot chill the flame and cause combustion problems. ture range, the fuel quantity required for the heat input
The separation of the dilution air and the combus- can be calculated. The fuel quantity is part of the mass
tion air allows the working temperature of the combus- flow rate of hot air, so, through a process of iteration,
tion chamber to be fixed independently of the final gas the airflow can be calculated. A proportion of the air is
exit temperature. For most applications the combustion supplied as combustion air, with the remainder being
chamber temperature will be considerably higher than supplied as dilution air.
the air heater exhaust temperature. It should be noted Air heaters are often installed horizontally. For dry-
that a minimum temperature of about 1200°C (2200°F) ers, the required warm air inlet connection is often at
should be maintained in the combustion space for sta- the top of the tower. In this case, a vertical air heater
ble, clean combustion. is supplied and arranged for up-firing. This con-
The refractory-lined combustor is designed to accom- figuration has the advantage of minimizing space
modate the flame. The typical flame length to diame- requirements. Where air heaters are required for sup-
ter ratio is about 1.8:1, but this can vary depending on plying hot gases to a rotary kiln, a ducting, sometimes
burner type, air-side pressure loss (burner pressure known as the upstand, is provided to act as a transition
drop), and fuel. The burner excess air level is typically between a horizontal air heater mounted at the ground
about 50%, which results in a high flame temperature. level and the inlet to the rotary kiln that may require
The burner firing rate is usually controlled by the an angled flexible connection at a high level. The duct-
demand from the process. This can be determined from ing also serves to ensure that the burner products of
Air Heaters 423
combustion and dilution air are well mixed and are at in the air heater is controlled by varying the speed of the
a uniform temperature before entering the process. ID fan or by operating an inlet damper at the fan.
When the air heater is located very close to the dryer The remainder of the air, the dilution air, passes
inlet, it may be necessary to include a static mixing around the annular space between the refractory-lined
device such as a choke ring to ensure good mixing of combustion chamber and the outer shell, hence the
hot gases and dilution air. The air heater can also be terminology “twin shell.” The dilution air mixes with
arranged for down-firing. This would be necessary for the hot products of combustion at the air heater outlet,
coal or other solid fuel–fired applications. giving the desired final temperature. The twin shell
Where there is need to get control of the air heater design may be used when the dilution air is already
mass flow or temperature over a wide range or where preheated or contains recycled gases at temperatures
it is undesirable to start up the air heater before the up to 200°C (390°F), otherwise a reverse flow air heater
process is fully operational, a start-up stack would be is preferred.
specified. In detergent drying applications worldwide, Combustion air and dilution air are supplied by sep-
residual fuel oil is often used as the local fuel of choice arate fans or blowers. Combustion air is supplied at a
where natural gas is not available and the cost of LPG pressure high enough to overcome the burner pressure
is excessive. In the best regulated applications where drop, while the dilution air is supplied at a relatively
the atomizing steam quality is good (dry and at suffi- low pressure but in a much larger quantity.
cient flow rate and pressure) and the fuel oil is hot and Dilution air keeps the inner shell cool, and conse-
at the required viscosity, there is inevitably a “puff” quently a minimal amount of refractory is required in
of smoke at start-up. The presence of soot particles in the combustor. The air-cooled shell means that heat
detergent is undesirable because it contaminates the losses are minimized and the shell temperature is
product, and therefore a start-up stack is always used on not much hotter than the air supply temperature. It is
oil-fired detergent dryers. The start-up stack will have important that an adequate airflow is maintained in the
a diversion damper or twin dampers, which will allow annular space around the inner combustor at all times.
the process connection to be isolated at light-off until If the airflow is too low, the inner shell could be over-
combustion is stabilized and the correct temperature is heated and may deform. The capacity range is typically
attained. It also allows operation of the process dryer at limited to about 3 to 1. Sometimes static mixing devices
a reduced hot-gas flow rate until operating conditions such as swirl plates are fitted in the annular space to
are stabilized. In normal operation, the start-up stack is improve the heat transfer when the airflow is low. The
isolated and the full gas flow rate from the air heater distance between the inner shell and the outer shell has
passes to the process. to be at least 25 mm (1 in.) for practical reasons and to
avoid potential hot spots where the shells may touch.
The dilution air inlet is preferably radial and sized large
enough to distribute air around the full circumference
of the air heater.
This type of air heater is typically used for spray dry-
16.4 Types ers and rotary dryers to dry detergents. The advantage
of the twin shell design is the lightweight refractory
16.4.1 Twin Shell, Twin Inlet Air Heater
construction and the overall efficiency, as heat losses
The most common type of air heater is the Peabody through the refractory are recovered in the dilution
twin shell design. This type of air heater is used for con- airstream.
ventional drying applications where the inlet air stream Drying processes inevitably produce a lot of dust, and
is at a low temperature, less than 100°C (212°F), and it is to be expected that some dust will be introduced
the required outlet temperature is no more than 800°C through the fans into the burner and air heater. The fan
(1500°F). air intakes should be positioned as far away as possible
The combustion chamber is refractory lined. The axi- from any sources of dust. Dust buildup in the burner
ally mounted burner operates with an excess air level of can cause combustion problems, so regular inspection
typically about 50%, giving a high flame temperature and cleaning is recommended. Dust buildup in the dilu-
in the combustion chamber. The burner is typically a tion air annulus can cause blockages and hot spots and,
swirl-stabilized register-type burner suitable for gas, therefore, in some applications provision is made for
oil, or in dual fuel configuration. Combustion air is typi- access for inspection and ease of cleaning. Rotary burn-
cally supplied at 15–25 mbar (6–10 wg) above the operat- ers or other burners with moving parts are not suited
ing pressure in the air heater, which is usually close to to a dusty environment and, if applied, will require
atmospheric pressure. The dryer or process usually has frequent cleaning if combustion and other operational
an induced draft (ID) fan downstream and so the draft problems are to be avoided.
16.4.2 Twin Shell, Twin Inlet, Reverse of the air heating/drying process. Recycle of exhaust
Flow Combustion Air gases may be inadvisable if the gases are very dusty or
contaminated.
This type of air heater is a variant of the standard type,
designed to reheat hot or corrosive gases that may have
been recycled from the drying process (see Figure 16.2). 16.4.3 Twin Shell, Single Inlet, Reverse Flow
The twin shell design means that the efficiency advan- This type of air heater is used to achieve high outlet
tages of the twin shell design can be retained but the shell temperatures, typically more than 1000°C (1800°F) (see
cooling is provided by the use of the combustion air. Figure 16.3). Dilution air or quench air is not required,
This design is preferred when hot and/or dirty gases and therefore all of the air passes through the burner.
are used instead of ambient air for quenching. In a The combustion air cools the inner shell and recov-
standard twin shell design, dirty gases may result in ers some preheat. In this type of air heater, the burner
deposits of particulates in the annular space around the excess air level is likely to be higher than 50%, but there
combustor. Similarly, a gas containing water vapor or will be a limitation on the maximum excess air level for
acid gas may condense on the cool outer surfaces if it is burner stability.
cooled below the dew point. In this design variant, there
is a relatively short path from the recycle gas stream to
16.4.4 All-Metal Air Heater
the air heater inlet. Where the recycle gas is already sup-
plied at a high temperature, the quantity required may This type is another variant of the standard design, but
be lower and the important requirement for cooling the the use of refractory is restricted to the burner throat
inner shell may be affected. (see Figure 16.4). The inner shell is cooled by the dilution
This type of air heater is often used in coal drying airflow in the annular gap.
applications where there is a further requirement to con- The main advantage of this type of air heater is the
trol the oxygen concentration of the hot gases in direct weight, cost, and size savings due to the elimination
contact with the coal to avoid spontaneous combustion. of the refractory lining. This type of air heater can be
The dryer exhaust gases are constantly recycled until installed, for example, on the roof of a building. The
the oxygen concentration is similar to the oxygen con- inner air-cooled shell will be fabricated from stainless
centration in the air heater products of combustion, typ- steel. Careful consideration has to be given to the annu-
ically less than 6%. lus air velocity and distribution of air in the annulus to
In other applications, part of the dryer exhaust gas avoid hot spots.
can be recycled back to the air heater as dilution air All-metal air heaters can only be used for gas and
and reheated by mixing with new products of combus- light oil firing applications with an outlet temperature
tion. This can serve to increase the overall efficiency up to 600°C (1100°F). The all-metal air heater in vertical
Combustion air inlet
Burner
Hot
gases
outlet
Fuel
Recycled gas inlet
FIGURE 16.2
Twin shell, twin-inlet reverse flow combustion air.
Air Heaters 425
Air inlet
Fuel Hot
gas/oil gases
inlet outlet
FIGURE 16.3
Twin shell single-inlet high-temperature air heater.
Dilution
Combustion
air inlet
air inlet
Fuel Hot
gas gases
inlet outlet
FIGURE 16.4
Twin shell twin-inlet all-metal air heater.
configuration is the basis of the Hamworthy Combustion (typically >2 MJ/Nm3 = 54 Btu/ft3). It consists of a volute
AMOxsafe gas combustion unit (GCU), which is used to plenum chamber in which the gas enters tangentially
dispose of excess boil-off gas (BOG) on LNG carriers. In and flows in a single uninterrupted stream until it
these applications, the hot gases are discharged directly leaves through an unobstructed annular exit port into
to atmosphere, but the exhaust temperature is limited to the burner throat. The radial and axial dimensions in
450°C (840°F). the scroll are progressively reduced from the gas inlet
so that the gas passage area is reduced as gas is dis-
charged. In this type of burner, the gas exits through the
16.4.5 Air Heaters with Low Heating Value Fuels
annular gap around the air register as a spinning torus
Air heaters can be designed for a wide range of fuels. In where it mixes with the expanding swirled combustion
the steel industry, blast furnace gas is often an economi- air as it enters the burner throat. The high turbulence
cal fuel. It has a low heating value, but since it is a by- and complete mixing result in rapid combustion with a
product gas, its use saves on expensive primary fuels. low excess air level using minimal gas and combustion
The basic air heater design and operating principle is air pressures. Stable combustion is assured despite the
the same as for other twin shell air heaters, but a low low heating value and the relatively low flame tempera-
heating value burner is applied. ture. The scroll burner is designed for blast furnace gas
The Peabody scroll burner, seen in Figure 16.5, is but can be used with a wide range of other low heating
often used for low pressure, low heating value gases value, low pressure gases. The low heating value burner
Blast furnace gas inlet

A B G F E
Operating quadrant B1
Stainless steel for air valves
helix Gas inlet
Windbox
Air inlet
Oil gun
C D A1
Flexible
hoses
Peephole Ignitor
Ignitor Diffuser
Gas scroll Oil supply
Steam supply Flame eye
FIGURE 16.5
Peabody scroll burner.
uses low pressure gas—typically 15–25 mbar (6–10 wg) In a typical FCCU, a pressurized air heater is required to
gas pressure at the burner. provide a source of hot gases to preheat the unit at start-
An auxiliary fuel such as residual fuel oil, natural gas, up. In normal operation, the process is self-sustaining
or coke oven gas is normally fired as a constant pilot and the burner in the air heater is not used, but an air-
with the low CV burner. This is useful at start-up. flow is maintained through the heater.
The pressurized air heater is normally a single inlet,
single shell design (see Figure 16.6). It is lined internally
16.4.6 Pressurized Air Heaters
with refractory. Heat losses will be somewhat greater
Pressurized air heaters are used as preheaters in process than the twin shell design, but since the air heater is
plants that normally operate at an elevated pressure. only used for a few weeks at start-up, perhaps every
Air inlet
Dilution
air
Combustion
air
Fuel Hot
gas/oil gases
inlet outlet
FIGURE 16.6
Pressurized air heater.
Air Heaters 427
FIGURE 16.7
Pressurized air heater with an internally mounted burner. The air inlet diverter damper mechanism should be noted.
2 or 3 years, this is inconsequential. The single air inlet When operating at an elevated pressure, the flame
is fitted with a diverter damper so that the air can be volume will be reduced in proportion with the ratio of
apportioned between combustion air, which goes to the the absolute pressure. Similarly, the burner size will be
burner, and dilution air, which is introduced annularly reduced due to the smaller volume of air required. Many
around the burner (see Figure 16.7). process licensors specify that the combustion chamber
The burner can either be a high intensity burner or a regis- in a pressurized air heater must be at least double the
ter burner. The advantage of the high intensity burner (like length of the expected flame length. This means that the
the John Zink HI or Hamworthy Combustion Vortomax) aspect ratio of a pressurized air heater may be 3 or 4 to 1.
is that it has a refractory-lined precombustor and, conse- Pressurized air heaters are connected by a flanged
quently, there is excellent stability, even when operating connection or sometimes by a butt weld connection
out of ratio or at high excess air levels that can occur during directly to the FCCU or other process equipment with-
start-up or preheating of the process. The burner has to be out any flexible bellows. They can be oriented vertically
properly constructed for this specific application. or horizontally with a top outlet. For expansion con-
Air has to be provided at a high pressure to overcome cerns, the air heater may be suspended on spring hang-
the operating pressure in the process, but the air-side ers from the structural steelwork or on flexible mounts.
pressure drop is typically in the range of 20–50 mbar Pressurized air heaters are often specified to conform
(8–20 wg), which is relatively small compared to the to specific pressure vessel design codes including ASME
operating pressure of 2–4 barg (30–60 psig). The main VIII Div 1, AD Merkblatter, and BS 5500.
disadvantage of the single air inlet is that it is not easy to
establish the correct split between combustion air and
dilution air. The airflow required for pilot ignition and
the start-up flame at turndown needs to be set carefully,
and the airflow then has to be increased as the firing rate
16.5 Duct Burner as Air Heater
increases. Since the air heater is operated infrequently
and only during start-ups when there are many compet- For gas-fired applications, a variant of the normal
ing demands on the operator, there are often problems type of air heater is the duct burner (see Chapter 15).
with light-up and control of the air heater. This is particularly useful where the requirement is
to reheat or increase the temperature of a stream of This type of system will require an ID fan to draw
warm air, since the burner can be installed in the duct the air heater products of combustion through the heat
carrying the warm air. The main requirement is that exchanger and a blower to deliver the process air-
the airstream should contain sufficient oxygen (typi- stream. The heat exchanger pressure drop is typically
cally more than 18% by volume) to allow completion about 10–15 mbar (4–6 wg) on the air side and the
of combustion. hot-gas side.
Turbine exhaust gas (TEG) heating systems in com-
bined cycle gas turbine (CCGT) applications use duct
burners in this way to recover useful heat from the gas
turbine. The temperature of TEG is often between 500°C
and 600°C (900°F to 1100°F) while leaving the turbine
16.7 Solid Fuel–Fired Air Heaters
and contains a high proportion of oxygen since turbines
operate at a very high excess air level. Combustion of Air heaters can use finely pulverized solid fuels such
fuel at the duct burner reduces the oxygen concentration as coal or sawdust. These are generally arranged in a
of the TEG and increases the temperature of the gases, down-fired configuration to aid ash removal. Coal must
which then pass to a boiler where they can transfer heat have a volatile content greater than 20% and the par-
in the production of steam. ticle size should be 80% less than 80 μm with 50% less
The main advantages of using a duct burner as an air than 50 μm. The ash content must be less than 15%.
heater are the relatively small foot-print and the need The air heater design varies from the traditional twin
for only one fan. The flame is much shorter than that shell design to ensure that excessive temperatures are
produced by a conventional register burner, and there- not attained in the combustion zone that may result
fore the overall length is reduced. The requirement in molten ash on the refractory-lined surfaces. This is
for internal refractory is reduced or eliminated com- achieved through stage-wise introduction of dilution
pletely. A duct burner tends to produce a fairly even air or arranging an air “curtain” between the flame zone
heat distribution. and the refractory-lined combustor in order to “freeze”
the ash in the gas stream.
A solid-fired air heater generally needs a second-
ary fuel to preheat the combustion chamber, to cure or
dry out the refractories prior to commissioning, and to
ignite the coal on start-up and also to stabilize combus-
16.6 Indirect-Fired Air Heaters
tion when the coal is of poor or variable quality.
Direct-fired air heaters supply a stream of warm or hot
air mixed with products of combustion, and therefore
they are not usually suitable for direct drying of food-
stuffs or other clean products, unless used with an out-
let heat exchanger.
16.8 Instrumentation and Control Systems
Indirect-fired air heaters are specified when a fairly
high temperature (more than 200°C ≈ 400°F) is required A twin shell air heater usually has three main control
in the process. Lower temperatures can be obtained loops: the burner firing rate, the product mass flow rate,
with steam heaters or with hot oil systems. and the air heater inlet pressure. These are briefly dis-
The indirect-fired air heater consists of a typical air cussed in this section.
heater, either twin shell or duct burner type, operat- The burner firing rate is set by the demand from the
ing at temperatures up to 800°C (1500°F). The hot gases process. This may be the product outlet temperature,
from the air heater pass to a heat exchanger (shell and product feed rate, or dryer outlet temperature. The
tube or plate type). The process air is heated by indirect demand signal directly acts on burner firing rate. The
contact with the hot flue gas. Temperatures up to 400°C fuel flow rate and the combustion airflow rate are con-
(750°F) can readily be obtained: higher temperatures trolled in ratio, with typically 50% excess air, but this
would require alloy steel construction. Due to the heat may increase on turndown. Due to the high excess air
exchanger design limitations, the air heater gas stream level overall resulting from the effect of dilution air, it
may not be cooled below 200°C (400°F), and therefore the is not necessary to measure the fuel and airflow rates
overall efficiency will be less than 75%. The efficiency and control the ratio precisely, but the air/fuel ratio
can be increased by recycling of the heat exchanger flue should be set up on commissioning and periodically
gas stream back to the air heater dilution air, but it is checked. The fuel/air ratio must never be allowed to go
unlikely that an overall efficiency greater than 92% can sub-stoichiometric and ideally should be maintained
be achieved with an increase in complexity. at a higher level to avoid excessive temperatures.
Air Heaters 429
The product mass flow rate to the process is controlled

by the air heater outlet temperature. If the burner fir-
16.9 Refractory Issues
ing rate is increased, the temperature at the heater out-
let will tend to increase, and so the dilution airflow rate The inner combustor of the twin shell air heater is lined
is increased. The heater outlet temperature (or dryer or with refractory capable of withstanding a high temper-
process inlet temperature) is normally controlled within ature. For gas firing, the hot face temperature may be
a very tight range. If the temperature is too low, the dryer 1650°C (3000°F) or higher and for oil firing, the hot face
cannot operate efficiently. If the temperature is too high, temperature may be in excess of 1700°C (3100°F).
the product may be damaged. For a given temperature, The refractory lining in a twin shell air heater need
the burner firing rate and the dilution airflow rate are only be 115 mm (4.5 in.) thick as the outer surface is
controlled in ratio unless the burner excess air level var- always air-cooled. The advantage of the thin lining
ies. Some applications (e.g., fertilizer drying) use differ- is that there is a very small thermal mass and the air
ent temperatures for different products and then it may heater can therefore respond rapidly to changes in load.
be necessary to analyze the required duty conditions In a pressurized air heater or other single shell design,
and turndown ranges carefully. It is not necessary to the refractory lining may need to be 200 mm (8 in.) or
measure the dilution airflow rate, but knowledge of the thicker to arrive at an acceptable shell temperature and
airflow rate can be helpful in commissioning and opera- to minimize heat losses. In any case, it may be neces-
tions, especially when there is a wide range of required sary to fit external personnel protection in the form of
operating conditions. an expanded mesh or an air-spaced weather-shield in
The operating pressure in the air heater or at the galvanized steel or aluminum.
inlet to the dryer is normally controlled at close to The hot face refractory selection is a high alumina
atmospheric pressure by the ID fan at the dryer or (~60%) firebrick or monolithic lining. Castable or
process outlet. In some applications, it is desirable to gunned monolithic linings are satisfactory in gas-fired
maintain a slight negative pressure close to the prod- applications. For oil- and coal-fired air heaters where
uct feed. The draft is maintained independently of control of the firing range is more limited, a brick lining
the mass flow rate by varying the fan speed with an is preferred as that will not require a complicated dry
inverter or by operation of an inlet vane or fan dis- out schedule with long periods where the heater has to
charge damper. be maintained at a low temperature.
FIGURE 16.8
Low NOx QL burner for air heaters.
The air heater outlet after mixing of the dilution air removal before being discharged to atmosphere, which
and products of combustion will also need to be refrac- may have the effect of absorbing some potential pollut-
tory lined. The extent and thickness of lining in this ants. In some applications (e.g., sugar beet pulp dryers
area will depend on the gas velocity, adequacy of mix- where the pulp is dried for animal feed after the sugar
ing, and the operating temperature. For residual fuel is extracted), NOx emissions at the dryer exhaust are
oil-fired applications where products of combustion often lower than those measured in the air heater.
may contain oxides of sulfur (SOx), it is just equally The typical burner performance will result in NOx
important to ensure that the lining is not too thick and emissions (corrected for dry gas at 3% O2) of between 100
that the shell temperature is maintained well above the and 200 mg/Nm3 for gas firing and 250–500 mg/Nm3 for
dew point (180°C–200°C or 360°F–400°F). oil firing. Low NOx burners may be fitted to air heaters
A radiation shield is required at the outlet end of the and then NOx emissions can be reduced by 50% or more.
annulus of a twin shell air heater to prevent radiation An example of a low NOx burner is the Coen QL burner
from the heater outlet, creating a hot spot on the outer (see Figure 16.8), a burner designed for low NOx at high
shell. This can be achieved by stepping in the heater out- excess air conditions.
let after the end of the inner combustor without unduly Due to the relatively high excess air level, CO emis-
restricting the dilution airflow. sions are not a problem if the flame pattern is designed
in such a way as to not quench the reaction. Common
ways to achieve this are porting additional air around
the burner or using a three zone burner such as a
Veriflame burner with a larger tertiary zone.
The operator should always be aware that incom-
16.10 Emissions
plete combustion of fuel oil (e.g., at start-up or due to
Emissions from direct-fired air heaters are not usually improper operation or lack of maintenance) could result
of great concern. Exhaust gases leaving the dryer may in unburned oil droplets or soot particles, which may
pass through various gas conditioning stages for dust contaminate the product.
17
Marine and Offshore Applications
Richard Price
CONTENTS
17.1 Introduction.................................................................................................................................................................... 431
17.2 Fuels................................................................................................................................................................................. 432
17.2.1 Fuel Oils.............................................................................................................................................................. 432
17.2.2 Low-Sulfur Marine Gas Oil.............................................................................................................................. 432
17.2.3 LNG Boil-Off Gas.............................................................................................................................................. 432
17.2.4 Produced Fuel Gas and Crude Oil.................................................................................................................. 432
17.3 Auxiliary Boiler Applications...................................................................................................................................... 432
17.3.1 Small “Donkey” Auxiliary Boilers.................................................................................................................. 432
17.3.2 Watertube Auxiliary Boilers............................................................................................................................ 433
17.4 LNG Carriers.................................................................................................................................................................. 434
17.4.1 LNG Carrier Main Propulsion Boilers............................................................................................................ 434
17.4.2 LNG Carrier Gas Combustion Units............................................................................................................... 435
17.5 Offshore Applications................................................................................................................................................... 436
17.5.1 FPSOs................................................................................................................................................................... 436
17.5.1.1 Tanker-Converted FPSOs.................................................................................................................. 436
17.5.1.2 FPSO New Boilers............................................................................................................................... 439
17.5.2 FSOs..................................................................................................................................................................... 439
17.5.3 Floating LNG...................................................................................................................................................... 440
17.5.3.1 LNG Re-Gasification Vessels............................................................................................................. 440
17.5.3.2 Floating LNG Production Vessels..................................................................................................... 441
boilers of duty 1–5 m-ton/h (1.1–5.5 U.S. ton/h) are com-

mon on container ships, bulk carriers, general cargo
17.1 Introduction
ships, and ferries. Larger auxiliary boilers of range 10–15
Since the first steam plant was installed on a boat circa m-ton/h (11–16.5 U.S. ton/h) are installed on large cruise
1803, all but a few ships have had a fired boiler of one ships, and on tankers of range 20–50 m-ton/h (22–55 U.S.
type or another. The earliest steam-propelled ships used ton/h) as the cargo (crude oil) pumps are steam turbine
coal as their fuel, and it was not until the early 1900s that driven, using saturated steam at 16–20 barg (240–290 psig).
the steam boilers used fuel oil and the importance of the Marine and offshore applications are wide ranging in
register burner established. Oil-fired, superheated, main application depending on the total steam demand, which
propulsion boilers rated at 60–63 barg (880–930 psig) is composed of turbine drives (pumps, gas compressors,
were applied on ships until the mid-1970s, when the die- and turbo-generators) and heating requirements.
sel engine took over the main propulsion role, primarily All ships are governed and regulated by
due to higher fuel efficiency. These early vessels raised International Maritime Organization (IMO), which is
steam in boilers with the propulsion by screw (propel- responsible for the safety and security of shipping and
ler) or paddle driven by reciprocating steam engines, the prevention of marine pollution, and classification
and later by steam turbines. societies (such as the American Bureau of Shipping—
Diesel engine ships (motor ships) still, however, have ABS, Det Norske Veritas—DNV, Lloyds Register of
a steam requirement, albeit on a much smaller duty. Shipping—LRS), which are responsible for verifying the
The steam is needed for heating duties; hence, auxiliary ship’s design, construction, and maintenance.
431
storage and offloading (FSO) vessels, where regular fuel

oil supplies are not readily available, use or produce
17.2 Fuels
their own fuel from their main process or cargo they
17.2.1 Fuel Oils are handling.
On FPSOs, crude oil taken from the oil field beneath
The stock marine fuel for boilers (and ships’ main
the sea floor is stabilized and reduced in pressure to a
engines) has for many years been residual heavy no. 6
condition where it can be stored in the cargo tanks. This
fuel oil, often referred to as “bunker C,” with viscosity
process results in the light-end gases (methane, ethane,
180–600 cSt at 50°C (122°F). Marine diesel oil is, however,
etc.) separating from the crude oil and thus providing a
always available onboard as the cold-start or back-up
suitable fuel for the vessel’s power and heating require-
fuel, with viscosity of 2–11 cSt at 40°C (100°F). The grade
ments. The produced gas is typically heavier than natu-
and quality of marine fuel oils are governed globally by
ral gas having ethane, propane, and butane of 10%–30%
the ISO standard ISO 8217, which defines limits for vis-
by volume, and on certain installations carbon dioxide
cosities, densities, sulfur, water, ash and flash point, and
up to 50%.
other properties. The heavy-fuel oil is maintained at a
Where produced gas quantities are low, normally on
temperature of 50°C (120°F) in the ship’s fuel tanks, from
heavy crude oil fields, it is common for the produced
where it is pumped and pre-heated to the appropriate
crude oil itself to be fired in steam boilers, as this repre-
temperature to achieve the required viscosity of the main
sents the lowest cost fuel option.
engines or burner fitted to the steam boilers.
Fuel gases and fuel oils with flashpoints lower than
60°C (140°F) (most crude oils) are defined as hazard-
17.2.2 Low-Sulfur Marine Gas Oil
ous fuels by marine classification societies. When these
New marine legislation from 2007 relating to sulfur fuels are taken into safe areas like the engine room, they
dioxide (SO2) emissions from ships has had a signifi- have to be contained within double gas pipes, ventilated
cant effect on marine fuel oils and ships’ engine room enclosures, and use certified hazardous area equipment
systems. This new legislation started in the European and instrumentation.
Union (EU) to reduce SO2 emissions from ships that are
a prime contributor to premature deaths and hospital
admissions. This legislation has resulted in parts of the
seas and main trade routes becoming emission control
areas (ECAs), where the ships must use low-sulfur fuel
(or have SO2 emissions equivalent to using low-sulfur 17.3 Auxiliary Boiler Applications
fuels). At the time of writing, the main ECAs are the
17.3.1 Small “Donkey” Auxiliary Boilers
English Channel (the busiest waterway in the world),
the North Sea, the Baltic Sea, and 200 nautical miles Modern small marine auxiliary boilers for duties of range
around the coasts of the United States. The ship owner 1–5 m-ton/h (1.1–5.5 U.S. ton/h) are typically cylindri-
now has to use low-sulfur marine gas oil (LSMGO), cal vertical, operating at 7 barg (100 psig), and side fired
which is ∼50% more expensive than heavy-fuel oil, and with the burner firing across the diameter of the circular
convert the engine room plant and fuel tanks to suit. furnace. Thus, there is a need for the flame shape of the
heavy oil firing burner to be short and wide. The furnace
construction is water-tube wall with a refractory floor.
17.2.3 LNG Boil-Off Gas
Other marine auxiliary boiler types are conventional
LNG, or liquefied natural gas, first began being trans- shell/fire-tube boilers, 2 or 3 pass, in either horizontal
ported as a liquid cargo at −163°C (−261°F) by ships in configuration, or vertical configuration, thus minimiz-
the 1960s. Since the cargo tank insulation is not 100% ing the boiler footprint.
perfect and the tank pressure is near ambient, a small The majority of burners on small auxiliary boilers
amount of the LNG evaporates, which has resulted in utilize simple pressure jet atomizers in the range of 1–4
LNG boil-off gas (BOG) as a fuel. This fuel stream, being MW (3.3 × 106 –14 × 106 Btu/h), with the HFO preheated
almost 100% methane, was first used on the main pro- to ∼120°C (250°F) to achieve an oil viscosity of 15–20 cSt,
pulsion boilers of LNG carriers in the early 1970s and and pumped to a pressure of 20–30 barg (290–440 psig).
gave rise to the first dual-fuel register burners on ships. The burner construction and operation is simple, with
often just on-off, or high–low operation. Larger pressure
jet burners may be fully modulating with a turndown
17.2.4 Produced Fuel Gas and Crude Oil
of just 3 or 4–1. Ignition of the heavy-fuel oil is by direct
Remote offshore installations, particularly floating spark of the main heavy-fuel oil spray from the pressure
production storage and offloading (FPSO) and floating jet nozzle, with a simple high tension (HT) spark.
Marine and Offshore Applications 433
tertiary air to the combustion zone immediately in

front of the rotary-cup. This gives a well-controlled
flame shape and turndowns of up to 10:1 with capabili-
ties of adjusting the air/fuel ratio throughout the range
of the burner.
The rotary-cup burner has a separate diesel oil igniter
that is used to ignite the main heavy-fuel oil flame.
17.3.2 Watertube Auxiliary Boilers

Crude oil and product tankers require larger boilers of
typical range 20–50 m-ton/h (22–55 U.S. ton/h) operat-
ing at 16–20 barg (240–290 psig), for the main purpose of
providing steam for the turbine drives of cargo pumps,
used to offload the cargo. It is common for two boilers to
be fitted to each tanker to provide redundancy of steam
supply, although some tankers are fitted with just one
boiler, with capacities up to 100 m-ton/h (110 U.S. ton/h)
(a) on the largest of very large crude carriers (VLCCs).
Modern marine boilers of this type are either a ver-
tical cylindrical watertube, with the burner firing
downward along the axis of a cylindrical furnace, or
a D-type watertube, with the burner firing downward
into a rectangular box shaped furnace. A single steam-
atomized register burner, such as the Hamworthy
Combustion type DF (see Figure 17.2), is applied for
boilers up to 50 m-ton/h (55 U.S. ton/h), and two burn-
ers with isolation air-door registers for larger boilers
50–100 m-ton/h (55–110 U.S. ton/h).
Marine boilers are very compact and highly rated
compared with equivalent land boilers of the same duty.
Furnace heat release rates (defined as the thermal heat
(b) input divided by the furnace volume) can be as high as
FIGURE 17.1
1.4 MW/m3, (168 Btu/[h ft3]), which is about two times
Hamworthy Combustion ElectroTec® rotary-cup burner. (a) Burner that of land boilers. The reasons for this are the following:
machine sectional view and (b) complete assembly with FD fan.
• Space in the engine room is limited, so the
Some small auxiliary marine boilers and larger boilers smallest physical size is preferred.
up to 10 m-ton/h (11 U.S. ton/h), which is the equivalent • There is no NOx limitation or legislation for
of approximately 8 MW (27 × 106 Btu/h) heat input, have marine boilers on ships (unlike the main pro-
applied rotary-cup burners, such as the Hamworthy pulsion engines, which have strict NOx limits).
ElectroTec® (see Figure 17.1).
Rotary-cup burners are capable of burning heavy and Hence, the burner is required to produce a relatively
dirty oils without the use of an atomizing fluid or high- small and compact flame, and this is achieved by good
pressure fuel supply. Also, the required heavy-fuel oil tem- fuel oil atomization, effective air mixing, and high mix-
perature is lower at ∼90°C (190°F) to achieve 55–70 cSt. ing energy provided by burner draft losses as high as
The rotary cup is driven by a constant speed motor 30–35 mbar (12–15 in. W.G.).
at typically 5500 rpm. Oil is delivered to the rear of Heavy-fuel oil and atomizing steam are delivered to a
the cup at low modulated pressure (from a separate fuel oil sprayer, which is usually of concentric tube con-
pump) and runs along the inside of the cup as a thin struction, with the steam helping to keep the fuel oil up
film, picking up speed. When the oil spills off the lip of to temperature all the way to the Y-type atomizer tip
the cup, it is in droplet form with a high radial veloc- (see Figure 17.3). The fuel oil pressure is typically 16–18
ity and so is mechanically atomized. The combustion barg (240–260 psig) at the burner maximum combustion
air fan is directly driven by another motor to blow air rating (MCR), reducing to ∼2 barg (30 psig) at the burner
through dampers to supply primary, secondary, and turndown, which is of the order of 8–10 to 1.
(a) (b)
FIGURE 17.2
Hamworthy Combustion DF register burner. (a) Windbox view with fuel valve train and (b) burner throat view.
gases are extracted from the boiler uptakes, cooled and

scrubbed by an inert gas system, which blows the gases
into the cargo tanks, as the crude oil is unloaded. The
oxygen content in the inert gas (and hence boiler exhaust
gas) must be below 5%, and so the excess air level of
the burner must be accurately controlled through the
burner load range.
17.4 LNG Carriers

As described in Section 17.2.3, LNG carriers produce
boil-off gas from the LNG cargo which can be used as
an alternative to heavy-fuel oil, thus providing a high-
FIGURE 17.3 fuel cost saving for the ship’s operation. The first dual-
Heavy-fuel oil sprayer and twin-fluid Y-jet atomizer.
fuel LNG carrier was built in the early 1970s with the
The atomizing steam is saturated and regulated to main propulsion boilers providing high-pressure super-
a constant pressure of between 5 and 10 barg (70 and heated steam to a steam turbine directly coupled to the
150 psig). Cold start is by firing marine diesel oil, with ship’s propeller via a gearbox. Over 200 steam LNG car-
compressed air as the atomizing medium at ∼5 barg riers with dual-fuel boilers were built up to circa 2005,
(70 psig). The burner includes a pressure jet diesel oil when a fundamental change to dual-fuel medium speed
igniter, which uses a simple HT spark. and standard heavy-fuel oil-fired engines, which pro-
Combustion air is provided by an electric motor driven vided a significant efficiency improvement, was real-
forced-draft fan, with airflow regulation either a twin- ized. To date, over 120 LNG carriers have since been
bladed damper on the burner windbox inlet, or a multi- ordered with engine propulsion.
vane inlet damper on the fan inlet. The burner is fired
with an excess air at MCR of typically 15% (providing
17.4.1 LNG Carrier Main Propulsion Boilers
∼3% oxygen in the boiler exhaust) so as to ensure com-
plete and clean combustion of the residual heavy-fuel oil. Two such boilers are fitted for propulsion purposes on
On tankers, the boiler exhaust gases are commonly LNG carriers, with each having 2 or 3 dual-fuel regis-
used as an inert gas for the cargo tanks. The exhaust ter burners in downward firing orientation, such as the
(a) (b)
FIGURE 17.4
Hamworthy Combustion HXG dual-fuel register burner. (a) Burner frontplate view and (b) burner throat view.
Hamworthy Combustion type HXG (Figure 17.4). Each

burner incorporates an air door for open/closed opera-
tion for when the burner is in/out of service, actuated
by twin air cylinders. The heavy-fuel oil sprayer and
atomizer use the same design and principles as that
described for the burners for auxiliary watertube boil-
ers in Section 17.3.2, although the atomizing steam is
often superheated to ∼300°C (570°F).
The needs of the LNG carrier to maneuver at very
slow speeds demand a high turndown of the propul-
sion boilers. Thus, individual burner turndowns when
fuel oil firing of 15–1 are common.
The LNG boil-off gas is compressed to a nominal 1 barg
(15 psig) and heated to ambient temperature. Thus, the
fuel gas pressure at the burners at MCR is ∼0.6 barg (9
psig), with a maximum burner turndown of 7–1. The fuel
gas is injected into the air steam via multiple gas nozzles,
which include staged injection so as to provide stable com-
bustion across all operating conditions and boiler loads.
The burners can also fire heavy-fuel oil and boil-off
gas in combination, should the boil-off gas availability FIGURE 17.5
Chentronics® high-energy igniter.
not be sufficient to meet the boiler requirements.
As with auxiliary boilers, the boilers are very com-
17.4.2 LNG Carrier Gas Combustion Units
pact compared to equivalent land boilers, with furnace
heat release rates of 1–1.2 MW/m3 (120–140 Btu/[h ft3]). With the change in LNG carrier propulsion type from
Flames are therefore very compact, and vigorous when steam turbine to engines, a new type of combustion
viewed, with burner draft losses between 25 and 30 mbar application resulted. As the boil-off gas from the cargo
(10 and 12 in. W.G.) at MCR. tanks is continuous, there always needs to be a method
Ignition is usually by direct sparking of the main of safely disposing of the gas, since simply venting the
heavy-fuel oil-atomized spray by a high-energy spark methane-rich stream is not allowed as part of normal
rod, typically of type Chentronics® (see Figure 17.5), operations by the IGC Code (The International Code
which delivers 20 sparks per second. The igniter rod is for the Construction and Equipment of Ships Carrying
advanced and retracted in and out of the fuel oil spray Liquefied Gases in Bulk), which governs the design,
zone by a pneumatic air cylinder. construction, and operation of LNG carriers.
(a) (b)
FIGURE 17.6
Hamworthy Combustion AMOxsafe® GCU. (a) GCU system arrangement on aft deck and (b) GCU body.
When there is more boil-off gas than required by the is shown in Figure 17.7, and the temperature profile
propulsion need of a steam LNG carrier, the total boil- taken from computational fluid dynamic (CFD) simu-
off gas is fired in the boilers, and the excess steam pro- lations (see Chapter 9) in Figure 17.8. The boil-off gas
duced “dumped” to a sea-water cooled condenser. LNG supply to the GCU uses the same compressors that sup-
carriers with engines however, cannot burn this excess ply the LNG carrier’s main engines, and hence the pres-
boil-off gas without engine load; so a requirement for sure at the burner is ∼2 barg (29 psig). Ignition of the
an alternative safe method of safely burning the gas AMOxsafe® GCU is by a small diesel oil igniter using a
is needed. Hence, the need for a gas combustion unit, pressure jet, which can also be used as a permanent pilot
or GCU, evolved, such as the Hamworthy Combustion when the nitrogen content of the boil-off gas stream is
AMOxsafe® GCU (see Figure 17.6). high after cargo tank loading, so as to ensure complete
The prime function of the GCU is to safely inciner- and safe combustion of the methane.
ate the boil-off gas, while maintaining an exhaust tem-
perature below 535°C (995°F), a figure selected taking
into account the auto-ignition temperature of methane
with a 50°C (120°F) margin. The GCU is sized for the
maximum boil-off gas rate of the cargo tanks, which is
17.5 Offshore Applications
based on the guarantee value of the tank design. This
is typically 0.15% of the tank volume per day, and an 17.5.1 FPSOs
LNG carrier of capacity 175,000 m3 (5.8 × 106 ft3) would
FPSO vessels can be either conversions of existing
have a maximum boil-off gas rate of ∼4.5 m-ton/h
crude oil tankers, or purpose built as new. Most have
(5 U.S. ton/h). This boil-off gas stream is near 100%
the benefit of produced fuel gas being available for
methane, although just after ship’s loading, it will con-
heating and power requirements, as described in
tain a high level of nitrogen that quickly reduces. Thus,
Section 17.2.4.
4.5 m-ton/h (5 U.S. ton/h) equates to a thermal rating of
the GCU of 70 MW (240 × 106 Btu/h).
17.5.1.1 Tanker-Converted FPSOs
To achieve an exhaust temperature of less than 535°C
(995°F), and an industry standard bulk average of 450°C The majority of FPSOs have been tanker conversions,
(840°F), at MCR, the products of combustion are mixed which have existing steam boilers in the engine room,
and cooled by large volumes of dilution air. with these being either main high-pressure propulsion
In the case of AMOxsafe® GCU, the Hamworthy boilers (if the tanker was propelled by a steam turbine)
Combustion DF burner is used in an up-fired configura- or lower-pressure auxiliary boilers.
tion, with combustion air supplied by centrifugal fans. Since the original main boilers were only designed
The products of combustion are then mixed with dilution to fire fuel oil, their performance on fuel gas must
air from axial fans, which is also used to cool the double be calculated to determine the impact effect on the
lining of the GCU body, in the same way as a twin-shell original boiler design and materials used. As the fuel
air heater. The flow schematic of the AMOxsafe® GCU gas flames provide less radiant heat transfer to the
Exhaust
Mixing
air jets
Cold air Cold air
Dilution Dilution
air fan air fan
Steel, twin shell

construction
Register burner with

dual swirler
Flame
technology
shield
Vent
Combustion BOG
air fan × 2 Double Control
block and valve
vent valves
FIGURE 17.7
AMOxsafe® GCU flow schematic.
furnace watertube walls compared to fuel oil firing, For main boiler conversions, the original register
the temperature at the furnace exit, and in particular burners that fired fuel oil are replaced with register
at the superheater, is increased. Thus, the tempera- burners with dual-fuel capability, being very similar to
ture of the superheater material must be considered, those for LNG carrier main boiler (described in Section
and to keep the original boiler rating, it is common 17.4.1), such as Hamworthy Combustion’s type DF, as
for the superheater to be replaced with material of a shown in Figure 17.9.
high-temperature rating. Since the fuel gas is considered to be a hazardous gas,
Since the exhaust gases at the superheater are hot- when it is taken into the engine room, pipelines must
ter when gas firing, the attemporator (which controls be double skinned with either an outer pipe pressurized
the steam delivery temperature usually by a tube with an inert gas (nitrogen) or ventilated under negative
coil in the boiler’s lower water drum) and the attem- pressure by extraction air fans at a minimum rate of nor-
porator control valve must also be verified as their mally 30 air changes per hour. All potential leak sources
operating duty will have changed. The reduction in must be fully enclosed in ventilated valve boxes (like the
radiant heat transfer in the boiler furnace results in one shown in Figure 17.10) or “boiler hood rooms,” with
an increase in the boiler exhaust temperature and all electrical items rated for hazardous area applications.
an overall reduction in boiler efficiency of typically Typical fuel gas supply pressures from the crude oil
3%–4%. process are 6–8 barg (90–120 psig), and this pressure
Temperature
(plane 2)
2.216e + 003
1.742e + 003
1.269e + 003
7.963e + 002
3.232e + 002
(K)
0 1.744 3.488 5.232 (m)
FIGURE 17.8
AMOxsafe® GCU temperature profile.
FIGURE 17.9
Hamworthy Combustion dual-fuel DF register burner.
is regulated and controlled to give about 2 barg (30 of interruption of the fuel gas supply, so as to keep the
psig) at the burner gas nozzles at the boiler maximum boilers online at all times. Burner ignition is commonly
rating. by direct high-energy spark of the main fuel, in the
Fuel gas is the main fuel consumed as this is the low- same way as the register burners for the main boilers of
est cost and cleanest available. There is usually a “back- LNG carriers, described in Section 17.4.1.
up” second fuel, primarily being marine diesel oil, The conversion to gas firing of the auxiliary boilers
which must be automatically switched to in the event on motor tankers as part of their conversion to FPSOs
FIGURE 17.10
Fuel gas valve enclosure for FPSO engine room boiler.
is simpler than that of main boilers in that there is no be supplied as complete modules, as per that shown in
steam superheater or feedwater economizer, and there Figure 17.11.
is commonly just one burner on each boiler. A boiler Superheated, high-pressure steam boilers and asso-
performance study to assess the suitability of the boiler ciated turbines provide higher efficiency (reduced fuel
to fire fuel gas and calculate the drop in efficiency, is, consumption) and hence reduced emission solutions for
however, still necessary. power generation than lower pressure, saturated steam
boilers. Where there is insufficient produced fuel gas
available, as is often the case later in the FPSO operation
17.5.1.2 FPSO New Boilers term as the oilfield reserves deplete, crude oil is fired as
the prime fuel. Hence, plant efficiency is an important
Often the existing boilers of tankers are not sufficient economic consideration. A main deck boiler module for
in capacity for the new duty of that of an FPSO, and so 2 × 71 m-ton/h (2 × 78 U.S. ton/h) steam boilers, pro-
new additional boilers are required. There are several viding steam at 63 barg (930 psig) and 515°C (959°F) to
possible reasons: the power generation and/or machin- fire fuel gas, crude oil, and/or diesel oil, is shown in 3D
ery drives that are to be used by new steam turbines, CAD format in Figure 17.12.
high crude oil process heating requirements, or that
the original tanker only had a single auxiliary boiler
17.5.2 FSOs
and redundancy of steam supply is required. As the
engine rooms of tankers have no spare space for such FSO units are tankers that are permanently moored to
additional boilers, the new boilers are placed on the receive crude oil as it is processed from a neighboring
aft main deck, or on the cargo (process) main deck. production platform, with the crude oil arriving on the
In order to reduce the time in the tanker conversion FSO via a subsea pipe. Steam is required from auxiliary
shipyard, it is common practice for the new boilers to boilers on the FSO for crude oil heating, and for the
FIGURE 17.11
Hamworthy Combustion’s triple 120 m-ton/h (132 U.S. ton/h) steam boiler module for FPSO.
cargo pumps that offload the crude oil periodically to

shuttle tankers, which then take the crude oil to refin-
eries. Produced gas is often available at the production
platform, and this is commonly supplied to the FSO for
the boilers, via a second subsea pipe.
The firing of the fuel gas on the auxiliary boilers is the
same as for an FPSO; however, as the fuel gas is cooled
(due to the supply lying on the sea floor), any condensed
liquids (water and hydrocarbons) must be removed
before use in the boilers. Hence, fuel gas “knock-out”
pots similar to the one shown in Figure 17.13 are com-
mon additions, with the recovered liquids drained to a
suitable slop tank.
17.5.3 Floating LNG

Floating liquefied natural gas, or FLNG, is a relatively
new market, covering the application of permanently
moored LNG carriers for storage or re-gasification duties,
or the offshore production of LNG over a gas field.
17.5.3.1 LNG Re-Gasification Vessels

FIGURE 17.12
Hamworthy Combustion’s high-pressure steam boiler module for With the demand for natural gas continuing to rise due
power generation. to low prices and environmental benefits in terms of
FIGURE 17.13
Hamworthy Combustion fuel gas knock-out pot on an FSO.
emissions, an alternative to installing long pipelines to

carry the gas from source to users has evolved by way
of LNG tankers called re-gasification vessels or floating
storage and re-gasification units (FSRUs).
The FSRUs, whether new purpose built or converted
from existing LNG carriers, are permanently moored at
the shore side, usually close to the natural gas demand.
The ships are fitted with LNG pumps and re-gasification
systems, so they can deliver the LNG in its gaseous state
(natural gas) directly into the natural gas pipe network.
The LNG tanks are then regularly refilled from conven-
tional LNG tankers, which tie up alongside, and a ship-
to-ship transfer of the LNG is carried out.
The re-gasification process requires a considerable
amount of heat, usually by way of steam, to convert the
LNG from a liquid state at −163°C (−261°F), to a gaseous
state at ambient temperature. Hence, there is a demand for
low-pressure steam, which has to be available on the FSRU
from its own boilers. The fuel for the boilers is LNG boil-
FIGURE 17.14
off gas from the cargo tanks or the re-gassed LNG should Hamworthy Combustion DFL low NOx register burner.
the boil-off gas quantity not be sufficient to meet demand.
Since the FSRU is permanently moored at the shore, it
is common for the steam boilers to become subject to the techniques such as flue gas recirculation or NOx removal
same exhaust gas emission limits as the local legislation by selective catalytic reduction (SCR) may be necessary.
onshore. The natural gas firing provides very clean com-
17.5.3.2 Floating LNG Production Vessels
bustion with practically zero stack solids, carbon mon-
oxide, and sulfur dioxide. However, since NOx levels of As onshore and shallow water gas supplies reduce,
marine boilers are high (due to the very high furnace heat the move to deeper offshore gas production has also
release ratings), low NOx burners such as Hamworthy evolved. For deep water, a floating facility is required
Combustion’s type DFL (see Figure 17.14) can be fitted. and also since these gas fields are far offshore, pipe-
Depending on the NOx limit imposed, which varies lines to shore are impractical. Hence, the facility must
greatly around the globe, other NOx reduction combustion not only produce sales quality natural gas, but also
refrigerate and store it at −163°C (−261°F) as LNG. This is are often steam turbine driven. Hence, another offshore
the FLNG FPSO. steam boiler requirement, where the boilers would fire
The power demands of FLNG FPSOs are much larger, the produced fuel gas or LNG boil-off gas, but also
about three to four times, than that of an equivalent with the capability to fire liquid condensate which is
oil-producing FPSO, with one of the main demands produced as a by-product of the raw gas separation and
being the large LNG liquefaction compressors, which treatment process.
Appendix A: Units and Conversions
TABLE A.1 TABLE A.2 (continued)

Prefixes Basic Units, Conversions, and Molecular Properties
Multiplier Prefix Symbol 1 kJ/kg = 0.43 Btu/lb, 1 Btu/lb = 2.326 kJ/kg, 1 kg/GJ = 1 g/MJ =
2.326 lbm/mmBtu
1018 Exa E
1 Btu/SCF = 37 kJ/m3, 1 m3/GJ = 37.3 ft3/mmBtu, 1 lbm/mmBtu =
1015 Peta P 0.430 kg/GJ = 0.430 g/MJ
1012 Tera T 1 Therm = 105 Btu = 1.055 × 105 kJ
109 Giga G 1 hp = 0.7064 Btu/s = 0.7457 kW = 745.7 W = 550 lbf · ft/s =
106 Mega M 42.41 Btu/min
1 boiler hp = 33,475 Btu/h, 1 Btu/h = 1.0551 kJ/h
103 Kilo k
1 barrel (42 gallons) of crude oil = 5,800,000 Btu = 6120 MJ
102 Hecta h 1 gallon of gasoline = 124,000 Btu = 131 MJ
10 Deca da 1 gallon of heating oil = 139,000 Btu = 146.7 MJ
10−1 Deci d 1 gallon of diesel fuel = 139,000 Btu = 146.7 MJ
10−2 Centi c 1 barrel of residual fuel oil = 6,287,000 Btu = 6633 MJ
10−3 Milli m 1 cubic foot of natural gas = 1,026 Btu = 1.082 MJ
1 gallon of propane = 91,000 Btu = 96 MJ
10−6 Micro m
1 short ton of coal = 20,681,000 Btu = 21,821 MJ
10−9 Nano n
10−12 Pico p
Force
10−15 Femto f 1 lbf = 4.4482 N = 32.174 lbm · ft/s2 or gc = 32.174 lbm ft/s2 lbf
10−18 Atto a
Source: Annamalai, K. and Puri, I.K., Ideal Gas Law
Combustion Science and Engineering, PV = RT ; PV = mRT ; PV = nRT , Pv = RT ,
Table A.1A, CRC Press, Boca
Raton, FL, p. 981, 2007. R = 8.314 kPa m 3 /k mol k = 0.08314 bar m 3/k mol K
= 1.986 Btu/lb mol °R = 1545 ft lbf /lb mol °R
TABLE A.2 = 0.7299 atm ft 3 /lb mol °R
Basic Units, Conversions, and Molecular Properties
Length/Velocity
Area 1 in. = 0.0254 m
1 acre = 4046.9 m2 1 ft = 12 in. = 0.3048 m
1 m2 = 10−6 km2 = 104 cm2 = 106 mm2 1 mile = 5280 ft = 1609.3 m
1 m2 = 10.764 ft2 = 1550 in.2 1 statute mile = 0.87 nmi = 1.609 km
1 ft2 = 144 in.2 = 0.0929 m2 1 nautical mile = 1.15 smi = 1.85 km
1 hectare = 10,000 m2 = 2.5 acres = 108,000 ft2 1 mi/h = 1.46667 ft/s = 0.447 m/s = 1.609 km/h
1 m/s = 3.2808 ft/s = 2.237 mi/h = 1.96 kt = 1.15 smi/h = 3.63 km/h
Density
Speed of light in vacuum, c = 2.998 × 108 m/s
1 g/cm3 = 1 kg/L = 1000 kg/m3 = 62.43 lbm/ft3 = 0.03613 lbm/in3
1 kg/m3 = 0.06243 lbm/ft3, 1 lbm/ft3 = 16.018 kg/m3 Speed of sound = γRT
Specific gravity = density/reference density
For liquids, reference density of water at 15.74°C (60°F) = 999 kg/m3, Mass
62.4 lb/ft3 1 teragram (Tg) = 1 million metric tonnes
For gases, reference density of air at 15.74°C (60°F) = 1.206 kg/m3 Mass of an electron = 0.5 MeV (1 Mev = 106 eV; for mass,
use E = me2) = 9.109 × 1031 kg
Energy Mass of proton = 940 MeV = 1.67 × 10−27 kg, Mass of neutron =
1 eV ≈ 1.602 × 10−19 J. 1.675 × 10−27 kg
1 mBtu = 1 kBtu = 1000 Btu, 1 mmBtu = 1000 kBtu = 106 Btu 1 lbm = 0.45359 kg = 7000 grains
1 TBtu = 109 Btu or 1 GBtu 1 short ton = 2000 lb = 907.2 kg
1 quad = 1015 Btu or 1.05 × 1015 kJ or 2.93 × 1011 kWh = 172.4 million 1 long ton = 2240 lb or 1016.1 kg
barrels of crude oil 1 metric ton = 1000 kg
1 kWh = 0.0036 GJ = 3.6 MJ = 3412 Btu, 1 hp h = 0.00268 GJ = 1 ounce = 28.3495 g
2.68 MJ = Btu 1 kg = 2.2046 lb
1 Btu = 778.14 ft lbf = 1.0551 kJ, 1 kJ = 0.94782 Btu = 25,037 lbm ft/s2
1 cal = 4.1868 J, 1 (food) cal = 1000 cal or 1 kcal (continued)
443
444 Appendix A: Units and Conversions
TABLE A.2 (continued) TABLE A.2 (continued)

Basic Units, Conversions, and Molecular Properties Basic Units, Conversions, and Molecular Properties
Molecular Properties Volume of 1 kmol (SI) and 1 lb mol (English) of an ideal gas at STP
1 Å = 1.0 × 10−10 m conditions as defined below:
NAvog = 6.023 × 1026 molecules/kmol for a molecular substance
(e.g., oxygen)
Chemists’
= 6.023 × 1026 atoms/atom mole for an atomic substance (e.g., He)
Scientific or U.S. Standard Standard or NTP (Gas
Boltzmann constant, kB = 1.38 × 10−26 kJ/molecule K
SATP (1976) or ISA CSA Industry)
Planck’s constant, hP = 6.626 × 10−37 kJ s/molecule
Stefan–Boltzmann constant, σ = 5.66961 × 10−11 kW/m2 K4 25°C (77°F), 15°C (60°F), 0°C (32°F), 20°C (65°F),
Charge of an electron = 1.602 × 10−19 coulombs, orbit radius (nm) = 101.3 kPa 101.33 kPa 101.33 kPa 101.33 kPa
0.0529n2, n: orbit number (14.7 psi, (1 atm, (1 atm, 14.7 psi, (1 atm)
Energy level of an orbit (eV) = 13.56/n2 29.92 in. of 14.696 psi, 29.92 in. of Hg)
Hg) 29.92 in. of Hg)
Numbers
24.5 m3/kmol 23.7 m3/kmol 22.4 m3/kmol 23.89 m3/kmol
ln x = 2.303 log10 x
(392 ft3/lb (375.6 ft3/lb (359.2 ft3/lb (382.7 ft3/lb
log10 x = 0.4343 ln x
mol) mol) mol) mol)
e = 2.718
π = 3.142 SATP, standard ambient temperature and pressure; ISA, international
1 deg = 0.0175 rad standard atmosphere; NTP, normal temperature and pressure.
Pressure
1 bar = 105 Pa, 1 mm Hg = 133.3 Pa
1 in Hg = 3.387 kPa = 0.491 psi Air Composition
1 in water (4°C) = 0.03613 psi
1 atm = 14.696°lbf/in.2 = 1.0133 bar = 10.3323 mm of H2O (4°C) = Species Mole % Mass % Molecular Weight
760 mm of Hg(0°C) Ar 0.934 1.288287 39.948
1 psi = 1 lbf/in.2 = 144 lbf/ft2 = 6.894 kPa = 6894 Pa = 27.653
CO2 0.0314 0.047715 44.01
in water (4°C)
N2 78.084 75.51721 28.01
Specific Heat O2 20.9476 23.14489 32
1 Btu/lb °F = 4.1868 kJ/kg °C
Ne 0.001818 0.001267 20.18
1 kJ/kg °C = 0.23885 Btu/lb °F
He 0.000524 7.24E−05 4.0026
Temperature Krypton 0.000114 0.00033 83.8
T (°C) = (T(°F) − 32) * (5/9).
Xe 8.70E−06 3.94E−05 131.3
T (°F) = T (°C) * 1.8 + 32.
To convert electron volts into the corresponding temperature H2 0.00005 3.48E−06 2.016
in Kelvin, multiply by 11,604. CH4 0.0002 0.000111 16.043
N2O 0.00005 7.6E−05 44.013
Volume
SO2, NO2, CO, I2 0.000235 — —
1 m3 = 1000 L
1 fluid ounce = 29.5735 cm3 = 0.0295735 L Source: Annamalai, K. and Puri, I.K., Combustion Science and
1 m3/kg = 1000 L/kg = 16.02 ft3/lb, 1 m3/GJ = 37.26 ft3/mmBtu Engineering, Table A.1A, CRC Press, Boca Raton, FL,
1 ft3/lbm = 0.062428 m3/kg p. 981, 2007.
1 U.S. gallon = 128 fluid ounce = 3.786 L Note: Molecular weight (mass) of air = 28.96 kg/kmol.
1 barrel = 42 U.S. gallons = 35 imperial gallons = 158.98 L = 5.615 ft3 =
231 in.3 = 0.1337 ft3
Appendix A: Units and Conversions 445
TABLE A.3 TABLE A.3 (continued)

Atomic Weights for Common Elements Atomic Weights for Common Elements
Name Symbol Atomic Number Atomic Weight Name Symbol Atomic Number Atomic Weight
Aluminum Al 13 26.98 Silver Ag 47 107.87
Antimony Sb 51 121.76 Sodium Na 11 22.99
Argon Ar 18 39.95 Strontium Sr 38 87.62
Arsenic As 33 74.92 Sulfur S 16 32.07
Barium Ba 56 137.32 Tantalum Ta 73 180.95
Beryllium Be 4 9.01 Thallium Tl 81 204.38
Bismuth Bi 83 208.98 Tin Sn 50 118.71
Boron B 5 10.811 Titanium Ti 22 47.87
Bromine Br 35 79.90 Tungsten W 74 183.84
Cadmium Cd 48 112.41 Uranium U 92 238.03
Calcium Ca 20 40.08 Vanadium V 23 50.94
Carbon C 6 12.01 Xenon Xe 54 131.29
Cesium Cs 55 132.91 Zinc Zn 30 65.39
Chlorine Cl 17 35.45 Zirconium Zr 40 92.22
Chromium Cr 24 52.00 Source: Annamalai, K. and Puri, I.K., Combustion Science and
Cobalt Co 27 58.93 Engineering, Table A.1B, CRC Press, Boca Raton, FL,
Copper Cu 29 63.55 p. 985, 2007.
Fluorine F 9 19.00
Germanium Ge 32 72.61
Gold Au 79 196.97
Helium He 2 4.00
Hydrogen H 1 1.01
Indium In 49 114.82
Iodine I 53 126.90
Iridium If 77 192.22
Iron Fe 26 55.85
Krypton Kr 36 83.80
Lead Pb 82 207.20
Lithium Li 3 6.94
Magnesium Mg 12 24.31
Manganese Mn 25 54.94
Mercury Hg 80 200.59
Molybdenum Mo 42 95.94
Neon Ne 10 20.18
Nickel Ni 28 58.69
Nitrogen N 7 14.01
Oxygen O 8 16.00
Palladium Pd 46 106.42
Phosphorus P 15 30.97
Platinum Pt 78 195.08
Plutonium Pu 94 244.00
Potassium K 19 39.10
Radium Ra 88 226.00
Radon Rn 86 222.00
Rhodium Rh 45 102.91
Selenium Se 34 78.96
Silicon Si 14 28.09
446
FIGURE A.1
Appendix A: Units and Conversions
Periodic table of the elements. (From Dragoset, R.A. et al., NIST, Physical Measurement Laboratory, U.S. National Institute of Standards and Technology, 2011.)
Appendix A: Units and Conversions 447
References
1. K. Annamalai and J.K. Puri, Combustion Science and Engineering, CRC Press, Boca Raton, FL, 2007, p. 981, Table A.1A.
2. R.A. Dragoset, A. Musgrove, C.W. Clark, and W.C. Martin, NIST, Physical Measurement laboratory, U.S. National Institute
of Standards Technology, 2011.
Appendix B: Physical Properties of Materials
TABLE B.1 TABLE B.1 (continued)

Areas and Circumferences of Circles and Drill Sizes Areas and Circumferences of Circles and Drill Sizes
Diameter Circumference Area Area Diameter Circumference Area Area
Drill Size (in.2) (in.2) (in.2) (ft2) Drill Size (in.2) (in.2) (in.2) (ft2)
80 0.0135 0.042 41 0.000 143 0.000 000 9 41 0.0960 0.301 59 0.007 24 0.000 050 3
79 0.0145 0.045 55 0.000 165 0.000 001 1 40 0.0980 0.307 88 0.007 54 0.000 052 4
1/64 in.2 0.0156 0.049 09 0.000 191 0.000 001 3 39 0.0995 0.312 59 0.007 78 0.000 054 0
78 0.0160 0.050 27 0.000 201 0.000 001 4 38 0.1015 0.318 87 0.008 09 0.000 056 2
77 0.0180 0.056 55 0.000 254 0.000 001 8 37 0.1040 0.326 73 0.008 49 0.000 059 0
76 0.0200 0.062 83 0.000 314 0.000 002 2 36 0.1065 0.334 58 0.008 91 0.000 061 9
75 0.0210 0.065 97 0.000 346 0.000 002 4 7/64 in.2 0.1094 0.343 61 0.009 40 0.000 065 2
74 0.0225 0.070 69 0.000 398 0.000 002 8 35 0.1100 0.345 58 0.009 50 0.000 066 0
73 0.0240 0.075 40 0.000 452 0.000 003 1 34 0.1110 0.348 72 0.009 68 0.000 067 2
72 0.0250 0.078 54 0.000 491 0.000 003 4 33 0.1130 0.355 00 0.010 03 0.000 069 6
71 0.0260 0.081 68 0.000 531 0.000 003 7 32 0.1160 0.364 43 0.010 57 0.000 073 4
70 0.0280 0.087 96 0.000 616 0.000 004 3 31 0.1200 0.376 99 0.011 31 0.000 078 5
69 0.0292 0.091 73 0.000 670 0.000 004 7 1/8 in.2 0.1250 0.392 70 0.012 27 0.000 085 2
68 0.0310 0.097 39 0.000 755 0.000 005 2 30 0.1285 0.403 70 0.012 96 0.000 090 1
1/32 in.2 0.0313 0.098 18 0.000 765 0.000 005 3 29 0.1360 0.427 26 0.014 53 0.000 100 9
67 0.0320 0.100 53 0.000 804 0.000 005 6 28 0.1405 0.441 39 0.015 49 0.000 107 7
66 0.0330 0.103 67 0.000 855 0.000 005 9 9/64 in.2 0.1406 0.441 79 0.015 53 0.000 107 9
65 0.0350 0.109 96 0.000 962 0.000 006 7 27 0.1440 0.442 39 0.016 29 0.000 113 1
64 0.0360 0.113 10 0.001 018 0.000 007 1 26 0.1470 0.461 82 0.016 97 0.000 117 9
63 0.0370 0.116 24 0.001 075 0.000 007 5 25 0.1495 0.469 67 0.017 55 0.000 121 9
62 0.0380 0.119 38 0.001 134 0.000 007 9 24 0.1520 0.477 52 0.018 15 0.000 126 0
61 0.0390 0.122 52 0.001 195 0.000 008 3 23 0.1540 0.483 81 0.018 63 0.000 129 4
60 0.0400 0.125 66 0.001 257 0.000 008 7 5/32 in.2 0.1562 0.490 87 0.019 17 0.000 133 1
59 0.0410 0.128 81 0.001 320 0.000 009 2 22 0.1570 0.493 23 0.019 36 0.000 134 4
58 0.0420 0.131 95 0.001 385 0.000 009 6 21 0.1590 0.499 51 0.019 86 0.000 137 9
57 0.0430 0.135 09 0.001 452 0.000 010 1 20 0.1610 0.505 80 0.020 36 0.000 141 4
56 0.0465 0.146 08 0.001 698 0.000 011 8 19 0.1660 0.521 51 0.021 64 0.000 150 3
3/64 in.2 0.0469 0.147 26 0.001 73 0.000 012 0 18 0.1695 0.532 50 0.022 56 0.000 156 7
55 0.0520 0.163 36 0.002 12 0.000 014 7 11/64 in.2 0.1719 0.539 96 0.023 20 0.000 161 1
54 0.0550 0.172 79 0.002 38 0.000 016 5 17 0.1730 0.543 50 0.023 51 0.000 163 2
53 0.0595 0.186 93 0.002 78 0.000 019 3 16 0.1770 0.556 06 0.024 61 0.000 170 9
1/16 in.2 0.0625 0.196 35 0.003 07 0.000 021 3 15 0.1800 0.565 49 0.025 45 0.000 176 7
52 0.0635 0.199 49 0.003 17 0.000 022 0 14 0.1820 0.571 77 0.026 02 0.000 180 7
51 0.0670 0.210 49 0.003 53 0.000 024 5 13 0.1850 0.581 20 0.026 88 0.000 186 7
50 0.0700 0.219 91 0.003 85 0.000 026 7 3/16 in.2 0.1875 0.589 05 0.027 61 0.000 191 7
49 0.0730 0.229 34 0.004 19 0.000 029 1 12 0.1890 0.593 76 0.028 06 0.000 194 8
48 0.0760 0.238 76 0.004 54 0.000 031 5 11 0.1910 0.600 05 0.028 65 0.000 199 0
5/64 in.2 0.0781 0.245 44 0.004 79 0.000 033 3 10 0.1930 0.606 33 0.029 40 0.000 203 2
47 0.0785 0.246 62 0.004 84 0.000 033 6 9 0.1960 0.615 75 0.030 17 0.000 209 5
46 0.0810 0.254 47 0.005 15 0.000 035 8 8 0.1990 0.625 18 0.031 10 0.000 216 0
45 0.0820 0.257 61 0.005 28 0.000 036 7 7 0.2010 0.631 46 0.031 73 0.000 220 4
44 0.0860 0.270 18 0.005 81 0.000 040 3 13/64 in.2 0.2031 0.638 14 0.032 41 0.000 224 8
43 0.0890 0.279 60 0.006 22 0.000 043 2 6 0.2040 0.640 89 0.032 69 0.000 227 0
42 0.0935 0.293 74 0.006 87 0.000 047 7 5 0.2055 0.645 60 0.033 17 0.000 230 3
3/32 in.2 0.0937 0.294 52 0.006 90 0.000 047 9 4 0.2090 0.656 59 0.034 31 0.000 238 2
(continued)
449
450 Appendix B: Physical Properties of Materials
TABLE B.1 (continued) TABLE B.1 (continued)

3 0.2130 0.669 16 0.035 63 0.000 247 5 9/16 in.2 0.5625 1.7672 0.2485 0.001 726
7/32 in.2 0.2187 0.687 22 0.037 58 0.000 261 0 37/64 in.2 0.5781 1.8162 0.2625 0.001 823
2 0.2210 0.694 29 0.038 36 0.000 266 4 19/32 in.2 0.5938 1.8653 0.2769 0.001 923
1 0.2280 0.716 28 0.040 83 0.000 283 5 39/64 in.2 0.6094 1.9144 0.2917 0.002 025
A 0.2340 0.735 13 0.043 01 0.000 298 7 5/8 in.2 0.6250 1.9635 0.3068 0.002 131
15/64 in.2 0.2344 0.736 31 0.043 14 0.000 299 6 41/64 in.2 0.6406 2.0126 0.3223 0.002 238
B 0.2380 0.747 70 0.044 49 0.000 308 9 21/32 in.2 0.6562 2.0617 0.3382 0.002 350
C 0.2420 0.760 27 0.046 00 0.000 319 4 43/64 in.2 0.6719 2.1108 0.3545 0.002 462
D 0.2460 0.772 83 0.047 53 0.000 330 1 11/16 in.2 0.6875 2.1598 0.3712 0.002 578
E = 1/4 in.2 0.2500 0.785 40 0.049 09 0.000 340 9 23/32 in.2 0.7188 2.2580 0.4057 0.002 818
F 0.2570 0.807 39 0.051 87 0.000 360 2 3/4 in.2 0.7500 2.3562 0.4418 0.003 068
G 0.2610 0.819 96 0.053 50 0.000 371 5 25/32 in.2 0.7812 2.4544 0.4794 0.003 329
17/64 in.2 0.2656 0.834 41 0.055 42 0.000 384 9 13/16 in.2 0.8125 2.5525 0.5185 0.003 601
H 0.2660 0.835 67 0.055 57 0.000 385 9 27/32 in.2 0.8438 2.6507 0.5591 0.003 883
I 0.2720 0.854 52 0.058 11 0.000 403 5 7/8 in.2 0.8750 2.7489 0.6013 0.004 176
J 0.2770 0.870 22 0.060 26 0.000 418 5 29/32 in.2 0.9062 2.8471 0.6450 0.004 479
K 0.2810 0.882 79 0.062 02 0.000 430 7 15/16 in.2 0.9375 2.9452 0.6903 0.004 794
9/32 in.2 0.2812 0.883 57 0.062 13 0.000 431 5 31/32 in.2 0.9688 3.0434 0.7371 0.005 119
L 0.2900 0.911 06 0.066 05 0.000 458 7 1 in.2 1.0000 3.1416 0.7854 0.005 454
M 0.2950 0.926 77 0.068 35 0.000 474 7 1 1/16 in.2 1.0625 3.3379 0.8866 0.006 157
19/64 in.2 0.2969 0.932 66 0.069 22 0.000 480 7 1 1/8 in.2 1.1250 3.5343 0.9940 0.006 903
N 0.3030 0.951 90 0.071 63 0.000 500 7 1 3/16 in.2 1.1875 3.7306 1.1075 0.007 691
5/16 in.2 0.3125 0.981 75 0.076 70 0.000 532 6 1 1/4 in.2 1.2500 3.9270 1.2272 0.008 522
O 0.3160 0.992 75 0.078 43 0.000 544 6 1 5/16 in.2 1.3125 4.1233 1.3530 0.009 396
P 0.3230 1.014 74 0.081 94 0.000 569 0 1 3/8 in.2 1.3750 4.3170 1.4849 0.010 31
21/64 in.2 0.3281 1.030 8 0.084 56 0.000 587 2 1 7/16 in.2 1.4375 4.5160 1.6230 0.011 27
Q 0.3320 1.043 0 0.086 57 0.000 601 2 1 1/2 in.2 1.5000 4.7124 1.7671 0.012 27
R 0.3390 1.065 0 0.090 26 0.000 626 8 1 9/16 in.2 1.5625 4.9087 1.9175 0.013 32
11/32 in.2 0.3437 1.079 8 0.092 81 0.000 644 5 1 5/8 in.2 1.6250 5.1051 2.0739 0.014 40
S 0.3480 1.093 3 0.095 11 0.000 660 5 1 11/16 in.2 1.6875 5.3014 2.2365 0.015 53
T 0.3580 1.124 7 0.100 6 0.000 699 0 1 3/4 in.2 1.7500 5.4978 2.4053 0.016 70
23/64 in.2 0.3594 1.129 0 0.101 4 0.000 704 4 1 13/16 in.2 1.8125 5.6941 2.5802 0.017 92
U 0.3680 1.156 1 0.106 4 0.000 738 6 1 7/8 in.2 1.8750 5.8905 2.7612 0.019 18
3/8 in.2 0.3750 1.178 1 0.110 5 0.000 767 0 1 15/16 in.2 1.9375 6.0868 2.9483 0.020 47
V 0.3770 1.184 4 0.111 6 0.000 775 2 2 in.2 2.0000 6.2832 3.1416 0.021 82
W 0.3860 1.212 7 0.117 0 0.000 812 7 2 1/16 in.2 2.0625 6.4795 3.3410 0.023 20
25/64 in.2 0.3906 1.227 2 0.119 8 0.000 832 2 2 1/8 in.2 2.1250 6.6759 3.5466 0.024 63
X 0.3970 1.247 2 0.123 8 0.000 859 6 2 3/16 in.2 2.1875 6.8722 3.7583 0.026 10
Y 0.4040 1.269 2 0.128 2 0.000 890 2 2 1/4 in.2 2.2500 7.0686 3.9761 0.027 61
13/32 in.2 0.4062 1.276 3 0.129 6 0.000 900 1 2 5/16 in.2 2.3125 7.2649 4.2000 0.029 17
Z 0.4130 1.297 5 0.134 0 0.000 930 3 2 3/8 in.2 2.3750 7.4613 4.4301 0.030 76
27/64 in.2 0.4219 1.325 4 0.139 8 0.000 970 8 2 7/16 in.2 2.4375 7.6576 4.6664 0.032 41
7/16 in.2 0.4375 1.3745 0.1503 0.001 044 2 1/2 in.2 2.5000 7.8540 4.9087 0.034 09
29/64 in.2 0.4531 1.4235 0.1613 0.001 120 2 9/16 in.2 2.5625 8.0503 5.1572 0.035 81
15/32 in.2 0.4687 1.4726 0.1726 0.001 198 2 5/8 in.2 2.6250 8.2467 5.4119 0.037 58
31/64 in.2 0.4844 1.5217 0.1843 0.001 280 2 11/16 in.2 2.6875 8.4430 5.6727 0.039 39
1/2 in.2 0.5000 1.5708 0.1964 0.001 364 2 3/4 in.2 2.7500 8.6394 5.9396 0.041 25
33/64 in.2 0.5156 1.6199 0.2088 0.001 450 2 13/16 in.2 2.8125 8.8357 6.2126 0.043 14
17/32 in.2 0.5313 1.6690 0.2217 0.001 539 2 7/8 in.2 2.8750 9.0323 6.4918 0.045 08
35/64 in.2 0.5469 1.7181 0.2349 0.001 631 2 15/16 in.2 2.9375 9.2284 6.7771 0.047 06
Appendix B: Physical Properties of Materials 451

3 in.2 3.0000 9.4248 7.0686 0.049 09 6 1/4 in.2 6.2500 19.649 30.680 0.2131
3 1/16 in.2 3.0625 9.6211 7.3662 0.051 15 6 3/8 in.2 6.3750 20.028 31.919 0.2217
3 1/8 in.2 3.1250 9.8175 7.6699 0.053 26 6 1/2 in.2 6.5000 20.420 33.183 0.2304
3 3/16 in.2 3.1875 10.014 7.9798 0.055 42 6 5/8 in.2 6.6250 20.813 34.472 0.2394
3 1/4 in.2 3.2500 10.210 8.2958 0.057 36 6 3/4 in.2 6.7500 21.206 35.785 0.2485
3 5/16 in.2 3.3125 10.407 8.6179 0.059 85 6 7/8 in.2 6.8750 21.598 37.122 0.2578
3 3/8 in.2 3.3750 10.603 8.9462 0.062 13 7 in.2 7.0000 21.991 38.485 0.2673
3 7/16 in.2 3.4375 10.799 9.2806 0.064 45 7 1/8 in.2 7.1250 22.384 39.871 0.2769
3 1/2 in.2 3.5000 10.996 9.6211 0.066 81 7 1/4 in.2 7.2500 22.777 41.283 0.2867
3 9/16 in.2 3.5625 11.192 9.9678 0.069 22 7 3/8 in.2 7.3750 23.169 42.718 0.2967
3 5/8 in.2 3.6250 11.388 10.321 0.071 67 7 1/2 in.2 7.5000 23.562 44.179 0.3068
3 11/16 in.2 3.6875 11.585 10.680 0.074 17 7 5/8 in.2 7.6250 23.955 45.664 0.3171
3 3/4 in.2 3.7500 11.781 11.045 0.076 70 7 3/4 in.2 7.7500 24.347 47.173 0.3276
3 13/16 in.2 3.8125 11.977 11.416 0.079 28 7 7/8 in.2 7.8750 24.740 48.707 0.3382
3 7/8 in.2 3.8750 12.174 11.793 0.081 90 8 in.2 8.0000 25.133 50.266 0.3491
3 15/16 in.2 3.9375 12.370 12.177 0.084 56 8 1/8 in.2 8.1250 25.525 51.849 0.3601
4 in.2 4.0000 12.566 12.566 0.087 26 8 1/4 in.2 8.2500 25.918 53.456 0.3712
4 1/16 in.2 4.0625 12.763 12.962 0.090 02 8 3/8 in.2 8.3750 26.301 55.088 0.3826
4 1/8 in.2 4.1250 12.959 13.364 0.092 81 8 1/2 in.2 8.5000 26.704 56.745 0.3941
4 3/16 in.2 4.1875 13.155 13.772 0.095 64 8 5/8 in.2 8.6250 27.096 58.426 0.4057
4 1/4 in.2 4.2500 13.352 14.186 0.098 52 8 3/4 in.2 8.7500 27.489 60.132 0.4176
4 5/16 in.2 4.3125 13.548 14.607 0.101 4 8 7/8 in.2 8.8750 27.882 61.862 0.4296
4 3/8 in.2 4.3750 13.745 15.033 0.104 3 9 in.2 9.0000 28.274 63.617 0.4418
4 7/16 in.2 4.4375 13.941 15.466 0.107 4 9 1/8 in.2 9.1250 28.667 65.397 0.4541
4 1/2 in.2 4.5000 14.137 15.904 0.110 4 9 1/4 in.2 9.2500 29.060 67.201 0.4667
4 9/16 in.2 4.5625 14.334 16.349 0.113 5 9 3/8 in.2 9.3750 29.452 69.029 0.4794
4 5/8 in.2 4.6250 14.530 16.800 0.1167 9 1/2 in.2 9.5000 29.845 70.882 0.4922
4 11/16 in.2 4.6875 14.726 17.257 0.1198 9 5/8 in.2 9.6250 30.238 72.760 0.5053
4 3/4 in.2 4.7500 14.923 17.721 0.1231 9 3/4 in.2 9.7500 30.631 74.662 0.5185
4 13/16 in.2 4.8125 15.119 18.190 0.1263 9 7/8 in.2 9.8750 31.023 76.589 0.5319
4 7/8 in.2 4.8750 15.315 18.665 0.1296 10 in.2 10.0000 31.416 78.540 0.5454
4 15/16 in.2 4.9375 15.512 19.147 0.1330 10 1/8 in.2 10.1250 31.809 80.516 0.5591
5 in.2 5.0000 15.708 19.635 0.1364 10 1/4 in.2 10.2500 32.201 82.516 0.5730
5 1/16 in.2 5.0625 15.904 20.129 0.1398 10 3/8 in.2 10.3750 32.594 84.541 0.5871
5 1/8 in.2 5.1250 16.101 20.629 0.1433 10 1/2 in.2 10.5000 32.987 86.590 0.6013
5 3/16 in.2 5.1875 16.297 21.135 0.1468 10 5/8 in.2 10.6250 33.379 88.664 0.6157
5 1/4 in.2 5.2500 16.493 21.648 0.1503 10 3/4 in.2 10.7500 33.772 90.763 0.6303
5 5/16 in.2 5.3125 16.690 22.166 0.1539 10 7/8 in.2 10.8750 34.165 92.886 0.6450
5 3/8 in.2 5.3750 16.886 22.691 0.1576 11 in.2 11.0000 34.558 95.033 0.6600
5 7/16 in.2 5.4375 17.082 23.221 0.1613 11 1/8 in.2 11.1250 34.950 97.205 0.6750
5 1/2 in.2 5.5000 17.279 23.758 0.1650 11 1/4 in.2 11.2500 35.343 99.402 0.6903
5 9/16 in.2 5.5625 17.475 24.301 0.1688 11 3/8 in.2 11.3750 35.736 101.6 0.7056
5 5/8 in.2 5.6250 17.671 24.851 0.1726 11 1/2 in.2 11.5000 36.128 103.9 0.7213
5 11/16 in.2 5.6875 17.868 25.406 0.1764 11 5/8 in.2 11.6250 36.521 106.1 0.7371
5 3/4 in.2 5.7500 18.064 25.967 0.1803 11 3/4 in.2 11.7500 36.914 108.4 0.7530
5 13/16 in.2 5.8125 18.261 26.535 0.1843 11 7/8 in.2 11.8750 37.306 110.8 0.7691
5 7/8 in.2 5.8750 18.457 27.109 0.1883 12 in.2 12.0000 37.699 113.1 0.7854
5 15/16 in.2 5.9375 18.653 27.688 0.1923 12 1/4 in.2 12.2500 38.485 117.9 0.819
6 in.2 6.0000 18.850 28.274 0.1963 12 1/2 in.2 12.5000 39.269 122.7 0.851
6 1/8 in.2 6.1250 19.242 29.465 0.2046 12 3/4 in.2 12.7500 40.055 127.7 0.886
(continued)

13 in.2 13.0000 40.841 132.7 0.921 24 1/4 in.2 24.2500 76.184 461.9 3.207
13 1/4 in.2 13.2500 41.626 137.9 0.957 24 1/2 in.2 24.5000 76.969 471.4 3.274
13 1/2 in.2 13.5000 42.412 143.1 0.995 24 3/4 in.2 24.7500 77.755 481.1 3.341
13 3/4 in.2 13.7500 43.197 148.5 1.031 25 in.2 25.0000 78.540 490.9 3.409
14 in.2 14.0000 43.982 153.9 1.069 25 1/4 in.2 25.2500 79.325 500.7 3.477
14 1/4 in.2 14.2500 44.768 159.5 1.109 25 1/2 in.2 25.5000 80.111 510.7 3.547
14 1/2 in.2 14.5000 45.553 165.1 1.149 25 3/4 in.2 25.7500 80.896 520.8 3.616
14 3/4 in.2 14.7500 46.339 170.9 1.185 26 in.2 26.0000 81.682 530.9 3.687
15 in.2 15.0000 47.124 176.7 1.228 26 1/4 in.2 26.2500 82.467 541.2 3.758
15 1/4 in.2 15.2500 47.909 182.7 1.269 26 1/2 in.2 26.5000 83.252 551.6 3.830
15 1/2 in.2 15.5000 48.695 188.7 1.309 26 3/4 in.2 26.7500 84.038 562.0 3.903
15 3/4 in.2 15.7500 49.480 194.8 1.352 27 in.2 27.0000 84.823 572.6 3.976
16 in.2 16.0000 50.266 201.1 1.398 27 1/4 in.2 27.2500 85.609 583.2 4.050
16 1/4 in.2 16.2500 51.051 207.4 1.440 27 1/2 in.2 27.5000 86.394 594.0 4.125
16 1/2 in.2 16.5000 51.836 213.8 1.485 27 3/4 in.2 27.7500 87.179 604.8 4.200
16 3/4 in.2 16.7500 52.622 220.4 1.531 28 in.2 28.0000 87.965 615.8 4.276
17 in.2 17.0000 53.407 227.0 1.578 28 1/4 in.2 28.2500 88.750 626.8 4.353
17 1/4 in.2 17.2500 54.193 233.7 1.619 28 1/2 in.2 28.5000 89.536 637.9 4.430
17 1/2 in.2 17.5000 54.978 240.5 1.673 28 3/4 in.2 28.7500 90.321 649.2 4.508
17 3/4 in.2 17.7500 55.763 247.5 1.719 29 in.2 29.0000 91.106 660.5 4.587
18 in.2 18.0000 56.548 254.5 1.769 29 1/4 in.2 29.2500 91.892 672.0 4.666
18 1/4 in.2 18.2500 57.334 261.6 1.816 29 1/2 in.2 29.5000 92.677 683.5 4.746
18 1/2 in.2 18.5000 58.120 268.8 1.869 29 3/4 in.2 29.7500 93.463 695.1 4.827
18 3/4 in.2 18.7500 58.905 276.1 1.920 30 in.2 30.0000 94.248 706.9 4.909
19 in.2 19.0000 59.690 283.5 1.969 31 in.2 31.0000 97.390 754.8 5.241
19 1/4 in.2 19.2500 60.476 291.0 2.022 32 in.2 32.0000 100.53 804.3 5.585
19 1/2 in.2 19.5000 61.261 298.7 2.075 33 in.2 33.0000 103.67 855.3 5.940
19 3/4 in.2 19.7500 62.047 306.4 2.125 34 in.2 34.0000 106.81 907.9 6.305
20 in.2 20.0000 62.832 314.2 2.182 35 in.2 35.0000 109.96 962.1 6.681
20 1/4 in.2 20.2500 63.617 322.1 2.237 36 in.2 36.0000 113.10 1017.9 7.069
20 1/2 in.2 20.5000 64.403 330.1 2.292 37 in.2 37.0000 116.24 1075.2 7.467
20 3/4 in.2 20.7500 65.188 338.2 2.348 38 in.2 38.0000 119.38 1134.1 7.876
21 in.2 21.0000 65.974 346.4 2.405 39 in.2 39.0000 122.52 1194.6 8.296
21 1/4 in.2 21.2500 66.759 354.7 2.463 40 in.2 40.0000 125.66 1256.6 8.727
21 1/2 in.2 21.5000 67.544 363.1 2.521 41 in.2 41.0000 128.81 1320.3 9.168
21 3/4 in.2 21.7500 68.330 371.5 2.580 42 in.2 42.0000 131.95 1385.4 9.621
22 in.2 22.0000 69.115 380.1 2.640 43 in.2 43.0000 135.09 1452.2 10.08
22 1/4 in.2 22.2500 69.901 388.8 2.700 44 in.2 44.0000 138.23 1520.5 10.56
22 1/2 in.2 22.5000 70.686 397.6 2.761 45 in.2 45.0000 141.37 1590.4 11.04
22 3/4 in.2 22.7500 71.471 406.5 2.823 46 in.2 46.0000 144.51 1661.9 11.54
23 in.2 23.0000 72.257 415.5 2.885 47 in.2 47.0000 147.66 1734.9 12.04
23 1/4 in.2 23.2500 73.042 424.6 2.948 48 in.2 48.0000 150.80 1809.6 12.57
23 1/2 in.2 23.5000 73.828 433.7 3.012 49 in.2 49.0000 153.94 1885.7 13.10
23 3/4 in.2 23.7500 74.613 443.0 3.076 50 in.2 50.0000 157.08 1963.5 13.64
24 in.2 24.0000 75.398 452.4 3.142 Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
TABLE B.2
Physical Properties of Pipe
Schedule
Nominal Wall Inside Weight of Moment Section Radius
Number
Pipe Size, Thickness ID Area Metal Sq. Ft. Outside Sq. Ft. Inside Weight Water per of Inertia Modulus Gyration
OD (in.) a b c (in.) (in.) (in.2) Area (in.2) Surface (per ft) Surface (per ft) per ft (lb) ft (lb) (in.4) (in.3) (in.)
— — 10S 0.049 0.307 0.0740 0.0548 0.106 0.0804 0.186 0.0321 0.00088 0.00437 0.1271
1/8 40 Std 40S 0.068 0.269 0.0568 0.0720 0.106 0.0705 0.245 0.0246 0.00106 0.00525 0.1215
0.405 80 XS 80S 0.095 0.215 0.0364 0.0925 0.106 0.0563 0.315 0.0157 0.00122 0.00600 0.1146
— — 10S 0.065 0.410 0.1320 0.0970 0.141 0.1073 0.330 0.0572 0.00279 0.01032 0.1694
1/4 40 Std 40S 0.088 0.364 0.1041 0.1250 0.141 0.0955 0.425 0.0451 0.00331 0.01230 0.1628
0.540 80 XS 80S 0.119 0.302 0.0716 0.1574 0.141 0.0794 0.535 0.0310 0.00378 0.01395 0.1547
— — 10S 0.065 0.545 0.2333 0.1246 0.177 0.1427 0.423 0.1011 0.00586 0.01737 0.2169
3/8 40 Std 40S 0.091 0.493 0.1910 0.1670 0.177 0.1295 0.568 0.0827 0.00730 0.02160 0.2090
0.675 80 XS 80S 0.126 0.423 0.1405 0.2173 0.177 0.1106 0.739 0.0609 0.00862 0.02554 0.1991
— — 10S 0.083 0.674 0.3570 0.1974 0.220 0.1765 0.671 0.1547 0.01431 0.0341 0.2692
40 Std 40S 0.109 0.622 0.3040 0.2503 0.220 0.1628 0.851 0.1316 0.01710 0.0407 0.2613
1/2 80 XS 80S 0.147 0.546 0.2340 0.3200 0.220 0.1433 1.088 0.1013 0.02010 0.0478 0.2505
0.840 160 — — 0.187 0.466 0.1706 0.3830 0.220 0.1220 1.304 0.0740 0.02213 0.0527 0.2402
— XXS — 0.294 0.252 0.0499 0.5040 0.220 0.0660 1.714 0.0216 0.02425 0.0577 0.2192
— — 5S 0.065 0.920 0.6650 0.2011 0.275 0.2409 0.684 0.2882 0.02451 0.0467 0.349
— — 10S 0.083 0.884 0.6140 0.2521 0.275 0.2314 0.857 0.2661 0.02970 0.0566 0.343
3/4 40 Std 40S 0.113 0.824 0.5330 0.3330 0.275 0.2157 1.131 0.2301 0.0370 0.0706 0.334
1.050 80 XS 80S 0.154 0.742 0.4320 0.4350 0.275 0.1943 1.474 0.1875 0.0448 0.0853 0.321
160 — — 0.218 0.614 0.2961 0.5700 0.275 0.1607 1.937 0.1284 0.0527 0.1004 0.304
— XXS — 0.308 0.434 0.1479 0.7180 0.275 0.1137 2.441 0.0641 0.0579 0.1104 0.284
— — 5S 0.065 1.185 1.1030 0.2553 0.344 0.3100 0.868 0.478 0.0500 0.0760 0.443
— — 10S 0.109 1.097 0.9450 0.4130 0.344 0.2872 1.404 0.409 0.0757 0.1151 0.428
1 40 Std 40S 0.133 1.049 0.8640 0.4940 0.344 0.2746 1.679 0.374 0.0874 0.1329 0.421
1.315 80 XS 80S 0.179 0.957 0.7190 0.6390 0.344 0.2520 2.172 0.311 0.1056 0.1606 0.407
160 — — 0.250 0.815 0.5220 0.8360 0.344 0.2134 2.844 0.2261 0.1252 0.1903 0.387
— XXS — 0.358 0.599 0.2818 1.0760 0.344 0.1570 3.659 0.1221 0.1405 0.2137 0.361
— — 5S 0.065 1.530 1.839 0.326 0.434 0.401 1.107 0.797 0.1038 0.1250 0.564
— — 10S 0.109 1.442 1.633 0.531 0.434 0.378 1.805 0.707 0.1605 0.1934 0.550
1 1/4 40 Std 40S 0.140 1.380 1.496 0.669 0.434 0.361 2.273 0.648 0.1948 0.2346 0.540
1.660 80 XS 80S 0.191 1.278 1.283 0.881 0.434 0.335 2.997 0.555 0.2418 0.2913 0.524
160 — — 0.250 1.160 1.057 1.107 0.434 0.304 3.765 0.458 0.2839 0.342 0.506
— XXS — 0.382 0.896 0.631 1.534 0.434 0.2346 5.214 0.2732 0.341 0.411 0.472
— — 5S 0.065 1.770 2.461 0.375 0.497 0.463 1.274 1.067 0.1580 0.1663 0.649
— — 10S 0.109 1.682 2.222 0.613 0.497 0.440 2.085 0.962 0.2469 0.2599 0.634
(continued)
453
TABLE B.2 (continued)
454

Schedule
Number
1 1/2 40 Std 40S 0.145 1.610 2.036 0.799 0.497 0.421 2.718 0.882 0.310 0.326 0.623
1.900 80 XS 80S 0.200 1.500 1.767 1.068 0.497 0.393 3.631 0.765 0.391 0.412 0.605
160 — — 0.281 1.338 1.406 1.429 0.497 0.350 4.859 0.608 0.483 0.508 0.581
— XXS — 0.400 1.100 0.950 1.885 0.497 0.288 6.408 0.412 0.568 0.598 0.549
— — 5S 0.065 2.245 3.960 0.472 0.622 0.588 1.604 1.716 0.315 0.2652 0.817
— — 10S 0.109 2.157 3.650 0.776 0.622 0.565 2.638 1.582 0.499 0.420 0.802
2 40 Std 40S 0.154 2.067 3.360 1.075 0.622 0.541 3.653 1.455 0.666 0.561 0.787
2.375 80 XS 80S 0.218 1.939 2.953 1.477 0.622 0.508 5.022 1.280 0.868 0.731 0.766
160 — — 0.343 1.689 2.240 2.190 0.622 0.442 7.444 0.971 1.163 0.979 0.729
— XXS — 0.436 1.503 1.774 2.656 0.622 0.393 9.029 0.769 1.312 1.104 0.703
— — 5S 0.083 2.709 5.76 0.728 0.753 0.709 2.475 2.499 0.710 0.494 0.988
— — 10S 0.120 2.635 5.45 1.039 0.753 0.690 3.531 2.361 0.988 0.687 0.975
2 1/2 40 Std 40S 0.203 2.469 4.79 1.704 0.753 0.646 5.793 2.076 1.530 1.064 0.947
2.875 80 XS 80S 0.276 2.323 4.24 2.254 0.753 0.608 7.661 1.837 0.193 1.339 0.924
160 — — 0.375 2.125 3.55 2.945 0.753 0.556 10.01 1.535 2.353 1.637 0.894
— XXS — 0.552 1.771 2.46 4.030 0.753 0.464 13.70 1.067 2.872 1.998 0.844
— — 5S 0.083 3.334 8.73 0.891 0.916 0.873 3.03 3.78 1.301 0.744 1.208
— — 10S 0.120 3.260 8.35 1.274 0.916 0.853 4.33 3.61 1.822 1.041 1.196
3 40 Std 40S 0.216 3.068 7.39 2.228 0.916 0.803 7.58 3.20 3.02 1.724 1.164
3.500 80 XS 80S 0.300 2.900 6.61 3.020 0.916 0.759 10.25 2.864 3.90 2.226 1.136
160 — — 0.437 2.626 5.42 4.210 0.916 0.687 14.32 2.348 5.03 2.876 1.094
— XXS — 0.600 2.300 4.15 5.470 0.916 0.602 18.58 1.801 5.99 3.43 1.047
— — 5S 0.083 3.834 11.55 1.021 1.047 1.004 3.47 5.01 1.960 0.980 1.385
3 1/2 — — 10S 0.120 3.760 11.10 1.463 1.047 0.984 4.97 4.81 2.756 1.378 1.372
4.000 40 Std 40S 0.226 3.548 9.89 2.68 1.047 0.929 9.11 4.28 4.79 2.394 1.337
80 XS 80S 0.318 3.364 8.89 3.68 1.047 0.881 12.51 3.85 6.28 3.14 1.307
— — 5S 0.083 4.334 14.75 1.152 1.178 1.135 3.92 6.40 2.811 1.249 1.562
— — 10S 0.120 4.260 14.25 1.651 1.178 1.115 5.61 6.17 3.96 1.762 1.549
4 40 Std 40S 0.237 4.026 12.73 3.17 1.178 1.054 10.79 5.51 7.23 3.21 1.510
4.500 80 XS 80S 0.337 3.826 11.50 4.41 1.178 1.002 14.98 4.98 9.61 4.27 1.477
120 — — 0.437 3.626 10.33 5.58 1.178 0.949 18.96 4.48 11.65 5.18 1.445
160 — — 0.531 3.438 9.28 6.62 1.178 0.900 22.51 4.02 13.27 5.90 1.416
— XXS — 0.674 3.152 7.80 8.10 1.178 0.825 27.54 3.38 15.29 6.79 1.374
— — 5S 0.109 5.345 22.44 1.868 1.456 1.399 6.35 9.73 6.95 2.498 1.929
— — 10S 0.134 5.295 22.02 2.285 1.456 1.386 7.77 9.53 8.43 3.03 1.920
5 40 Std 40S 0.258 5.047 20.01 4.30 1.456 1.321 14.62 8.66 15.17 5.45 1.878
5.563 80 XS 80S 0.375 4.813 18.19 6.11 1.456 1.260 20.78 7.89 20.68 7.43 1.839
120 — — 0.500 4.563 16.35 7.95 1.456 1.195 27.04 7.09 25.74 9.25 1.799
160 — — 0.625 4.313 14.61 9.70 1.456 1.129 32.96 6.33 30 10.8 1.760
— XXS — 0.750 4.063 12.97 11.34 1.456 1.064 38.55 5.62 33.6 12.1 1.722
— — 5S 0.109 6.407 32.20 2.231 1.734 1.677 5.37 13.98 11.85 3.58 2.304
— — 10S 0.134 6.357 31.70 2.733 1.734 1.664 9.29 13.74 14.4 4.35 2.295
6 40 Std 40S 0.280 6.065 28.89 5.58 1.734 1.588 18.97 12.51 28.14 8.5 2.245
6.625 80 XS 80S 0.432 5.761 26.07 8.40 1.734 1.508 28.57 11.29 40.5 12.23 2.195
120 — — 0.562 5.501 23.77 10.70 1.734 1.440 36.39 10.30 49.6 14.98 2.153
160 — — 0.718 5.189 21.15 13.33 1.734 1.358 45.30 9.16 59 17.81 2.104
— XXS — 0.864 4.897 18.83 15.64 1.734 1.282 53.16 8.17 66.3 20.03 2.060
— — 5S 0.109 8.407 55.5 2.916 2.258 2.201 9.91 24.07 26.45 6.13 3.01
— — 10S 0.148 8.329 54.5 3.94 2.258 2.180 13.40 23.59 35.4 8.21 3.00
20 — — 0.250 8.125 51.8 6.58 2.258 2.127 22.36 22.48 57.7 13.39 2.962
30 — — 0.277 8.071 51.2 7.26 2.258 2.113 24.70 22.18 63.4 14.69 2.953
40 Std 40S 0.322 7.981 50.0 8.40 2.258 2.089 28.55 21.69 72.5 16.81 2.938
8 60 — — 0.406 7.813 47.9 10.48 2.258 2.045 35.64 20.79 88.8 20.58 2.909
8.625 80 XS 80S 0.500 7.625 45.7 12.76 2.258 1.996 43.39 19.80 105.7 24.52 2.878
100 — — 0.593 7.439 43.5 14.96 2.258 1.948 50.87 18.84 121.4 28.14 2.847
120 — — 0.718 7.189 40.6 17.84 2.258 1.882 60.63 17.60 140.6 32.6 2.807
140 — — 0.812 7.001 38.5 19.93 2.258 1.833 67.76 16.69 153.8 35.7 2.777
— XXS — 0.875 6.875 37.1 21.30 2.258 1.800 72.42 16.09 162 37.6 2.757
160 —. — 0.906 6.813 36.5 21.97 2.258 1.784 74.69 15.80 165.9 38.5 2.748
— — 5S 0.134 10.482 86.3 4.52 2.815 2.744 15.15 37.4 63.7 11.85 3.75
— — 10S 0.165 10.420 85.3 5.49 2.815 2.728 18.70 36.9 76.9 14.3 3.74
20 — — 0.250 10.250 82.5 8.26 2.815 2.683 28.04 35.8 113.7 21.16 3.71
— — — 0.279 10.192 81.6 9.18 2.815 2.668 31.20 35.3 125.9 23.42 3.70
30 — — 0.307 10.136 80.7 10.07 2.815 2.654 34.24 35.0 137.5 25.57 3.69
10 40 Std 40S 0.365 10.020 78.9 11.91 2.815 2.623 40.48 34.1 160.8 29.9 3.67
10.750 60 XS 80S 0.500 9.750 74.7 16.10 2.815 2.553 54.74 32.3 212 39.4 3.63
80 — — 0.593 9.564 71.8 18.92 2.815 2.504 64.33 31.1 244.9 45.6 3.60
100 — — 0.718 9.314 68.1 22.63 2.815 2.438 76.93 29.5 286.2 53.2 3.56
120 — — 0.843 9.064 64.5 26.24 2.815 2.373 89.20 28.0 324 60.3 3.52
140 — — 1.000 8.750 60.1 30.6 2.815 2.291 104.13 26.1 368 68.4 3.47
160 — — 1.125 8.500 56.7 34.0 2.815 2.225 115.65 24.6 399 74.3 3.43
— — 5S 0.165 12.420 121.2 6.52 3.34 3.25 19.56 52.5 129.2 20.27 4.45
— — 10S 0.180 12.390 120.6 7.11 3.34 3.24 24.20 52.2 140.5 22.03 4.44
20 — — 0.250 12.250 117.9 9.84 3.34 3.21 33.38 51.1 191.9 30.1 4.42
30 — — 0.330 12.090 114.8 12.88 3.34 3.17 43.77 49.7 248.5 39.0 4.39
— Std 40S 0.375 12.000 113.1 14.58 3.34 3.14 49.56 49.0 279.3 43.8 4.38
(continued)
455
456

Schedule
Number
12 40 — — 0.406 11.938 111.9 15.74 3.34 3.13 53.53 48.5 300 47.1 4.37
12.750 — XS 80S 0.500 11.750 108.4 19.24 3.34 3.08 65.42 47.0 362 56.7 4.33
60 — — 0.562 11.626 106.2 21.52 3.34 3.04 73.16 46.0 401 62.8 4.31
80 — — 0.687 11.376 101.6 26.04 3.34 2.978 88.51 44.0 475 74.5 4.27
100 — — 0.843 11.064 96.1 31.5 3.34 2.897 107.20 41.6 562 88.1 4.22
120 — — 1.000 10.750 90.8 36.9 3.34 2.814 125.49 39.3 642 100.7 4.17
140 — — 1.125 10.500 86.6 41.1 3.34 2.749 139.68 37.5 701 109.9 4.13
160 — — 1.312 10.126 80.5 47.1 3.34 2.651 160.27 34.9 781 122.6 4.07
10 — — 0.250 13.500 143.1 10.80 3.67 3.53 36.71 62.1 255.4 36.5 4.86
20 — — 0.312 13.376 140.5 13.42 3.67 3.5 45.68 60.9 314 44.9 4.84
30 Std — 0.375 13.250 137.9 16.05 3.67 3.47 54.57 59.7 373 53.3 4.82
40 — — 0.437 13.126 135.3 18.62 3.67 3.44 63.37 58.7 429 61.2 4.80
— XS — 0.500 13.000 132.7 21.21 3.67 3.4 72.09 57.5 484 69.1 4.78
— — — 0.562 12.876 130.2 23.73 3.67 3.37 80.66 56.5 537 76.7 4.76
14 60 — — 0.593 12.814 129.0 24.98 3.67 3.35 84.91 55.9 562 80.3 4.74
14.000 — — — 0.625 12.750 127.7 26.26 3.67 3.34 89.28 55.3 589 84.1 4.73
— — — 0.687 12.626 125.2 28.73 3.67 3.31 97.68 54.3 638 91.2 4.71
80 — — 0.750 12.500 122.7 31.2 3.67 3.27 106.13 53.2 687 98.2 4.69
— — — 0.875 12.250 117.9 36.1 3.67 3.21 122.66 51.1 781 111.5 4.65
100 — — 0.937 12.126 115.5 38.5 3.67 3.17 130.73 50.0 825 117.8 4.63
120 — — 1.093 11.814 109.6 44.3 3.67 3.09 150.67 47.5 930 132.8 4.58
140 — — 1.250 11.500 103.9 50.1 3.67 3.01 170.22 45.0 1127 146.8 4.53
160 — — 1.406 11.188 98.3 55.6 3.67 2.929 189.12 42.6 1017 159.6 4.48
10 — — 0.250 15.500 188.7 12.37 4.19 4.06 42.05 81.8 384 48 5.57
20 — — 0.312 15.376 185.7 15.38 4.19 4.03 52.36 80.5 473 59.2 5.55
30 Std — 0.375 15.250 182.6 18.41 4.19 3.99 62.58 79.1 562 70.3 5.53
— — — 0.437 15.126 179.7 21.37 4.19 3.96 72.64 77.9 648 80.9 5.50
40 XS — 0.500 15.000 176.7 24.35 4.19 3.93 82.77 76.5 732 91.5 5.48
— — — 0.562 14.876 173.8 27.26 4.19 3.89 92.66 75.4 813 106.6 5.46
— — — 0.625 14.750 170.9 30.2 4.19 3.86 102.63 74.1 894 112.2 5.44
16 60 — — 0.656 14.688 169.4 31.6 4.19 3.85 107.50 73.4 933 116.6 5.43
16.000 — — — 0.687 14.626 168.0 33.0 4.19 3.83 112.36 72.7 971 121.4 5.42
— — 0.750 14.500 165.1 35.9 4.19 3.8 122.15 71.5 1047 130.9 5.40
80 — — 0.842 14.314 160.9 40.1 4.19 3.75 136.46 69.7 1157 144.6 5.37
— — — 0.875 14.250 159.5 41.6 4.19 3.73 141.35 69.1 1193 154.1 5.36
100 — — 1.031 13.938 152.6 48.5 4.19 3.65 164.83 66.1 1365 170.6 5.30
120 — — 1.218 13.564 144.5 56.6 4.19 3.55 192.29 62.6 1556 194.5 5.24
140 — — 1.437 13.126 135.3 65.7 4.19 3.44 223.50 58.6 1760 220.0 5.17
160 — — 1.593 12.814 129.0 72.1 4.19 3.35 245.11 55.9 1894 236.7 5.12
10 — — 0.250 17.500 240.5 13.94 4.71 4.58 47.39 104.3 549 61.0 6.28
20 — — 0.312 17.376 237.1 17.34 4.71 4.55 59.03 102.8 678 75.5 6.25
— Std — 0.375 17.250 233.7 20.76 4.71 4.52 70.59 101.2 807 89.6 6.23
30 — — 0.437 17.126 230.4 24.11 4.71 4.48 82.06 99.9 931 103.4 6.21
— XS — 0.500 17.000 227.0 27.49 4.71 4.45 93.45 98.4 1053 117.0 6.19
40 — — 0.562 16.876 223.7 30.8 4.71 4.42 104.75 97.0 1172 130.2 6.17
— — — 0.625 16.750 220.5 34.1 4.71 4.39 115.98 95.5 1289 143.3 6.15
18 — — — 0.687 16.626 217.1 37.4 4.71 4.35 127.03 94.1 1403 156.3 6.13
18.000 60 — — 0.750 16.500 213.8 40.6 4.71 4.32 138.17 92.7 1515 168.3 6.10
— — — 0.875 16.250 207.4 47.1 4.71 4.25 160.04 89.9 1731 192.8 6.06
80 — — 0.937 16.126 204.2 50.2 4.71 4.22 170.75 88.5 1834 203.8 6.04
100 — — 1.156 15.688 193.3 61.2 4.71 4.11 207.96 83.7 2180 242.2 5.97
120 — — 1.375 15.250 182.6 71.8 4.71 3.99 244.14 79.2 2499 277.6 5.90
140 — — 1.562 14.876 173.8 80.7 4.71 3.89 274.23 75.3 2750 306 5.84
160 — — 1.781 14.438 163.7 90.7 4.71 3.78 308.51 71.0 3020 336 5.77
10 — — 0.250 19.500 298.6 15.51 5.24 5.11 52.73 129.5 757 75.7 6.98
— — — 0.312 19.376 294.9 19.30 5.24 5.07 65.40 128.1 935 93.5 6.96
20 Std — 0.375 19.250 291.0 23.12 5.24 5.04 78.60 126.0 1114 111.4 6.94
— — — 0.437 19.126 287.3 26.86 5.24 5.01 91.31 124.6 1286 128.6 6.92
30 XS — 0.500 19.000 283.5 30.6 5.24 4.97 104.13 122.8 1457 145.7 6.90
— — — 0.562 18.876 279.8 34.3 5.24 4.94 116.67 121.3 1624 162.4 6.88
20 40 — — 0.593 18.814 278.0 36.2 5.24 4.93 122.91 120.4 1704 170.4 6.86
20.000 — — — 0.625 18.750 276.1 38.0 5.24 4.91 129.33 119.7 1787 178.7 6.85
— — — 0.687 18.626 272.5 41.7 5.24 4.88 141.71 118.1 1946 194.6 6.83
— — — 0.750 18.500 268.8 45.4 5.24 4.84 154.20 116.5 2105 210.5 6.81
60 — — 0.812 18.376 265.2 48.9 5.24 4.81 166.40 115.0 2257 225.7 6.79
— — — 0.875 18.250 261.6 52.6 5.24 4.78 178.73 113.4 2409 240.9 6.77
80 — — 1.031 17.938 252.7 61.4 5.24 4.70 208.87 109.4 2772 277.2 6.72
100 — — 1.281 17.438 238.8 75.3 5.24 4.57 256.10 103.4 3320 332 6.63
120 — — 1.500 17.000 227.0 87.2 5.24 4.45 296.37 98.3 3760 376 6.56
140 — — 1.750 16.500 213.8 100.3 5.24 4.32 341.10 92.6 4220 422 6.48
160 — — 1.968 16.064 202.7 111.5 5.24 4.21 379.01 87.9 4590 459 6.41
10 — — 0.250 23.500 434 18.65 6.28 6.15 63.41 188.0 1316 109.6 8.40
— — — 0.312 23.376 430 23.20 6.28 6.12 78.93 186.1 1629 135.8 8.38
20 Std — 0.375 23.250 425 27.83 6.28 6.09 94.62 183.8 1943 161.9 8.35
— — — 0.437 23.126 420 32.4 6.28 6.05 109.97 182.1 2246 187.4 8.33
— XS — 0.500 23.000 415 36.9 6.28 6.02 125.49 180.1 2550 212.5 8.31
(continued)
457
458

Schedule
Number
24 30 — — 0.562 22.876 411 41.4 6.28 5.99 140.80 178.1 2840 237.0 8.29
24.000 — — — 0.625 22.750 406 45.9 6.28 5.96 156.03 176.2 3140 261.4 8.27
40 — — 0.687 22.626 402 50.3 6.28 5.92 171.17 174.3 3420 285.2 8.25
— — — 0.750 22.500 398 54.8 6.28 5.89 186.24 172.4 3710 309 8.22
60 — — 0.968 22.064 382 70.0 6.28 5.78 238.11 165.8 4650 388 8.15
80 — — 1.218 21.564 365 87.2 6.28 5.65 296.36 158.3 5670 473 8.07
100 — — 1.531 20.938 344 108.1 6.28 5.48 367.40 149.3 6850 571 7.96
120 — — 1.812 20.376 326 126.3 6.28 5.33 429.39 141.4 7830 652 7.87
140 — — 2.062 19.876 310 142.1 6.28 5.20 483.13 134.5 8630 719 7.79
160 — — 2.343 19.314 293 159.4 6.28 5.06 541.94 127.0 9460 788 7.70
10 — — 0.312 29.376 678 29.1 7.85 7.69 98.93 293.8 3210 214 10.50
30 20 — — 0.500 29.000 661 46.3 7.85 7.59 157.53 286.3 5040 336 10.43
30.000 30 — — 0.625 28.750 649 57.6 7.85 7.53 196.08 281.5 6220 415 10.39
a ASA B36.10 Steel-pipe schedule numbers.
b ASA B36.10 Steel-pipe nominal wall-thickness designations.
c ASA B36.19 Stainless-steel-pipe schedule numbers.

TABLE B.3
Commercial Copper Tubinga
Size, OD Wall Thickness Flow Area Surface Area
Metal Inside Outside Weight
in. mm in. mm gage in.2 mm2 Area (in.2) (ft2/ft) (ft2/ft) (lb/ft)
1/8 3.2 0.030 0.76 A 0.003 1.9 0.012 0.017 0.033 0.035
3/16 4.76 0.030 0.76 A 0.013 8.4 0.017 0.034 0.049 0.058
1/4 6.4 0.030 0.76 A 0.028 18.1 0.021 0.050 0.066 0.080
1/4 6.4 0.049 1.24 18 0.018 11.6 0.031 0.038 0.066 0.120
5/16 7.94 0.032 0.81 21A 0.048 31.0 0.028 0.065 0.082 0.109
3/8 9.53 0.032 0.81 21A 0.076 49.0 0.033 0.081 0.098 0.134
3/8 9.53 0.049 1.24 18 0.060 38.7 0.050 0.072 0.098 0.195
1/2 12.7 0.032 0.81 21A 0.149 96.1 0.047 0.114 0.131 0.182
1/2 12.7 0.035 0.89 20L 0.145 93.6 0.051 0.113 0.131 0.198
1/2 12.7 0.049 1.24 18K 0.127 81.9 0.069 0.105 0.131 0.269
1/2 12.7 0.065 1.65 16 0.108 69.7 0.089 0.97 0.131 0.344
5/8 15.9 0.035 0.89 20A 0.242 156 0.065 0.145 0.164 0.251
5/8 15.9 0.040 1.02 L 0.233 150 0.074 0.143 0.164 0.285
5/8 15.9 0.049 1.24 18K 0.215 139 0.089 0.138 0.164 0.344
3/4 19.1 0.035 0.89 20A 0.363 234 0.079 0.178 0.196 0.305
3/4 19.1 0.042 1.07 L 0.348 224 0.103 0.174 0.196 0.362
3/4 19.1 0.049 1.24 18K 0.334 215 0.108 0.171 0.196 0.418
3/4 19.1 0.065 1.65 16 0.302 195 0.140 0.162 0.196 0.542
3/4 19.1 0.083 2.11 14 0.268 173 0.174 0.151 0.196 0.674
7/8 22.2 0.045 1.14 L 0.484 312 0.117 0.206 0.229 0.455
7/8 22.2 0.065 1.65 16K 0.436 281 0.165 0.195 0.229 0.641
7/8 22.2 0.083 2.11 14 0.395 255 0.206 0.186 0.229 0.800
1 25.4 0.065 1.65 16 0.594 383 0.181 0.228 0.262 0.740
1 25.4 0.083 2.11 14 0.546 352 0.239 0.218 0.262 0.927
1 1/8 28.6 0.050 1.27 L 0.825 532 0.176 0.268 0.294 0.655
1 1/8 28.6 0.065 1.65 16K 0.778 502 0.216 0.261 0.294 0.839
1 1/4 31.8 0.065 1.65 16 0.985 636 0.242 0.293 0.327 0.938
1 1/4 31.8 0.083 2.11 14 0.923 596 0.304 0.284 0.327 1.18
1 3/8 34.9 0.055 1.40 L 1.257 811 0.228 0.331 0.360 0.884
1 3/8 34.9 0.065 1.65 16K 1.217 785 0.267 0.326 0.360 1.04
1 1/2 38.1 0.065 1.65 16 1.474 951 0.294 0.359 0.393 1.14
1 1/2 38.7 0.083 2.11 14 1.398 902 0.370 0.349 0.393 1.43
1 5/8 41.3 0.060 1.52 L 1.779 1148 0.295 0.394 0.425 1.14
1 5/8 41.3 0.072 1.83 K 1.722 1111 0.351 0.388 0.425 1.36
2 50.8 0.083 2.11 14 2.642 1705 0.500 0.480 0.628 1.94
2 50.8 0.109 2.76 12 2.494 1609 0.620 0.466 0.628 2.51
2 1/8 54.0 0.070 1.78 L 3.095 1997 0.449 0.520 0.556 1.75
2 1/8 54.0 0.083 2.11 14K 3.016 1946 0.529 0.513 0.556 2.06
2 5/8 66.7 0.080 2.03 L 4.77 3078 0.645 0.645 0.687 2.48
2 5/8 66.7 0.095 2.41 13K 4.66 3007 0.760 0.637 0.687 2.93
3 1/8 79.4 0.090 2.29 L 6.81 4394 0.950 0.771 0.818 3.33
3 1/8 79.4 0.109 2.77 12K 6.64 4284 1.034 0.761 0.818 4.00
(continued)

Commercial Copper Tubinga
Size, OD Wall Thickness Flow Area Surface Area
Metal Inside Outside Weight
in. mm in. mm gage in.2 mm2 Area (in.2) (ft2/ft) (ft2/ft) (lb/ft)
3 5/8 92.1 0.100 2.54 L 9.21 5942 1.154 0.897 0.949 4.29
3 5/8 92.1 0.120 3.05 11K 9.00 5807 1.341 0.886 0.949 5.12
4 1/8 104.8 0.110 2.79 L 11.92 7691 1.387 1.022 1.080 5.38
4 1/8 104.8 0.134 3.40 10K 11.61 7491 1.682 1.009 1.080 6.51
Sources: Kreith, F., The CRC Handbook of Mechanical Engineering, CRC Press, Boca Raton, FL, 1998; Baukal, C.E.
(ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
Notes: The table above gives dimensional data and weights of copper tubing used for automotive, plumbing,
refrigeration, and heat exchanger services. For additional data see the standards handbooks of the
Copper Development Association, Inc., the ASTM standards, and the “SAE Handbook.”
Dimensions in this table are actual specified measurements, subject to accepted tolerances. Trade size
designations are usually by actual OD, except for water and drainage tube (plumbing), which
measures 1/8 in. larger OD. A 1/2 in. plumbing tube, for example, measures 5/8 in. OD, and a 2 in.
plumbing tube measures 2 1/8 in. OD.
Key to gage sizes
Standard-gage wall thicknesses are listed by numerical designation (14–21), BWG or Stubs gage. These
gage sizes are standard for tubular heat exchangers. The letter A designates SAE tubing sizes for
automotive service. Letter designations K and L are the common sizes for plumbing services, soft or
hard temper.
Other materials
These same dimensional sizes are also common for much of the commercial tubing available in
aluminum, mild steel, brass, bronze, and other alloys. Tube weights in this table are based on copper
at 0.323 lb/in3. For other materials the weights should be multiplied by the following approximate
factors:
Aluminum 0.30
Mild steel 0.87
Brass 0.95
Monel 0.96
Stainless steel 0.89
a Compiled and computed.
TABLE B.4
SAE Grades for Steel Bolts
Proof Strength Tensile
SAE Grade No. Size Range Incl. (kpsi)a Strength (kpsi)a Material Head Marking
1 1
1 to 1 Low- or medium-
4 2 carbon steel
1 3
2 to 55 74
4 4
7 1
to 1 33 60
8 2
1
5 to 1 85 120 Medium-carbon steel,
4 Q&T
1 1
1 to 1 74 105
8 2
1
5.2 to 1 85 120 Low-carbon martensite
4
steel, Q & T
1 1
7 to 1 105 133 Medium-carbon alloy
4 2
steel, Q & Tb
1 1
8 to 1 120 150 Medium-carbon alloy
4 2 steel, Q & T
1
8.2 to 1 120 150 Low-carbon martensite
4 steel, Q & T
Sources: Russel, Burdsall & Ward Corp., Helpful Hints, Mentor, OH, Chapter 23; Kreith, F., The CRC Handbook of Mechanical
Engineering, CRC Press, Boca Raton, FL, 1998; Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press,
Boca Raton, FL, 2001.
a Minimum values.
b Roll threaded after heat treatment.

TABLE B.5
ASTM Grades for Steel Bolts
Proof Strength Tensile
ASTM Designation Size Range Incl. (kpsi)a Strength (kpsi)a Material Head Marking
1
A307 to 4 Low-carbon steel
4
1
A325 type 1 to 1 85 120 Medium-carbon steel, Q & T
2 A325
1 1
1 to 1 74 105
8 2
1
A325 type 2 to 1 85 120 Low-carbon steel, Q & T
2
A325
1 1
1 to 1 74 105
8 2
1
A325 type 3 to 1 85 120 Weathering steel, Q & T
2 A325
1 1
1 to 1 74 105
8 2
A354 grade BC Alloy steel, Q & T

BC
1
A354 grade BD to 4 120 150 Alloy steel, Q & T
4
1
A449 to 1 85 120 Medium-carbon steel. Q & T
4
1 1
1 to 1 74 105
8 2
3
1 to 3 55 90
4
1 1
A490 type to 1 120 150 Alloy steel, Q & T
2 2 A490
A490 type 3 Weathering steel, Q & T

A490
Sources: Russel, Burdsall & Ward Corp., Helpful Hints, Mentor, OH, Chapter 23; Kreith, F., The CRC Handbook of Mechanical
Engineering, CRC Press, Boca Raton, FL, 1998; Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca
Raton, FL, 2001.
a Minimum values.
TABLE B.6
Metric Mechanical Property Classes for Steel Bolts, Screws, and Studs
Proof Strength Tensile Strength
Property Class Size Range Incl. (MPa) (MPa) Material Head Marking
4.6 M5–M36 225 400 Low- or medium-carbon steel
4.6
4.8 M1.6–M16 310 420 Low- or medium-carbon steel

4.8
5.8 M5–M24 380 520 Low- or medium-carbon steel

5.8
8.8 M16–M36 600 830 Medium-carbon steel, Q & T

8.8
9.8 M1.6–M16 650 900 Medium-carbon steel Q & T

9.8
10.9 M5–M36 830 1040 Low-carbon martensite steel, Q & T

10.9
12.9 M1.6–M36 970 1220 Alloy steel, Q & T

12.9
Sources: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001. Kreith, F., The CRC Handbook of
Mechanical Engineering, CRC Press, Boca Raton, FL, 1998; Russel, Burdsall & Ward Corp., Helpful Hints, Mentor, OH,
Chapter 23.
References
1. Baukal, C.E. (ed.), The John Zink Combustion Handbook,
2. Kreith, F., The CRC Handbook of Mechanical Engineering,
3. Russel, Bursdall & Ward Corp., Helpful Hints, Mentor,
OH, Chapter 23.
Appendix C: Properties of Gases and Liquids
TABLE C.1 TABLE C.1 (continued)

Ideal Gas Properties of Air Ideal Gas Properties of Air
(SI Units) (SI Units)
s° (kJ/ s° (kJ/
T (K) h (kJ/kg) pr u (kJ/kg) vr kg k) T (K) h (kJ/kg) pr u (kJ/kg) vr kg k)
200 199.97 0.3363 142.56 1707.0 1.29559 590 596.52 15.31 427.15 110.6 2.39140
210 209.97 0.3987 149.69 1512.0 1.34444 600 607.02 16.28 434.78 105.8 2.40902
220 219.97 0.4690 156.82 1346.0 1.39105 610 617.53 17.30 442.42 101.2 2.42644
230 230.02 0.5477 164.00 1205.0 1.43557 620 628.07 18.36 450.09 96.92 2.44356
240 240.02 0.6355 171.13 1084.0 1.47824 630 638.63 19.84 457.78 92.84 2.46048
250 250.05 0.7329 178.28 979.0 1.51917 640 649.22 20.64 465.50 88.99 2.47716
260 260.09 0.8405 185.45 887.8 1.55848 650 659.84 21.86 473.25 85.34 2.49364
270 270.11 0.9590 192.60 808.0 1.59634 660 670.47 23.13 481.01 81.89 2.50985
280 280.13 1.0889 199.75 738.0 1.63279 670 681.14 24.46 488.81 78.61 2.52589
285 285.14 1.1584 203.33 760.1 1.65055 680 691.82 25.85 496.62 75.50 2.54175
290 290.16 1.2311 206.91 676.1 1.66802 690 702.52 27.29 504.45 72.56 2.55731
295 295.17 1.3068 210.49 647.9 1.68515 700 713.27 28.80 512.33 69.76 2.57277
300 300.19 1.3860 214.07 621.2 1.70203 710 724.04 30.38 520.23 67.07 2.58810
305 305.22 1.4686 217.67 596.0 1.71865 720 734.82 32.02 528.14 64.53 2.60319
310 310.24 1.5546 221.25 572.3 1.73498 730 745.62 33.72 536.07 62.13 2.61803
315 315.27 1.6442 224.85 549.8 1.75106 740 756.44 35.50 544.02 59.82 2.63280
320 320.29 1.7375 228.42 528.6 1.76690 750 767.29 37.35 551.99 57.63 2.64737
325 325.31 1.8345 232.02 508.4 1.78249 760 778.18 39.27 560.01 55.54 2.66176
330 330.34 1.9352 235.61 489.4 1.79783 770 789.11 41.31 568.07 53.39 2.67595
340 340.42 2.149 242.82 454.1 1.82790 780 800.03 43.35 576.12 51.64 2.69013
350 350.49 2.379 250.02 422.2 1.85708 790 810.99 45.55 584.21 49.86 2.70400
360 360.58 2.626 257.24 393.4 1.88543 800 821.95 47.75 592.30 48.08 2.71787
370 370.67 2.892 264.46 367.2 1.91313 820 843.98 52.59 608.59 44.84 2.74504
380 380.77 3.176 271.69 343.4 1.94001 840 866.08 57.60 624.95 41.85 2.77170
390 390.88 3.481 278.93 321.5 1.96633 860 888.27 63.09 641.40 39.12 2.79783
400 400.98 3.806 286.16 301.6 1.99194 880 910.56 68.98 657.95 36.61 2.82344
410 411.12 4.153 293.43 283.3 2.01699 900 932.93 75.29 674.58 34.31 2.84856
420 421.26 4.522 300.69 266.6 2.04142 920 955.38 82.05 691.28 32.18 2.87324
430 431.43 4.915 307.99 251.1 2.06533 940 977.92 89.28 708.08 30.22 2.89748
440 441.61 5.332 315.30 236.8 2.08870 960 1000.55 97.00 725.02 28.40 2.92128
450 451.80 5.775 322.62 223.6 2.11161 980 1023.25 105.2 741.98 26.73 2.94468
460 462.02 6.245 329.97 211.4 2.13407 1000 1046.04 114.0 758.94 25.17 2.96770
470 472.24 6.742 337.32 200.1 2.15604 1020 1068.89 123.4 776.10 23.72 2.99034
480 482.49 7.268 344.70 189.5 2.17760 1040 1091.85 133.3 793.36 22.39 3.01260
490 492.74 7.824 352.08 179.7 2.19876 1060 1114.86 143.9 810.62 21.14 3.03449
500 503.02 8.411 359.49 170.6 2.21952 1080 1137.89 155.2 827.88 19.98 3.05608
510 513.32 9.031 366.92 162.1 2.23993 1100 1161.07 167.1 845.33 18.896 3.07732
520 523.63 9.684 374.36 154.1 2.25997 1120 1184.28 179.7 862.79 17.886 3.09825
530 533.98 10.37 381.84 146.7 2.27967 1140 1207.57 193.1 880.35 16.946 3.11883
540 544.35 11.10 389.34 139.7 2.29906 1160 1230.92 207.2 897.91 16.064 3.13916
550 554.74 11.68 396.86 133.1 2.31809 1180 1254.34 222.2 915.57 15.241 3.15916
560 565.17 12.66 404.42 127.0 2.33685 1200 1277.79 238.0 933.33 14.470 3.17888
570 575.59 13.50 411.97 121.2 2.35531 1220 1301.31 254.7 951.09 13.747 3.19834
580 586.04 14.38 419.55 115.7 2.37348 1240 1324.93 272.3 968.95 13.069 3.21751
(continued)
465
466 Appendix C: Properties of Gases and Liquids
TABLE C.1 (continued) TABLE C.2

(SI Units) (English Units)
s° (kJ/ T h (Btu/ u (Btu/ s° (Btu/
T (K) h (kJ/kg) pr u (kJ/kg) vr kg k) (°R) lb) pr lb) vr lb °R)
1260 1348.55 290.8 986.90 12.435 3.23638 360 85.97 0.3363 61.29 396.6 0.50369
1280 1372.24 310.4 1004.76 11.835 3.25510 380 90.75 0.4061 64.70 346.6 0.51663
1300 1395.97 330.9 1022.82 11.275 3.27345 400 95.53 0.4858 68.11 305.0 0.52890
1320 1419.76 352.5 1040.88 10.747 3.29160 420 100.32 0.5760 71.52 270.1 0.54058
1340 1443.60 375.3 1058.94 10.247 3.30959 440 105.11 0.6776 74.93 240.6 0.55172
1360 1467.49 399.1 1077.10 9.780 3.32724 460 109.90 0.7913 78.36 215.33 0.56235
1380 1491.44 424.2 1095.26 9.337 3.34474 480 114.69 0.9182 81.77 193.65 0.57255
1400 1515.42 450.5 1113.52 8.919 3.36200 500 119.48 1.0590 85.20 174.90 0.58233
1420 1539.44 478.0 1131.77 8.526 3.37901 520 124.27 1.2147 88.62 158.58 0.59172
1440 1563.51 506.9 1150.13 8.153 3.39586 537 128.34 1.3593 91.53 146.34 0.59945
1460 1587.63 537.1 1168.49 7.801 3.41247 540 129.06 1.3860 92.04 144.32 0.60078
1480 1611.79 568.8 1186.95 7.468 3.42892 560 133.86 1.5742 95.47 131.78 0.60950
1500 1635.97 601.9 1205.41 7.152 3.44516 580 138.66 1.7800 98.90 120.70 0.61793
1520 1660.23 636.5 1223.87 6.854 3.46120 600 143.47 2.005 102.34 110.88 0.62607
1540 1684.51 672.8 1242.43 6.569 3.47712 620 148.28 2.249 105.78 102.12 0.63395
1560 1708.82 710.5 1260.99 6.301 3.49276 640 153.09 2.514 109.21 94.30 0.64159
1580 1733.17 750.0 1279.65 6.046 3.50829 660 157.92 2.801 112.67 87.27 0.64902
1600 1757.57 791.2 1298.30 5.804 3.52364 680 162.73 3.111 116.12 80.96 0.65621
1620 1782.00 834.1 1316.96 5.574 3.53879 700 167.56 3.446 119.58 75.25 0.66321
1640 1806.46 878.9 1335.72 5.355 3.55381 720 172.39 3.806 123.04 70.07 0.67002
1660 1830.96 925.6 1354.48 5.147 3.56867 740 177.23 4.193 126.51 65.38 0.67665
1680 1855.50 974.2 1373.24 4.949 3.58335 760 182.08 4.607 129.99 61.10 0.68312
1700 1880.1 1025 1392.7 4.761 3.5979 780 186.94 5.051 133.47 57.20 0.68942
1750 1941.6 1161 1439.8 4.328 3.6336 800 191.81 5.526 136.97 53.63 0.69558
1800 2003.3 1310 1487.2 3.944 3.6684 820 196.69 6.033 140.47 50.35 0.70160
1850 2065.3 1475 1534.9 3.601 3.7023 840 201.56 6.573 143.98 47.34 0.70747
1900 2127.4 1655 1582.6 3.295 3.7354 860 206.46 7.149 147.50 44.57 0.71323
1950 2189.7 1852 1630.6 3.022 3.7677 880 211.35 7.761 151.02 42.01 0.71886
2000 2252.1 2068 1678.7 2.776 3.7994 900 216.26 8.411 154.57 39.64 0.72438
2050 2314.6 2303 1726.3 2.555 3.8303 920 221.18 9.102 158.12 37.44 0.72979
2100 2377.4 2559 1775.3 2.356 3.8605 940 226.11 9.834 161.68 35.41 0.73509
2150 2440.3 2837 1823.8 2.175 3.8901 960 231.06 10.61 165.26 33.52 0.74030
2200 2503.2 3138 1872.4 2.012 3.9191 980 236.02 11.43 168.83 31.76 0.74540
2250 2566.4 3464 1921.3 1.864 3.9474 1000 240.98 12.30 172.43 30.12 0.75042
1040 250.95 14.18 179.66 27.17 0.76019
Source: Kreith, F. (ed.), The CRC Handbook of Thermal
1080 260.97 16.28 186.93 24.58 0.76964
Engineering, CRC Press, Boca Raton, FL, 2000.
Note: The properties pr (relative pressure) and vr (relative 1120 271.03 18.60 194.25 22.30 0.77880
specific volume) are dimensionless quantities used in the 1160 281.14 21.18 201.63 20.29 0.78767
analysis of isentropic processes, and should not be con- 1200 291.30 24.01 209.05 18.51 0.79628
fused with the properties pressure and specific volume. 1240 301.52 27.13 216.53 16.93 0.80466
h is enthalpy, u is internal energy, and s° is entropy. 1280 311.79 30.55 224.05 15.52 0.81280
1320 322.11 34.31 231.63 14.25 0.82075
1360 332.48 38.41 239.25 13.12 0.82848
1400 342.90 42.88 246.93 12.10 0.83604
1440 353.37 47.75 254.66 11.17 0.84341
1480 363.89 53.04 262.44 10.34 0.85062
1520 374.47 58.78 270.26 9.578 0.85767
1560 385.08 65.00 278.13 8.890 0.86456
1600 395.74 71.73 286.06 8.263 0.87130
1650 409.13 80.89 296.03 7.556 0.87954
Appendix C: Properties of Gases and Liquids 467
TABLE C.2 (continued) TABLE C.2 (continued)

(English Units) (English Units)
T h (Btu/ u (Btu/ s° (Btu/ T h (Btu/ u (Btu/ s° (Btu/
(°R) lb) pr lb) vr lb °R) (°R) lb) pr lb) vr lb °R)
1700 422.59 90.95 306.06 6.924 0.88758 4150 1133.6 3858 849.09 0.3985 1.14446
1750 436.12 101.98 316.16 6.357 0.89542 4200 1148.7 4067 860.81 0.3826 1.14809
1800 449.71 114.0 326.32 5.847 0.90308 4300 1179.0 4513 884.28 0.3529 1.15522
1850 463.37 127.2 336.55 5.388 0.91056 4400 1209.4 4997 907.81 0.3262 1.16221
1900 477.09 141.5 346.85 4.974 0.91788 4500 1239.9 5521 931.39 0.3019 1.16905
1950 490.88 157.1 357.20 4.598 0.92504 4600 1270.4 6089 955.04 0.2799 1.17575
2000 504.71 174.0 367.61 4.258 0.93205 4700 1300.9 6701 978.73 0.2598 1.18232
2050 518.61 192.3 378.08 3.949 0.93891 4800 1331.5 7362 1002.5 0.2415 1.18876
2100 532.55 212.1 388.60 3.667 0.94564 4900 1362.2 8073 1026.3 0.2248 1.19508
2150 546.54 233.5 399.17 3.410 0.95222 5000 1392.9 8837 1050.1 0.2096 1.20129
2200 560.59 256.6 409.78 3.176 0.95868 5100 1423.6 9658 1074.0 0.1956 1.20738
2250 574.69 281.4 420.46 2.961 0.96501 5200 1454.4 10539 1098.0 0.1828 1.21336
2300 588.82 308.1 431.16 2.765 0.97123 5300 1485.3 11481 1122.0 0.1710 1.21923
2350 603.00 336.8 441.91 2.585 0.97732
Sources: Adapted from Moran, M.J. and Shapiro, H.N.,
2400 617.22 367.6 452.70 2.419 0.98331
Fundamentals of Engineering Thermodynamics, 3rd edn.,
2450 631.48 400.5 463.54 2.266 0.98919 Wiley, New York, 1995, as based on Keenan, J.H. and
2500 645.78 435.7 474.40 2.125 0.99497 Kaye, J., Gas Tables, Wiley, New York, 1945; Kreith, F.
2550 660.12 473.3 485.31 1.996 1.00064 (ed.), The CRC Handbook of Thermal Engineering, CRC
2600 674.49 513.5 496.26 1.876 1.00623 Press, Boca Raton, FL, 2000.
Note: The properties pr (relative pressure) and vr (relative spe-
2650 688.90 556.3 507.25 1.765 1.01172
cific volume) are dimensionless quantities used in the
2700 703.35 601.9 518.26 1.662 1.01712 analysis of isentropic processes, and should not be con-
2750 717.83 650.4 529.31 1.566 1.02244 fused with the properties pressure and specific volume.
2800 732.33 702.0 540.40 1.478 1.02767 h is enthalpy, u is internal energy, and s° is entropy.
2850 746.88 756.7 551.52 1.395 1.03282
2900 761.45 814.8 562.66 1.318 1.03788
2950 776.05 876.4 573.84 1.247 1.04288
3000 790.68 941.4 585.04 1.180 1.04779
3050 805.34 1011 596.28 1.118 1.05264
3100 820.03 1083 607.53 1.060 1.05741
3150 834.75 1161 618.82 1.006 1.06212
3200 849.48 1242 630.12 0.9546 1.06676
3250 864.24 1328 641.46 0.9069 1.07134
3300 879.02 1418 652.81 0.8621 1.07585
3350 893.83 1513 664.20 0.8202 1.08031
3400 908.66 1613 675.60 0.7807 1.08470
3450 923.52 1719 687.04 0.7436 1.08904
3500 938.40 1829 698.48 0.7087 1.09332
3550 953.30 1946 709.95 0.6759 1.09755
3600 968.21 2068 721.44 0.6449 1.10172
3650 983.15 2196 732.95 0.6157 1.10584
3700 998.11 2330 744.48 0.5882 1.10991
3750 1013.1 2471 756.04 0.5621 1.11393
3800 1028.1 2618 767.60 0.5376 1.11791
3850 1043.1 2773 779.19 0.5143 1.12183
3900 1058.1 2934 790.80 0.4923 1.12571
3950 1073.2 3103 802.43 0.4715 1.12955
4000 1088.3 3280 814.06 0.4518 1.13334
4050 1103.4 3464 825.72 0.4331 1.13709
4100 1118.5 3656 837.40 0.4154 1.14079

Ideal Gas Properties of Nitrogen, N2 Ideal Gas Properties of Nitrogen, N2
– –
T (K) h (kJ/kmol) ū (kJ/kmol) s–° (kJ/kmol · K) T (K) h (kJ/kmol) ū (kJ/kmol) s–° (kJ/kmol · K)
0 0 0 0 690 20,297 14,560 216.314
220 6,391 4,562 182.639 700 20,604 14,784 216.756
230 6,683 4,770 183.938 710 20,912 15,008 217.192
240 6,975 4,979 185.180 720 21,220 15,234 217.624
250 7,266 5,188 186.370 730 21,529 15,460 218.059
260 7,558 5,396 187.514 740 21,839 15,686 218.472
270 7,849 5,604 188.614 750 22,149 15,913 218.889
280 8,141 5,813 189.673 760 22,460 16,141 219.301
290 8,432 6,021 190.695 770 22,772 16,370 219.709
298 8,669 6,190 191.502 780 23,085 16,599 220.113
300 8,723 6,229 191.682 790 23,398 16,830 220.512
310 9,014 6,437 192.638 800 23,714 17,061 220.907
320 9,306 6,645 193.562 810 24,027 17,292 221.298
330 9,597 6,853 194.459 820 24,342 17,524 221.684
340 9,888 7,061 195.328 830 24,658 17,757 222.067
350 10,180 7,270 196.173 840 24,974 17,990 222.447
360 10,471 7,478 196.995 850 25,292 18,224 222.822
370 10,763 7,687 197.794 860 25,610 18,459 223.194
380 11,055 7,895 198.572 870 25,928 18,695 223.562
390 11,347 8,104 199.331 880 26,248 18,931 223.927
400 11,640 8,314 200.071 890 26,568 19,168 224.288
410 11,932 8,523 200.794 900 26,890 19,407 224.647
420 12,225 8,733 201.499 910 27,210 19,644 225.002
430 12,518 8,943 202.189 920 27,532 19,883 225.353
440 12,811 9,153 202.863 930 27,854 20,122 225.701
450 13,105 9,363 203.523 940 28,178 20,362 226.047
460 13,399 9,574 204.170 950 28,501 20,603 226.389
470 13,693 9,786 204.803 960 28,826 20,844 226.728
480 13,988 9,997 205.424 970 29,151 21,086 227.064
490 14,285 10,210 206.033 980 29,476 21,328 227.398
500 14,581 10,423 206.630 990 29,803 21,571 227.728
510 14,876 10,635 207.216 1,000 30,129 21,815 228.057
520 15,172 10,848 207.792 1,020 30,784 22,304 228.706
530 15,469 11,062 208.358 1,040 31,442 22,795 229.344
540 15,766 11,277 208.914 1,060 32,101 23,288 229.973
550 16,064 11,492 209.461 1,080 32,762 23,782 230.591
560 16,363 11,707 209.999 1,100 33,426 24,280 231.199
570 16,662 11,923 210.528 1,120 34,092 24,780 231.799
580 16,962 12,139 211.049 1,140 34,760 25,282 232.391
590 17,262 12,356 211.562 1,160 35,430 25,786 232.973
600 17,563 12,574 212.066 1,180 36,104 26,291 233.549
610 17,864 12,792 212.564 1,200 36,777 26,799 234.115
620 18,166 13,011 213.055 1,220 37,452 27,308 234.673
630 18,468 13,230 213.541 1,240 38,129 27,819 235.223
640 18,772 13,450 214.018 1,260 38,807 28,331 235.766
650 19,075 13,671 214.489 1,280 39,488 28,845 236.302
660 19,380 13,892 214.954 1,300 40,170 29,361 236.831
670 19,685 14,114 215.413 1,320 40,853 29,378 237.353
680 19,991 14,337 215.866 1,340 41,539 30,398 237.867

– –
1360 42,227 30,919 238.376 2850 95,859 72,163 264.895
1380 42,915 31,441 238.878 2900 97,705 73,593 265.538
1400 43,605 31,964 239.375 2950 99,556 75,028 266.170
1420 44,295 32,489 239.865 3000 101,407 76,464 266.793
1440 44,988 33,014 240.350 3050 103,260 77,902 267.404
1460 45,682 33,543 240.827 3100 105,115 79,341 268.007
1480 46,377 34,071 241.301 3150 106,972 80,782 268.601
1500 47,073 34,601 241.768 3200 108,830 82,224 269.186
1520 47,771 35,133 242.228 3250 110,690 83,668 269.763
1540 48,470 35,665 242.685 Source: Cengel, Y. and Boles, M., Thermodynamics: An
1560 49,168 36,197 243.137 Engineering Approach, 7th edn., McGraw-Hill,
1580 49,869 36,732 243.585 New York, 2011.
–
1600 50,571 37,268 244.028 h is enthalpy, ū is internal energy, and s̄
° is entropy.
1620 51,275 37,806 244.464
1640 51,980 38,344 244.896 TABLE C.4
1660 52,686 38,884 245.324
Ideal Gas Properties of Nitrogen, N2
1680 53,393 39,424 245.747
(English Units)
1700 54,099 39,965 246.166 –
1720 54,807 40,507 246.580 T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R)
1740 55,516 41,049 246.990 300 2,082.0 1,486.2 41.695
1760 56,227 41,594 247.396 320 2,221.0 1,585.5 42.143
1780 56,938 42,139 247.798 340 2,360.0 1,684.4 42.564
1800 57,651 42,685 248.195 360 2,498.9 1,784.0 42.962
1820 58,363 43,231 248.589 380 2,638.0 1,883.4 43.337
1840 59,075 43,777 248.979 400 2,777.0 1,982.6 43.694
1860 59,790 44,324 249.365 420 2,916.1 2,082.0 44.034
1880 60,504 44,873 249.748 440 3,055.1 2,181.3 44.357
1900 61,220 45,423 250.128 460 3,194.1 2,280.6 44.665
1920 61,936 45,973 250.502 480 3,333.1 2,379.9 44.962
1940 62,654 46,524 250.874 500 3,472.2 2,479.3 45.246
1960 63,381 47,075 251.242 520 3,611.3 2,578.6 45.519
1980 64,090 47,627 251.607 537 3,729.5 2,663.1 45.743
2000 64,810 48,181 251.969 540 3,750.3 2,678.0 45.781
2050 66,612 49,567 252.858 560 3,889.5 2,777.4 46.034
2100 68,417 50,957 253.726 580 4,028.7 2,876.9 46.278
2150 70,226 52,351 254.578 600 4,167.9 2,976.4 46.514
2200 72,040 53,749 255.412 620 4,307.1 3,075.9 46.742
2250 73,856 55,149 256.227 640 4,446.4 3,175.5 46.964
2300 75,676 56,553 257.027 660 4,585.8 3,275.2 47.178
2350 77,496 57,958 257.810 680 4,725.3 3,374.9 47.386
2400 79,320 59,366 258.580 700 4,864.9 3,474.8 47.588
2450 81,149 60,779 259.332 720 5,004.5 3,574.7 47.785
2500 82,981 62,195 260.073 740 5,144.3 3,674.7 47.977
2550 84,814 63,613 260.799 760 5,284.1 3,774.9 48.164
2600 86,650 65,033 261.512 780 5,424.2 3,875.2 48.345
2650 88,488 66,455 262.213 800 5,564.4 3,975.7 48.522
2700 90,328 67,880 262.902 820 5,704.7 4,076.3 48.696
2750 92,171 69,306 263.577 840 5,845.3 4,177.1 48.865
2800 94,014 70,734 264.241 860 5,985.9 4,278.1 49.031
(continued)

– –
T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R) T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R)
880 6,126.9 4,379.4 49.193 1860 13,426.5 9,732.8 54.729
900 6,268.1 4,480.8 49.352 1900 13,742 9,968 54.896
920 6,409.6 4,582.6 49.507 1940 14,058 10,205 55.061
940 6,551.2 4,684.5 49.659 1980 14,375 10,443 55.223
960 6,693.1 4,786.7 49.808 2020 14,694 10,682 55.383
980 6,835.4 4,889.3 49.955 2060 15,013 10,923 55.540
1000 6,977.9 4,992.0 50.099 2100 15,334 11,164 55.694
1020 7,120.7 5,095.1 50.241 2140 15,656 11,406 55.846
1040 7,263.8 5,198.5 50.380 2180 15,978 11,649 55.995
1060 7,407.2 5,302.2 50.516 2220 16,302 11,893 56.141
1080 7,551.0 5,406.2 50.651 2260 16,626 12,138 56.286
1100 7,695.0 5,510.5 50.783 2300 16,951 12,384 56.429
1120 7,839.3 5,615.2 50.912 2340 17,277 12,630 56.570
1140 7,984.0 5,720.1 51.040 2380 17,604 12,878 56.708
1160 8,129.0 5,825.4 51.167 2420 17,392 13,126 56.845
1180 8,274.4 5,931.0 51.291 2460 18,260 13,375 56.980
1200 8,420.0 6,037.0 51.143 2500 18,590 13,625 57.112
1220 8,566.1 6,143.4 51.534 2540 18,919 13,875 57.243
1240 8,712.6 6,250.1 51.653 2580 19,250 14,127 57.372
1260 8,859.3 6,357.2 51.771 2620 19,582 14,379 57.499
1280 9,006.4 6,464.5 51.887 2660 19,914 14,631 57.625
1300 9,153.9 6,572.3 51.001 2700 20,246 14,885 57.750
1320 9,301.8 6,680.4 52.114 2740 20,580 15,139 57.872
1340 9,450.0 6,788.9 52.225 2780 20,914 15,393 57.993
1360 9,598.6 6,897.8 52.335 2820 21,248 15,648 58.113
1380 9,747.5 7,007.0 52.444 2860 21,584 15,905 58.231
1400 9,896.9 7,116.7 52.551 2900 21,920 16,161 58.348
1420 10,046.6 7,226.7 52.658 2940 22,256 16,417 58.463
1440 10,196.6 7,337.0 52.763 2980 22,593 16,675 58.576
1460 10,347.0 7,447.6 52.867 3020 22,930 16,933 58.688
1480 10,497.8 7,558.7 52.969 3060 23,268 17,192 58.800
1500 10,648.0 7,670.1 53.071 3100 23,607 17,451 58.910
1520 10,800.4 7,781.9 53.171 3140 23,946 17,710 59.019
1540 10,952.2 7,893.9 53.271 3180 24,285 17,970 59.126
1560 11,104.3 8,006.4 53.369 3220 24,625 18,231 59.232
1580 11,256.9 8,119.2 53.465 3260 24,965 18,491 59.338
1600 11,409.7 8,232.3 53.561 3300 25,306 18,753 59.442
1620 11,562.8 8,345.7 53.656 3340 25,647 19,014 59.544
1640 11,716.4 8,459.6 53.751 3380 25,989 19,277 59.646
1660 11,870.2 8,573.6 53.844 3420 26,331 19,539 59.747
1680 12,024.3 8,688.1 53.936 3460 26,673 19,802 59.846
1700 12,178.9 8,802.9 54.028 3500 27,016 20,065 59.944
1720 12,333.7 8,918.0 54.118 3540 27,359 20,329. 60.041
1740 12,488.8 9,033.4 54.208 3580 27,703 20,593 60.138
1760 12,644.3 9,149.2 54.297 3620 28,046 20,858 60.234
1780 12,800.2 9,265.3 54.385 3660 28,391 21,122 60.328
1800 12,956.3 9,381.7 54.472 3700 28,735 21,387 60.422
1820 13,112.7 9,498.4 54.559 3740 29,080 21,653 60.515
1840 13,269.5 9,615.5 54.645 3780 29,425 21,919 60.607
TABLE C.4 (continued) TABLE C.5

Ideal Gas Properties of Nitrogen, N2 Ideal Gas Properties of Oxygen, O2
(English Units) (SI Units)
– –
T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R) T (K) h (kJ/kmol) ū (kJ/kmol) s–° (kJ/kmol · K)
3820 29,771 22,185 60.698 0 0 0 0
3860 30,117 22,451 60.788 220 6,404 4,575 196.171
3900 30,463 22,718 60.877 230 6,694 4,782 197.461
3940 30,809 22,985 60.966 240 6,984 4,989 198.696
3980 31,156 23,252 61.053 250 7,275 5,197 199.885
4020 31,503 23,520 61.139 260 7,566 5,405 201.027
4060 31,850 23,788 61.225 270 7,858 5,613 202.128
4100 32,198 24,056 61.310 280 8,150 5,822 203.191
4140 32,546 24,324 61.395 290 8,443 6,032 204.218
4180 32,894 24,593 61.479 298 8,682 6,203 205.033
4220 33,242 24,862 61.562 300 8,736 6,242 205.213
4260 33,591 25,131 61.644 310 9,030 6,453 206.177
4300 33,940 25,401 61.726 320 9,325 6,664 207.112
4340 34,289 25,670 61.806 330 9,620 6,877 208.020
4380 34,638 25,940 61.887 340 9,916 7,090 208.904
4420 34,988 26,210 61.966 350 10,213 7,303 209.765
4460 35,338 26,481 62.045 360 10,511 7,518 210.604
4500 35,688 26,751 62.123 370 10,809 7,733 211.423
4540 36,038 27,022 62.201 380 11,109 7,949 212.222
4580 36,389 27,293 62.278 390 11,409 8,166 213.002
4620 36,739 27,565 62.354 400 11,711 8,384 213.765
4660 37,090 27,836 62.429 410 12,012 8,603 214.510
4700 37,441 28,108 62.504 420 12,314 8,822 215.241
4740 37,792 28,379 62.578 430 12,618 9,043 215.955
4780 38,144 28,651 62.652 440 12,923 9,264 216.656
4820 38,495 28,924 62.725 450 13,228 9,487 217.342
4860 38,847 29,196 62.798 460 13,525 9,710 218.016
4900 39,199 29,468 62.870 470 13,842 9,935 218.676
5000 40,080 30,151 63.049 480 14,151 10,160 219.326
5100 40,962 30,834 63.223 490 14,460 10,386 219.963
5200 41,844 31,518 63.395 500 14,770 10,614 220.589
5300 42,728 32,203 63.563 510 15,082 10,842 221.206
520 15,395 11,071 221.812
Source: Cengel, Y. and Boles, M., Thermodynamics: An Engineering
Approach, 7th edn., McGraw-Hill, New York, 2011. 530 15,708 11,301 222.409
–
h is enthalpy, ū is internal energy, and s̄
° is entropy. 540 16,022 11,533 222.997
550 16,338 11,765 223.576
560 16,654 11,998 224.146
570 16,971 12,232 224.708
580 17,290 12,467 225.262
590 17,609 12,703 225.808
600 17,929 12,940 226.346
610 18,250 13,178 226.877
620 18,572 13,417 227.400
630 18,895 13,657 227.918
640 19,219 13,898 228.429
650 19,544 14,140 228.932
660 19,870 14,383 229.430
670 20,197 14,626 229.920
680 20,524 14,871 230.405
(continued)

Ideal Gas Properties of Oxygen, O2 Ideal Gas Properties of Oxygen, O2
– –
690 20,854 15,116 230.885 1360 44,198 32,891 254.404
700 21,184 15,364 231.358 1380 44,923 33,449 254.932
710 21,514 15,611 231.827 1400 45,648 34,008 255.454
720 21,845 15,859 232.291 1420 46,374 34,567 255.968
730 22,177 16,107 232.748 1440 47,102 35,129 256.475
740 22,510 16,357 233.201 1460 47,831 35,692 256.978
750 22,844 16,607 233.649 1480 48,561 36,256 257.474
760 23,178 16,859 234.091 1500 49,292 36,821 257.965
770 23,513 17,111 234.528 1520 50,024 37,387 258.450
780 23,850 17,364 234.960 1540 50,756 37,952 258.928
790 24,186 17,618 235.387 1560 51,490 38,520 259.402
800 24,523 17,872 235.810 1580 52,224 39,088 259.870
810 24,861 18,126 236.230 1600 52,961 39,658 260.333
820 25,199 18,382 236.644 1620 53,696 40,227 260.791
830 25,537 18,637 237.055 1640 54,434 40,799 261.242
840 25,877 18,893 237.462 1660 55,172 41,370 261.690
850 26,218 19,150 237.864 1680 55,912 41,944 262.132
860 26,559 19,408 238.264 1700 56,652 42,517 262.571
870 26,899 19,666 238.660 1720 57,394 43,093 263.005
880 27,242 19,925 239.051 1740 58,136 43,669 263.435
890 27,584 20,185 239.439 1760 58,880 44,247 263.861
900 27,928 20,445 239.823 1780 59,624 44,825 264.283
910 28,272 20,706 240.203 1800 60,371 45,405 264.701
920 28,616 20,967 240.580 1820 61,118 45,986 265.113
930 28,960 21,228 240.953 1840 61,866 46,568 265.521
940 29,306 21,491 241.323 1860 62,616 47,151 265.925
950 29,652 21,754 241.689 1880 63,365 47,734 266.326
960 29,999 22,017 242.052 1900 64,116 48,319 266.722
970 30,345. 22,280 242.411 1920 64,868 48,904 267.115
980 30,692 22,544 242.768 1940 65,620 49,490 267.505
990 31,041 22,809 242.120 1960 66,374 50,078 267.891
1000 31,389 23,075 243.471 1980 67,127 50,665 268.275
1020 32,088 23,607 244.164 2000 67,881 51,253 268.655
1040 32,789 24,142 244.844 2050 69,772 52,727 269.588
1060 33,490 24,677 245.513 2100 71,668 54,208 270.504
1080 34,194 25,214 246.171 2150 73,573 55,697 271.399
1100 34,899 25,753 246.818 2200 75,484 57,192 272.278
1120 35,606 26,294 247.454 2250 77,397 58,690 273.136
1140 36,314 26,836 248.081 2300 79,316 60,193 273.891
1160 37,023 27,379 248.698 2350 81,243 61,704 274.809
1180 37,734 27,923 249.307 2400 83,174 63,219 275.625
1200 38,447 28,469 249.906 2450 85,112 64,742 276.424
1220 39,162 29,018 250.497 2500 87,057 66,271 277.207
1240 39,877 29,568 251.079 2550 89,004 67,802 277.979
1260 40,594 30,118 251.653 2600 90,956 69,339 278.738
1280 41,312 30,670 252.219 2650 92,916 70,883 279.485
1300 42,033 31,224 252.776 2700 94,881 72,433 280.219
1320 42,753 31,778 253.325 2750 96,852 73,987 280.942
1340 43,475 32,334 253.868 2800 98,826 75,546 281.654

(SI Units) (English Units)
– –
T (K) h (kJ/kmol) ū (kJ/kmol) s–° (kJ/kmol · K) T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R)
2850 100,808 77,112 282.357 920 6,486.7 4,659.7 52.852
2900 102,793 78,682 283.048 940 6,636.1 4,769.4 53.012
2950 104,785 80,258 283.728 960 6,786.0 4,879.5 53.170
3000 106,780 81,837 284.399 980 6,936.4 4,990.3 53.326
3050 108,778 83,419 285.060 1000 7,087.5 5,101.6 53.477
3100 110,784 85,009 285.713 1020 7,238.9 5,213.3 53.628
3150 112,795 86,601 286.355 1040 7,391.0 5,325.7 53.775
3200 114,809 88,203 286.989 1060 7,543.6 5,438.6 53.921
3250 116,827 89,804 287.614 1080 7,696.8 5,552.1 54.064
1100 7,850.4 5,665.9 54.204
Source: Cengel, Y. and Boles, M., Thermodynamics: An
Engineering Approach, 7th edn., McGraw-Hill, 1120 8,004.5 5,780.3 54.343
New York, 2011. 1140 8,159.1 5,895.2 54.480
–
° is entropy. 1160 8,314.2 6,010.6 54.614
TABLE C.6 1180 8,469.8 6,126.5 54.748
1200 8,625.8 6,242.8 54.879
Ideal Gas Properties of Oxygen, O2
1220 8,782.4 6,359.6 55.008
(English Units) 1240 8,939.4 6,476.9 55.136
–
T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R) 1260 9,096.7 6,594.5 55.262
300 2,073.5 1,477.8 44.927 1280 9,254.6 6,712.7 55.386
320 2,212.6 1,577.1 45.375 1300 9,412.9 6,831.3 55.508
340 2,351.7 1,676.5 45.797 1320 9,571.9 6,950.2 55.630
360 2,490.8 1,775.9 46.195 1340 9,730.7 7,069.6 55.750
380 2,630.0 1,875.3 46.571 1360 9,890.2 7,189.4 55.867
400 2,769.1 1,974.8 46.927 1380 10,050.1 7,309.6 55.984
420 2,908.3 2,074.3 47.267 1400 10,210.4 7,430.1 56.099
440 3,047.5 2,173.8 47.591 1420 10,371.0 7,551.1 56.213
460 3,186.9 2,273.4 47.900 1440 10,532.0 7,672.4 56.326
480 3,326.5 2,373.3 48.198 1460 10,693.3 7,793.9 56.437
500 3,466.2 2,473.2 48.483 1480 10,855.1 7,916.0 56.547
520 3,606.1 2,573.4 48.757 1500 11,017.1 8,038.3 56.656
1520 11,179.6 8,161.1 56.763
537 3,725.1 2,658.7 48.982
1540 11,342.4 8,284.2 56.869
540 3,746.2 2,673.8 49.021
1560 11,505.4 8,407.4 56.975
560 3,886.6 2,774.5 49.276
1580 11,668.8 8,531.1 57.079
580 4,027.3 2,875.5 49.522
1600 11,832.5 8,655.1 57.182
600 4,168.3 2,976.8 49.762
1620 11,996.6 8,779.5 57.284
620 4,309.7 3,078.4 49.993
1640 12,160.9 8,904.1 57.385
640 4,451.4 3,180.4 50.218
1660 12,325.5 9,029.0 57.484
660 4,593.5 3,282.9 50.437
1680 12,490.4 9,154.1 57.582
680 4,736.2 3,385.8 50.650
1700 12,655.6 9,279.6 57.680
700 4,879.3 3,489.2 50.858
1720 12,821.1 9,405.4 57.777
720 5,022.9 3,593.1 51.059
1740 12,986.9 9,531.5 57.873
740 5,167.0 3,697.4 51.257
1760 13,153.0 9,657.9 57.968
760 5,311.4 3,802.4 51.450
1780 13,319.2 9,784.4 58.062
780 5,456.4 3,907.5 51.638
1800 13,485.8 9,911.2 58.155
800 5,602.0 4,013.3 51.821
1820 13,652.5 10,038.2 58.247
820 5,748.1 4,119.7 52.002
1840 13,819.6 10,165.6 58.339
840 5,894.8 4,226.6 52.179
1860 13,986.8 10,293.1 58.428
860 6,041.9 4,334.1 52.352
1900 14,322 10,549 58.607
880 6,189.6 4,442.0 52.522
1940 14,658 10,806 58.782
900 6,337.9 4,550.6 52.688
(continued)

– –
1980 14,995 11,063 58.954 3980 32,623 24,720 65.078
2020 15,333 11,321 59.123 4020 32,989 25,006 65.169
2060 15,672 11,581 59.289 4060 33,355 25,292 65.260
2100 16,011 11,841 59.451 4100 33,722 25,580 65.350
2140 16,351 12,101 59.612 4140 34,089 25,867 64.439
2180 16,692 12,363 59.770 4180 34,456 26,155 65.527
2220 17,036 12,625 59.926 4220 34,824 26,144 65.615
2260 17,376 12,888 60.077 4260 35,192 26,733 65.702
2300 17,719 13,151 60.228 4300 35,561 27,022 65.788
2340 18,062 13,416 60.376 4340 35,930 27,312 65.873
2380 18,407 13,680 60.522 4380 36,300 27,602 65.958
2420 18,572 13,946 60.666 4420 36,670 27,823 66.042
2460 19,097 14,212 60.808 4460 37,041 28,184 66.125
2500 19,443 14,479 60.946 4500 37,412 28,475 66.208
2540 19,790 14,746 61.084 4540 37,783 28,768 66.290
2580 20,138 15,014 61.220 4580 38,155 29,060 66.372
2620 20,485 15,282 61.354 4620 38,528 29,353 66.453
2660 20,834 15,551 61.486 4660 38,900 29,646 66.533
2700 21,183 15,821 61.616 4700 39,274 29,940 66.613
2740 21,533 16,091 61.744 4740 39,647 30,234 66.691
2780 21,883 16,362 61.871 4780 40,021 30,529 66.770
2820 22,232 16,633 61.996 4820 40,396 30,824 66.848
2860 22,584 16,905 62.120 4860 40,771 31,120 66.925
2900 22,936 17,177 62.242 4900 41,146 31,415 67.003
2940 23,288 17,450 62.363 5000 42,086 32,157 67.193
2980 23,641 17,723 62.483 5100 43,021 32,901 67.380
3020 23,994 17,997 62.599 5200 43,974 33,648 67.562
3060 24,348 18,271 62.716 5300 44,922 34,397 67.743
3100 24,703 18,546 62.831 Source: Cengel, Y. and Boles, M., Thermodynamics: An Engineering
3140 25,057 18,822 62.945 Approach, 7th edn., McGraw-Hill, New York, 2011.
–
° is entropy.
3220 25,769 19,374 63.169
3260 26,175 19,651 63.279
3300 26,412 19,928 63.386
3340 26,839 20,206 63.494
3380 27,197 20,485 63.601
3420 27,555 20,763 63.706
3460 27,914 21,043 63.811
3500 28,273 21,323 63.914
3540 28,633 21,603 64.016
3580 28,994 21,884 64.114
3620 29,354 22,165 64.217
3660 29,716 22,447 64.316
3700 30,078 22,730 64.415
3740 30,440 23,013 64.512
3780 30,803 23,296 64.609
3820 31,166 23,580 64.704
3860 31,529 23,864 64.800
3900 31,894 24,149 64.893
3940 32,258 24,434 64.986

Ideal Gas Properties of CO2 Ideal Gas Properties of CO2
– –
0 0 0 0 690 26,631 20,894 249.952
220 6,601 4,772 202.966 700 27,125 21,305 250.663
230 6,938 5,026 204.464 710 27,622 21,719 251.368
240 7,280 5,285 205.920 720 28,121 22,134 252.065
250 7,627 5,548 207.337 730 28,622 22,522 252.755
260 7,979 5,817 208.717 740 29,124 22,972 253.439
270 8,335 6,091 210.062 750 29,629 23,393 254.117
280 8,697 6,369 211.376 760 30,135 23,817 254.787
290 9,063 6,651 212.660 770 30,644 24,242 255.452
298 9,364 6,885 213.685 780 31,154 24,669 256.110
300 9,431 6,939 213.915 790 31,665 25,097 256.762
310 9,807 7,230 215.146 800 32,179 25,527 257.408
320 10,186 7,526 216.351 810 32,694 25,959 258.048
330 10,570 7,826 217.534 820 33,212 26,394 258.682
340 10,959 8,131 218.694 830 33,730 26,829 259.311
350 11,351 8,439 219.831 840 34,251 27,267 259.934
360 11,748 8,752 220.948 850 34,773 27,706 260.551
370 12,148 9,068 222.044 860 35,296 28,125 261.164
380 12,552 9,392 223.122 870 35,821 28,588 261.770
390 12,960 9,718 224.182 880 36,347 29,031 262.371
400 13,372 10,046 225.225 890 36,876 29,476 262.968
410 13,787 10,378 226.250 900 37,405 29,922 263.559
420 14,206 10,714 227.258 910 37,935 30,369 264.146
430 14,628 11,053 228.252 920 38,467 30,818 264.728
440 15,054 11,393 229.230 930 39,000 31,268 265.304
450 15,483 11,742 230.194 940 39,535 31,719 265.877
460 15,916 12,091 231.144 950 40,070 32,171 266.444
470 16,351 12,444 232.080 960 40,607 32,625 267.007
480 16,791 12,800 233.004 970 41,145 33,081 267.566
490 17,232 13,158 233.916 980 41,685 33,537 268.119
500 17,678 13,521 234.814 990 42,226 33,995 268.670
510 18,126 13,885 235.700 1000 42,769 34,455 269.215
520 18,576 14,253 236.575 1020 43,859 35,378 270.293
530 19,029 14,622 237.439 1040 44,953 36,306 271.354
540 19,485 14,996 238.292 1060 46,051 37,238 272.400
550 19,945 15,372 239.135 1080 47,153 38,174 273.430
560 20,407 15,751 239.962 1100 48,258 39,112 274.445
570 20,870 16,131 240.789 1120 49,369 40,057 275.444
580 21,337 16,515 241.602 1140 50,484 41,006 276.430
590 21,807 16,902 242.405 1160 51,602 41,957 277.403
600 22,280 17,291 243.199 1180 52,724 42,913 278.361
610 22,754 17,683 243.983 1200 53,848 43,871 297.307
620 23,231 18,076 244.758 1220 54,977 44,834 280.238
630 23,709 18,471 245.524 1240 56,108 45,799 281.158
640 24,190 18,869 246.282 1260 57,244 46,768 282.066
650 24,674 19,270 247.032 1280 58,381 47,739 282.962
660 25,160 19,672 247.773 1300 59,522 48,713 283.847
670 25,648 20,078 248.507 1320 60,666 49,691 284.722
680 26,138 20,484 249.233 1340 61,813 50,672 285.586
(continued)

– –
1360 62,963 51,656 286.439 2850 152,908 129,212 330.896
1380 64,116 52,643 287.283 2900 156,009 131,898 331.975
1400 65,271 53,631 288.106 2950 159,117 134,589 333.037
1420 66,427 54,621 288.934 3000 162,226 137,283 334.084
1440 67,586 55,614 289.743 3050 165,341 139,982 335.114
1460 68,748 56,609 290.542 3100 168,456 142,681 336.126
1480 66,911 57,606 291.333 3150 171,576 145,385 337.124
1500 71,078 58,606 292.114 3200 174,695 148,089 338.109
1520 72,246 59,609 292.888 3250 177,822 150,801 339.069
1540 73,417 60,613 292.654 Source: Cengel, Y. and Boles, M., Thermodynamics: An
1560 74,590 61,620 294.411 Engineering Approach, 7th edn., McGraw-Hill,
1580 76,767 62,630 295.161 New York, 2011.
–
° is entropy.
1620 78,123 64,653 296.632
1640 79,303 65,668 297.356
TABLE C.8
1660 80,486 66,592 298.072 Ideal Gas Properties of CO2
1680 81,670 67,702 298.781 (English Units)
1700 82,856 68,721 299.482 –
T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R)
1720 84,043 69,742 300.177
300 2,108.2 1,512.4 46.353
1740 85,231 70,764 300.863
320 2,256.6 1,621.1 46.832
1760 86,420 71,787 301.543
340 2,407.3 1,732.1 47.289
1780 87,612 72,812 302.217
360 2,560.5 1,845.6 47.728
1800 88,806 73,840 302.884
380 2,716.4 1,961.8 48.148
1820 90,000 74,868 303.544
400 2,874.7 2,080.4 48.555
1840 91,196 75,897 304.198
420 3,035.7 2,201.7 48.947
1860 92,394 76,929 304.845
440 3,199.4 2,325.6 49.329
1880 93,593 77,962 305.487
460 3,365.7 2,452.2 49.698
1900 94,793 78,996 306.122
480 3,534.7 2,581.5 50.058
1920 95,995 80,031 306.751
500 3,706.2 2,713.3 50.408
1940 97,197 81,067 307.374
520 3,880.3 2,847.7 50.750
1960 98,401 82,105 307.992
537 4,027.5 2,963.8 51.032
1980 99,606 83,144 308.604
540 4,056.8 2,984.4 51.082
2000 100,804 84,185 309.210
560 4,235.8 3,123.7 51.408
2050 103,835 86,791 310.701
580 4,417.2 3,265.4 51.726
2100 106,864 89,404 312.160
600 4,600.9 3,409.4 52.038
2150 109,898 92,023 313.589
620 4,786.6 3,555.6 52.343
2200 112,939 94,648 314.988
640 4,974.9 3,704.0 52.641
2250 115,984 97,277 316.356
660 5,165.2 3,854.6 52.934
2300 119,035 99,912 317.695
680 5,357.6 4,007.2 53.225
2350 122,091 102,552 319.011
700 5,552.0 4,161.9 53.503
2400 125,152 105,197 320.302
720 5,748.4 4,318.6 53.780
2450 128,219 107,849 321.566
740 5,946.8 4,477.3 54.051
2500 131,290 110,504 322.808
760 6,147.0 4,637.9 54.319
2550 134,368 113,166 324.026
780 6,349.1 4,800.1 54.582
2600 137,449 115,832 325.222
800 6,552.9 4,964.2 54.839
2650 140,533 118,500 326.396
820 6,758.3 5,129.9 55.093
2700 143,620 121,172 327.549
840 6,965.7 5,297.6 55.343
2750 146,713 123,849 328.684
860 7,174.7 5,466.9 55.589
2800 149,808 126,528 329.800
880 7,385.3 5,637.7 55.831
900 7,597.6 5,810.3 56.070

– –
920 7,811.4 5,984.4 56.305 1980 20,753 16,821 65.543
940 8,026.8 6,160.1 56.536 2020 21,284 17,273 65.809
960 8,243.8 6,337.4 56.765 2060 21,818 17,727 66.069
980 8,462.2 6,516.1 56.990 2100 22,353 18,182 66.327
1000 8,682.1 6,696.2 57.212 2140 22,890 18,640 66.581
1020 8,903.4 6,877.8 57.432 2180 23,429 19,101 66.830
1040 9,126.2 7,060.9 57.647 2220 23,970 19,561 67.076
1060 9,350.3 7,245.3 57.861 2260 24,512 20,024 67.319
1080 9,575.8 7,431.1 58.072 2300 25,056 20,489 67.557
1100 9,802.6 7,618.1 58.281 2340 25,602 20,955 67.792
1120 10,030.6 7,806.4 58.485 2380 26,150 21,423 68.025
1140 10,260.1 7,996.2 58.689 2420 26,699 21,893 68.253
1160 10,490.6 8,187.0 58.889 2460 27,249 22,364 68.479
1180 10,722.3 8,379.0 59.088 2500 27,801 22,837 68.702
1200 10,955.3 8,572.3 59.283 2540 28,355 23,310 68.921
1220 11,189.4 8,766.6 59.477 2580 28,910 23,786 69.138
1240 11,424.6 8,962.1 59.668 2620 29,465 24,262 69.352
1260 11,661.0 9,158.8 59.858 2660 30,023 24,740 69.563
1280 11,898.4 9,356.5 60.044 2700 30,581 25,220 69.771
1300 12,136.9 9,555.3 60.229 2740 31,141 25,701 69.977
1320 12,376.4 9,755.0 60.412 2780 31,702 26,181 70.181
1340 12,617.0 9,955.9 60.593 2820 32,264 26,664 70.382
1360 12,858.5 10,157.7 60.772 2860 32,827 27,148 70.580
1380 13,101.0 10,360.5 60.949 2900 33,392 27,633 70.776
1400 13,344.7 10,564.5 61.124 2940 33,957 28,118 70.970
1420 13,589.1 10,769.2 61.298 2980 34,523 28,605 71.160
1440 13,834.5 10,974.8 61.469 3020 35,090 29,093 71.350
1460 14,080.8 11,181.4 61.639 3060 35,659 29,582 71.537
1480 14,328.0 11,388.9 61.800 3100 36,228 30,072 71.722
1500 14,576.0 11,597.2 61.974 3140 36,798 30,562 71.904
1520 14,824.9 11,806.4 62.138 3180 37,369 31,054 72.085
1540 15,074.7 12,016.5 62.302 3220 37,941 31,546 72.264
1560 15,325.3 12,227.3 62.464 3260 38,513 32,039 72.441
1580 15,576.7 12,439.0 62.624 3300 39,087 32,533 72.616
1600 15,829.0 12,651.6 62.783 3340 39,661 33,028 72.788
1620 16,081.9 12,864.8 62.939 3380 40,236 33,524 72.960
1640 16,335.7 13,078.9 63.095 3420 40,812 34,020 73.129
1660 16,590.2 13,293.7 63.250 3460 41,388 34,517 73.297
1680 16,845.5 13,509.2 63.403 3500 41,965 35,015 73.462
1700 17,101.4 13,725.4 63.555 3540 42,543 35,513 73.627
1720 17,358.1 13,942.4 63.704 3580 43,121 36,012 73.789
1740 17,615.5 14,160.1 63.853 3620 43,701 36,512 73.951
1760 17,873.5 14,378.4 64.001 3660 44,280 37,012 74.110
1780 18,132.2 14,597.4 64.147 3700 44,861 37,513 74.267
1800 18,391.5 14,816.9 64.292 3740 45,442 38,014 74.423
1820 18,651.5 15,037.2 64.435 3780 46,023 38,517 74.578
1840 18,912.2 15,258.2 64.578 3820 46,605 39,019 74.732
1860 19,173.4 15,479.7 64.719 3860 47,188 39,522 74.884
1900 19,698 15,925 64.999 3900 47,771 40,026 75.033
1940 20,224 16,372 65.272 3940 48,355 40,531 75.182
(continued)
TABLE C.8 (continued)

Ideal Gas Properties of CO2
(English Units)
–
T (°R) h (Btu/lbmol) ū (Btu/lbmol) s–° (Btu/lbmol · °R)
3980 48,939 41,035 75.330
4020 49,524 41,541 75.477
4060 50,109 42,047 75.622
4100 50,695 42,553 75.765
4140 51,282 43,060 75.907
4180 51,868 43,568 76.048
4220 52,456 44,075 76.188
4260 53,044 44,584 76.327
4300 53,632 45,093 76.464
4340 54,221 45,602 76.601
4380 54,810 46,112 76.736
4420 55,400 46,622 76.870
4460 55,990 47,133 77.003
4500 56,581 47,645 77.135
4540 57,172 48,156 77.266
4580 57,764 48,668 77.395
4620 58,356 49,181 77.581
4660 58,948 49,694 77.652
4700 59,541 50,208 77.779
4740 60,134 50,721 77.905
4780 60,728 51,236 78.029
4820 61,322 51,750 78.153
4860 61,916 52,265 78.276
4900 62,511 52,781 78.398
5000 64,000 54,071 78.698
5100 65,491 55,363 78.994
5200 66,984 56,658 79.284
5300 68,471 57,954 79.569
Source: Cengel, Y. and Boles, M., Thermodynamics: An Engineering
Approach, 7th edn., McGraw-Hill, New York, 2011.
–
° is entropy.
TABLE C.9
Properties of Gases at Atmospheric Pressure (101.3 kPa = 14.7 psia): Air (Gas Constant = 286.8 J/(kg K) = 53.3 ft lbf/lbm °R; γ = cp/cv = 1.4)
Temp, T Density, p Specific Heat, cp Kinematic Viscosity, v Thermal Conductivity, k Thermal Diffusivity, α
K °R kg/m3 lbm/ft3 J/kg · K Btu/lbm · °R m2/s ft2/s W/m · K Btu/h · ft · °R m2/s ft2/h Prandtl Number, Pr
100 180 3.601 0.225 1026.6 0.245 1.923 × 10−6 2.070 × 10−5 0.009246 0.005342 0.02501 × 10−6 0.0869 0.770
150 270 2.368 0.148 1009.9 0.241 4.343 4.674 0.013735 0.007936 0.05745 0.223 0.753
200 360 1.768 0.110 1006.1 0.240 7.490 8.062 0.01809 0.01045 0.10165 0.394 0.739
250 450 1.413 0.0882 1005.3 0.240 9.49 10.2 0.02227 0.02287 0.13161 0.510 0.722
300 540 1.177 0.0735 1005.7 0.240 15.68 16.88 0.02624 0.01516 0.22160 0.859 0.708
350 630 0.998 0.0623 1009.0 0.241 20.76 22.35 0.03003 0.01735 0.2983 1.156 0.697
400 720 0.883 0.0551 1014.0 0.242 25.90 27.88 0.03365 0.01944 0.3760 1.457 0.689
450 810 0.783 0.489 1020.7 0.244 28.86 31.06 0.037.7 0.02142 0.4222 1.636 0.683
500 900 0.705 0.0440 1029.5 0.245 37.90 40.80 0.04038 0.02333 0.5564 2.356 0.680
550 990 0.642 0.0401 1039.2 0.248 44.34 47.73 0.04360 0.02519 0.6532 2.531 0.680
600 1000 0.589 0.0367 1055.1 0.252 51.34 55.26 0.04659 0.02682 0.7512 2.911 0.680
650 1170 0.543 0.0339 1063.5 0.254 58.51 62.98 0.00953 0.02862 0.8578 3.324 0.682
700 1260 0.503 0.0314 1075.2 0.257 66.25 7131 0.05230 0.030023 0.9672 3.748 0.684
750 1350 0.471 0.0594 1085.6 0.259 73.91 79.56 0.05509 0.03183 1.0774 4.175 0.686
800 1440 0.441 0.0275 1097.8 0.262 8229 88.58 0.05779 0.03339 1.1951 4.631 0.689
850 1530 0.415 0.0259 1109.5 0.265 90.75 97.68 0.06028 0.03483 1.3097 5.075 0.692
900 1620 0.393 0.0245 1121.2 0.268 99.3 107 0.06279 0.03628 1.4278 5.530 0.696
950 1710 0.372 0.0232 1132.1 0.270 108.2 116.5 0.06525 0.03770 1.5510 6.010 0.699
1000 1800 0.352 0.0220 1141.7 0.273 117.8 126.8 0.06752 0.03901 1.6779 6502 0.702
1100 1980 0.320 0.0120 1160 0.277 138.6 149.2 0.0732 0.0423 1.969 7.630 0.704
1200 2160 0.295 0.0184 1179 0.282 159.1 171.3 0.0782 0.0423 1.969 7.630 0.707
1300 2340 0.271 0.0189 1197 0.286 182.1 196.0 0.0837 0.0434 2.583 10.01 0.705
1400 2520 0.252 0.0157 1214 0.290 205.5 221.2 0.0891 0.0515 2.920 11.32 0.705
1500 2700 0.236 0.0147 1230 0.294 229.1 246.6 0.0946 0.0547 3.262 1264 0.705
1600 2880 0.221 0.0138 1248 0.298 254.5 273.9 0.100 0.0578 3.609 13.98 0.705
1700 36060 0.208 0.0130 1267 0.303 280.5 301.9 0.105 0.0607 3.977 15.41 0.705
1800 3240 0.197 0.0123 1287 0.307 308.1 331.6 0.111 0.0641 4.379 16.97 0.704
1900 3420 0.186 0.0115 1309 0.383 338.5 364.4 0.117 0.0676 4.811 18.64 0.704
2000 3600 0.176 0.0110 1338 0.320 369.0 397.2 0.124 0.0716 5.260 20.38 0.702
2100 3780 0.168 0.0105 1372 0.328 399.6 430.1 0.131 0.0757 5.715 22.15 0.700
2200 3960 0.160 0.0100 1419 0.339 432.6 465.6 0.139 0.0803 6120 2372 0.707
2300 4140 0.154 0.00955 1482 0.354 464.0 499.4 0.149 0.0861 6.540 25.34 0.710
2400 4320 0.146 0.00905 1574 0376 504.0 542.5 0.161 0.0930 7.020 27.20 0.718
2500 4500 0.139 0.00868 1688 0.403 543.5 585.0 0.175 0.101 7.441 28.83 0.730
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
479
480
TABLE C.10
Properties of Gases at Atmospheric Pressure (101.3 kPa = 14.7 psia): Nitrogen (Gas Constant = 296.8 J/(kg K) = 55.16 ft lbf/lbm °R; γ = cp/cv = 1.40)
100 180 3.4808 0.2173 1072.2 0.2561 1.971 × 10−6 2.122 × 10−5 0.009450 0.005460 0.025319 × 10−4 0.09811 0.786
200 360 1.7108 0.1068 1042.9 0.2491 7.568 8.146 0.01824 0.01054 0.10224 0.3962 0.747
300 540 1.1421 0.0713 1040.8 0.2486 15.63 16.82 0.02620 0.01514 0.22044 0.8542 0.713
400 720 0.8538 0.0533 1045.9 0.2498 25.74 27.71 0.03335 0.01927 0.3734 1.447 0.691
500 900 0.6824 0.0426 1055.5 0.2521 37.66 40.54 0.03984 0.02302 0.5530 2.143 0.684
600 1080 0.5687 0.0355 1075.6 0.2569 51.19 55.10 0.4580 0.02646 0.7486 2.901 0.686
700 1260 0.4934 0.0308 1096.9 0.2620 65.13 7010 0.05123 0.02960 0.9466 3.668 0.691
800 1440 0.4277 0.0267 1122.5 0.2681 81.46 87.68 0.05609 0.03241 1.1685 4.528 0.700
900 1620 0.3796 0.0237 1146.4 0.2738 91.06 98.02 0.06070 0.03507 1.3946 5.404 0.711
1000 1800 0.3412 0.0213 1167.7 0.2789 117.2 126.2 0.06475 0.03741 1.6250 6.297 0.724
1100 1980 0.3108 0.0194 1185.7 0.2382 136.0 146.4 0.06850 0.03958 1.8591 7.204 0.736
1200 2160 0.2851 0.0178 1203.7 0.2875 156.1 168.0 0.07184 0.04151 2.0932 8.111 0.748
TABLE C.11
Properties of Gases at Atmospheric Pressure (101.3 kPa = 14.7 psia): Oxygen (Gas Constant = 260 J/(kg K) = 48.3 ft lbf/lbm °R; γ = cp/cv = 1.40)
100 180 3.9118 0.2492 947.9 0.2264 1.946 × 10−6 2.095 × 10−5 0.00903 0.00522 0.023876 × 10−4 0.09252 0.815
150 270 26190 0.1635 917.8 0.2192 4.387 4.722 0.01367 0.00790 0.05688 0.2204 0.773
200 360 1.9559 0.1221 913.1 0.2181 7.593 8.173 0.01824 0.01054 0.10214 0.3958 0.745
250 450 1.5618 0.0975 915.7 0.2187 11.45 12.32 0.02259 0.01305 0.15794 0.6120 0.725
300 540 1.3007 0.0812 920.3 0.2198 15.86 17.07 0.02676 0.01546 0.22353 0.8662 0.709
350 630 1.1133 0.0695 929.1 0.2219 20.80 22.39 0.03070 0.01774 0.2968 1.150 0.702
400 720 0.9755 0.0609 942.0 0.2250 26.18 2818 0.03461 0.02000 0.3768 1.460 0.695
450 810 0.8682 0.0542 956.7 0.2285 31.99 34.43 0.03828 0.02212 0.4609 1.786 0.694
500 900 0.7801 0.0487 972.2 0.2322 38.37 41.27 0.04173 0.02411 0.5502 2.132 0.697
550 990 0.7096 0.0443 988.1 0.2360 45.05 48.49 0.04517 0.02610 06441 2.496 0.700
600 1080 0.6508 0.0406 1004.4 0.2399 52.15 56.13 0.04882 0.02792 0.7399 2.867 0.704
TABLE C.12
Properties of Gases at Atmospheric Pressure (101.3 kPa = 14.7 psia): Carbon Dioxide (Gas Constant = 188.9 J/(kg K) = 35.11 ft lbf/lbm °R; γ = cp/cv = 1.30)
K °R kg/m 3 lbm/ft3 J/kg · K Btu/lbm · °R 2
m /s 2
ft /s W/m · K Btu/h · ft · °R m2/s ft2/h Prandtl Number, Pr
220 396 2.4733 0.1544 783 0.187 4.490 × 10 −6 4.833 × 10 −5 0.010805 0.006243 0.05920 × 10 −4 0.2294 0.818
250 450 2.1657 0.1352 804 0.192 5.813 6.257 0.012884 0.007444 0.2868 0.793
300 540 1.7973 0.1122 871 0.208 8.321 8.957 0.016572 0.009575 0.10588 0.4103 0.770
350 630 1.5362 0.0959 900 0.215 11.19 12.05 0.02047 0.01183 0.14808 0.5738 0.755
400 720 1.3424 0.0838 942 0.225 14.39 15.49 0.02461 0.01422 0.19463 0.7542 0.738
450 810 1.1918 0.0744 980 0.234 17.90 19.27 0.02897 0.01674 0.24813 0.9615 0.721
500 900 1.0732 0.0670 1013 0.242 21.67 23.33 0.03352 0.01937 0.3084 1.195 0.702
550 990 0.9739 0.0608 1047 0.250 25.74 27.71 0.03821 0.02208 0.3750 1.453 0.685
600 1080 0.8938 0.0558 1076 0.257 30.02 32.31 0.04313 0.02491 0.4483 1.737 0.668
TABLE C.13
Properties of Gases at Atmospheric Pressure (101.3 kPa = 14.7 psia): Water Vapor or Steam (Gas Constant = 461.5 J/(kg K) = 85.78 ft lbf/lbm °R; γ = cp/cv = 1.33)
K °R kg/m 3 lbm/ft 3 J/kg · K Btu/lbm · °R m /s 2 ft /s2 W/m · K Btu/h · ft · °R m2/s ft2/h Prandtl Number, Pr
380 684 0.5863 0.0366 2060 0.492 2.16 × 10 −6 2.33 × 10 −5 0.0246 0.0142 0.2036 × 10 −4 0.789 1.060
400 720 0.5542 0.0346 2014 0.481 2.42 2.61 0.0261 0.0151 0.2338 0.906 1.040
450 810 0.4902 0.0306 1980 0.473 3.11 3.35 0.0299 0.0173 0.307 1.19 1.010
500 900 0.4005 0.0275 1985 0.474 3.86 4.16 0.0339 0.0196 0.387 1.50 0.996
550 990 0.4005 0.0250 1997 0.477 4.70 5.06 0.0379 0.0219 0.475 1.84 0.991
600 1080 0.3652 0.0228 2026 0.484 5.66 6.09 0.0422 0.0244 0.573 2.22 0.986
650 1170 0.3380 0.0211 2056 0.491 6.64 7.15 0.0464 0.0268 0.666 2.58 0.995
700 1260 13140 0.0196 2085 0.498 7.75 8.31 0.0505 0.0292 0.772 2.99 1.000
750 1350 0.2931 0.0183 2119 0.506 8.88 9.56 0.0549 0.0317 0.883 3.42 0.005
800 1440 0.2739 0.0171 2152 0.514 10.20 10.98 0.0592 0.0342 1.001 3.88 1.010
850 1530 0.2579 0.0161 2186 0.522 11.52 12.40 0.0637 0.0368 1.130 4.38 1.019
481
TABLE C.14
Thermodynamic Properties of Steam: Temperature Table (SI Units)
Specific Volume (m3/kg) Internal Energy (kJ/kg) Enthalpy (kJ/kg) Entropy (kJ/kg K)
Psat
Tsat (°C) (kPa) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
0 0.61 0.001000 206.13 206.13 0.00 2373.9 2373.9 0.0 2500.0 2500.0 −0.0012 9.1590 9.1578
5 0.87 0.001000 147.20 147.20 21.04 2361.1 2382.1 21.0 2489.6 2510.6 0.0757 8.9510 9.0267
10 1.23 0.001000 106.36 106.36 42.02 2347.8 2389.8 42.0 2478.4 2520.4 0.1509 8.7511 8.9020
15 1.71 0.001001 78.036 78.037 62.95 2333.7 2396.7 63.0 2466.8 2529.7 0.2244 8.5582 8.7827
20 2.34 0.001002 57.801 57.802 83.86 2319.9 2403.7 83.9 2455.0 2538.9 0.2965 8.3718 8.6684
25 3.17 0.001003 43.446 43.447 104.75 2305.5 2410.3 104.8 2443.1 2547.9 0.3672 8,1919 8.5591
30 4.24 0.001004 32.907 32.908 125.63 2291.6 2417.2 125.6 2431.2 2556.8 0.4367 8.0180 8.4546
35 5.62 0.001006 25.250 25.251 146.50 2277.3 2423.8 146.5 2419.2 2565.7 0.5049 7.8496 8.3545
40 7.37 0.001008 19.536 19.537 167.37 2263.2 2430.6 167.4 2407.3 2574.6 0.5720 7.6864 8.2584
45 9.58 0.001010 15.262 15.263 188.24 2249.1 2437.3 188.3 2395.3 2583.5 0.6381 7.5281 8.1662
50 12.33 0.001012 12.046 12.047 209.12 2234.7 2443.8 209.1 2383.2 2592.3 0.7031 7.3745 8.0776
55 15.74 0.001014 9.5771 9.5781 230.01 2220.4 2450.4 230.0 2371.1 2601.1 0.7672 7.2253 7.9925
60 19.92 0.001017 7.6776 7.6786 250.91 2206.0 2456.9 250.9 2358.9 2609.8 0.8303 7.0804 7.9107
65 25.00 0.001020 6.1996 6.2006 271.83 2191.6 2463.4 271.9 2346.6 2618.4 0.8926 6.9394 7.8320
70 31.15 0.001023 5.0452 5.0462 292.76 2177.0 2469.7 292.8 2334.2 2626.9 0.9540 6.8023 7.7563
75 38.54 0.001026 4.1328 4.1338 313.70 2162.3 2476.0 313.7 2321.6 2635.4 1.0146 6.6687 7.6834
80 47.35 0.001029 3.4074 3.4085 334.67 2147.6 2482.3 334.7 2309.8 2643.7 1.0744 6.5387 7.6131
85 57.80 0.001032 2.8276 2.8286 355.65 2132.8 2488.4 355.7 2296.2 2651.9 1.1335 6.4118 7.5453
90 70.10 0.001036 2.3604 2.3614 376.66 2117.8 2494.5 376.7 2283.3 2660.0 1.1917 6.2881 7.4798
95 84.52 0.001039 1.9806 1.9817 397.69 2102.8 2500.5 397.8 2270.2 2668.0 1.2493 6.1673 7.4166
100 101.32 0.001043 1.6689 1.6699 418.75 2087.9 2506.6 418.9 2257.0 2675.8 1.3062 6.0492 7.3554
105 120.80 0.001047 1.4142 1.4152 439.83 2072.8 2512.6 440.0 2243.6 2683.6 1.3624 5.9338 7.2962
110 143.27 0.001051 1.2063 1.2074 460.95 2057.2 2518.2 461.1 2230.0 2691.1 1.4179 5.8209 7.2388
115 169.07 0.001056 1.0350 1.0361 482.10 2041.3 2523.4 482.3 2216.3 2698.6 1.4728 5.7105 7.1833
120 198.55 0.001060 0.89100 0.8921 503.28 2025.4 2528.7 503.5 2202.3 2705.8 1.5271 5.6023 7.1293
125 232.11 0.001065 0.76938 0.7704 524.51 2009.6 2534.1 524.8 2188.2 2712.9 1.5807 5.4962 7.0770
130 270.15 0.001070 0.66702 0.6681 545.78 1993.6 2539.4 546.1 2173.8 2719.9 1.6338 5.3922 7.0261
135 313.09 0.001075 0.58074 0.5818 567.09 1977.3 2544.4 567.4 2159.2 2726.6 1.6864 5.2902 6.9766
140 361.39 0.001080 0.50739 0.5085 588.46 1960.9 2549.3 588.8 2144.3 2733.1 1.7384 5.1900 6.9284
145 415.53 0.001085 0.44462 0.4457 609.88 1944.3 2554.2 610.3 2129.1 2739.4 1.7899 5.0916 6.8815
150 475.99 0.001091 0.39100 0.3921 631.35 1927.5 2558.8 631.9 2113.6 2745.5 1.8409 4.9948 6.8358
155 543.30 0.001096 0.34514 0.3462 652.89 1910.3 2563.2 653.5 2097.8 2751.3 1.8915 4.8996 6.7911
160 618.00 0.001102 0.30566 0.3068 674.50 1892.8 2567.3 675.2 2081.7 2756.9 1.9416 4.8059 6.7475
165 700.68 0.001108 0.27131 0.2724 696.18 1875.1 2571.3 697.0 2065.2 2762.2 1.9912 4.7135 6.7048
170 791.86 0.001114 0.24141 0.2425 717.93 1857.2 2575.2 718.8 2048.4 2767.2 2.0405 4.6224 6.6630
175 892.20 0.001121 0.21538 0.2165 739.77 1839.0 2578.8 740.8 2031.2 2772.0 2.0894 4.5325 6.6220
180 1,002.3 0.001127 0.19266 0.1938 761.69 1820.5 2582.1 762.8 2013.6 2776.4 2.1380 4.4437 6.5817
185 1,122.9 0.001134 0.17272 0.1739 783.70 1803.6 2585.3 785.0 1995.5 2780.5 2.1862 4.3559 6.5421
190 1,254.5 0.001141 0.15513 0.1563 805.80 1782.4 2588.3 807.2 1977.1 2784.3 2.2341 4.2691 6.5032
195 1,398.0 0.001148 0.13964 0.1408 828.01 1762.9 2590.9 829.6 1958.1 2787.8 2.2817 4.1834 6.4651
200 1,553.9 0.001156 0.12597 0.1271 850.32 1743.0 2593.3 852.1 1938.8 2790.9 2.3290 4.0986 6.4276
205 1,723.1 0.001164 0.11386 0.1150 872.74 1722.7 2595.4 874.7 1918.9 2793.6 2.3761 4.0147 6.3908
210 1,906.3 0.001172 0.10307 0.1042 895.28 1702.0 2597.3 897.5 1898.5 2796.0 2.4230 3.9314 6.3544
215 2,104.3 0.001180 0.09345 0.0946 917.94 1681.0 2598.9 920.4 1877.6 2798.0 2.4696 3.8485 6.3181
220 2,317.8 0.001189 0.08486 0.0860 940.73 1659.5 2600.2 943.5 1856.2 2799.7 2.5161 3.7661 6.2821
225 2,547.8 0.001198 0.07716 0.0784 963.66 1637.6 2601.3 966.7 1834.2 2800.9 2.5623 3.6841 6.2464
230 2,795.0 0.001208 0.07022 0.0714 986.73 1615.4 2602.2 990.1 1811.7 2801.8 2.6084 3.6025 6.2109
235 3,060.3 0.001218 0.06400 0.0652 1010.0 15927 2602.7 1033.7 1788.6 2802.3 2.6544 3.5213 6.1757

Thermodynamic Properties of Steam: Temperature Table (SI Units)
Psat
Tsat (°C) (kPa) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
240 3,344.7 0.001228 0.05851 0.0597 1033.6 1569.1 2602.5 1037.5 1764.8 2802.3 2.7002 3.4404 6.1406
245 3,649.0 0.001239 0.05353 0.0548 1056.9 1545.2 2602.1 1061.4 1740.5 2801.9 2.7460 3.3597 6.1057
250 3,974.2 0.001250 0.04893 0.0502 1080.7 1521.0 2601.7 1085.6 1715.5 2801.2 2.7917 3.2792 6.0708
255 4,321.3 0.001262 0.04471 0.0460 1104.6 1496.7 2601.3 1110.1 1689.9 2800.0 2,8373 3.1986 6.0359
260 4,691.2 0.001275 0.04086 0.0421 1128.8 1471.9 2600.7 1134.8 1663.5 2798.3 2.8829 3.1180 6.0009
265 5,085.0 0.001288 0.03738 0.0387 1153.2 1446.4 2599.6 1159.8 1636.5 2796.3 2.9286 3.0372 5.9657
270 5,503.8 0.001302 0.03424 0.0355 1177.9 1420.3 2598.1 1185.1 1608.7 2793.7 2.9743 2.9560 5.9303
275 5,948.6 0.001317 0.03139 0.0327 1202.8 1393.4 2596.3 1210.7 1580.1 2790.8 3.0200 2.8745 5.8945
280 6,420.5 0.001333 0.02878 0.0301 1228.1 1366.0 2594.0 1236.6 1550.8 2787.4 3.0660 2.7924 5.8584
285 6,920.8 0.001349 0.02639 0.0277 1253.7 1337.9 2591.6 1263.0 1520.6 2783.6 3.1121 2.7097 5.8218
290 7,450.6 0.001366 0.02418 0.0255 1279.6 1297.7 2577.3 1289.8 1477.9 2767.7 3.1585 2.6262 5.7847
295 8,011.1 0.001385 0.02214 0.0235 1306.0 1265.5 2571.5 1317.1 1442.8 2759.9 3.2052 2.5417 5.7469
300 8,603.7 0.001404 0.02025 0.0217 1332.8 1232.0 2564.8 1344.9 1406.2 2751.1 3.2523 2.4560 5.7083
305 9,214.4 0.001425 0.01850 0.0199 1360.2 1197.5 2557.6 1373.3 1367.9 2741.2 3.3000 2.3688 5.6687
310 9,869.4 0.001447 0.01688 0.0183 1388.0 1161.1 2549.2 1402.3 1327.8 2730.1 3.3483 2.2797 5.6279
315 10,561. 0.001470 0.01538 0.0169 1416.5 1123.0 2539.6 1432.1 1285.5 2717.6 3.3973 2.1884 5.5858
320 11,289. 0.001499 0.01398 0.0155 1445.7 1083.0 2528.7 1462.6 1240.9 2703.5 3.4473 2.0947 5.5420
325 12,056. 0.001528 0.01267 0.0142 1475.5 1040.9 2516.4 1494.0 1193.6 2687.3 3.4984 1.9979 5.4962
330 12,862. 0.001561 0.01143 0.0130 1506.2 996.3 2502.5 1526.3 1143.3 2669.6 3.5507 1.8973 5.4480
335 13,712. 0.001598 0.01026 0.0119 1537.8 949.0 2486.8 1559.7 1089.6 2649.3 3.6045 1.7922 5.3967
340 14,605. 0.001639 0.00914 0.0108 1570.4 898.7 2469.0 1594.3 1032.2 2626.5 3.6601 1.6820 5.3420
345 15,545. 0.001686 0.00808 0.0098 1606.3 842.1 2448.4 1632.5 967.7 2600.2 3.7176 1.5658 5.2834
350 16,535. 0.001741 0.00706 0.0088 1643.0 780.5 2423.5 1671.8 897.2 2569.0 3.7775 1.4416 5.2191
355 17,577. 0.001808 0.00605 0.0079 1682.1 710.9 2393.0 1713.9 817.3 2531.2 3.8400 1.3054 5.1454
360 18,675. 0.001896 0.00504 0.0069 1726.2 629.5 2355.7 1761.6 723.7 2485.3 3.9056 1.1531 5.0587
365 19,833. 0.002016 0.00400 0.0060 1777.9 531.0 2308.9 1817.8 610.3 2428.1 3.9746 0.9822 4.9569
370 21,054. 0.002225 0.00274 0.0050 1843.3 394.1 2237.3 1890.1 451.9 2342.0 4.0476 0.7555 4.8030
374.4 22,090. 0.00315 0.00000 0.00315 2029.6 0.0 2029.6 2099.3 0.0 2099.3 4.4298 0.0 4.4298
Source: Kraus, A.D., Welty, J.R., and Aziz, A., Introduction to Thermal and Fluid Engineering, CRC Press, Boca Raton, FL, 2012.
TABLE C.15
Thermodynamic Properties of Steam: Pressure Table (SI Units)
Tsat
Psat (kPa) (°C) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
1.00 7.0 0.001000 129.08 129.08 29.40 2356.1 2385.5 29.4 2485.2 2514.6 0.1058 8.8704 8.9763
1.50 13.0 0.001001 88.067 88.068 54.68 2339.3 2394.0 54.7 2471.4 2526.3 0.1956 8.6337 8.8292
2.00 17.5 0.001001 67.073 67.074 73.41 2326.8 2400.2 73.4 2460.9 2534.3 0.2607 8.4642 8.7249
2.50 21.1 0.001002 54.290 54.291 88.41 2316.7 2405.1 88.4 2452.4 2540.8 0.3120 8.3322 8.6442
3.00 24.1 0.001003 45.751 45.752 100.96 2308.0 2409.0 101.0 2445.3 2546.2 0.3545 8.2240 8.5786
3.50 26.7 0.001003 39.483 39.484 111.81 2300.9 2412.7 111.8 2439.1 2550.9 0.3908 8.1324 8.5233
4.00 29.0 0.001004 34.779 34.780 121.37 2294.5 2415.9 121.4 2433.6 2555.0 0.4226 8.0529 8.4756
4.50 31.0 0.001005 31.128 31.129 129.95 2288.6 2418.6 130.0 2428.7 2558.7 0.4509 7.9827 8.4336
5.00 32.9 0.001005 28.194 28.195 137.73 2283.3 2421.0 137.7 2424.3 2562.0 0.4764 7.9197 8.3961
5.50 34.6 0.001006 25.773 25.774 144.86 2278.4 2423.3 144.9 2420.2 2565.0 0.4996 7.8626 8.3622
6.00 36.2 0.001006 23.742 23.743 151.45 2274.0 2425.4 151.5 2416.4 2567.9 0.5209 7.8104 8.3313
6.50 37.7 0.001007 22.013 22.014 157.58 2269.8 2427.4 157.6 2412.9 2570.5 0.5407 7.7623 8.3030
7.00 39.0 0.001007 20.522 20.523 163.31 2266.0 2429.3 163.3 2409.6 2572.9 0.5590 7.7177 8.2768
7.50 40.3 0.001008 19.225 19.226 168.70 2262.3 2431.0 168.7 2406.5 2575.2 0.5763 7.6762 8.2524
8.00 41.5 0.001008 18.086 18.087 173.79 2258.9 2432.7 173.8 2403.6 2577.4 0.5924 7.6372 8.2296
8.50 42.7 0.001009 17.080 17.081 178.61 2255.6 2434.2 178.6 2400.8 2579.4 0.6077 7.6006 8.2083
9.00 43.8 0.001009 16.185 16.186 183.19 2252.5 2435.7 183.2 2398.2 2581.4 0.6222 7.5660 8.1881
9.50 44.8 0.001010 15.383 15.384 187.56 2249.5 2437.1 187.6 2395.7 2583.2 0.6359 7.5332 8.1691
10.00 45.8 0.001010 14.660 14.661 191.74 2246.7 2438.4 191.7 2393.3 2585.0 0.6490 7.5021 8.1511
15.00 54.0 0.001014 10.020 10.021 225.83 2223.2 2449.0 225.8 2373.5 2599.4 0.7544 7.2548 8.0092
20.00 60.1 0.001017 7.6483 7.6493 251.28 2205.7 2457.0 251.3 2358.7 2610.0 0.8314 7.0779 7.9093
25.00 65.0 0.001020 6.2015 6.2025 271.80 2191.6 2463.3 271.8 2346.6 2618.4 0.8925 6.9396 7.8321
30.00 69.1 0.001022 5.2277 5.2287 289.09 2379.5 2468.6 289.1 2336.3 2625.5 0.9433 6.8260 7.7693
35.00 72.7 0.001024 4.5249 4.5259 304.11 2169.0 2473.1 304.1 2327.4 2631.5 0.9869 6.7295 7.7164
40.00 75.9 0.001026 3.9918 3.9929 317.42 2159.7 2477.1 317.5 2319.4 2636.8 1.0253 6.6455 7.6707
45.00 78.7 0.001028 3.5744 3.5755 329.40 2151.3 2480.7 329.4 2312.2 2641.6 1.0595 6.5710 7.6305
50.00 81.3 0.001030 3.2389 3.2398 340.31 2143.6 2483.9 340.4 2305.5 2645.9 1.0904 6.5042 7.5946
60.00 86.0 0.001033 2.7305 2.7316 359.66 2129.9 2489.6 359.7 2293.7 2653.5 1.1446 6.3880 7.5326
70.00 90.0 0.001036 2.3638 2.3648 376.49 2117.9 2494.4 376.6 2283.4 2659.9 1.1913 6.2891 7.4804
80.00 93.5 0.001038 2.0859 2.0869 391.43 2107.2 2498.7 391.5 2274.1 2665.6 1.2323 6.2029 7.4352
90.00 96.7 0.001041 1.8667 1.8678 404.90 2097.7 2502.6 405.0 2265.7 2670.7 1.2689 6.1265 7.3954
100.00 99.6 0.001043 1.6898 1.6908 417.20 2089.0 2506.2 417.3 2258.0 2675.3 1.3020 6.0578 7.3598
101.32 100.0 0.001043 1.66895 1.6700 418.74 2087.9 2506.6 418.8 2257.0 2675.8 1.3062 6.0493 7.3554
125.00 106.0 0.001048 1.36965 1.3707 444.01 2069.7 2513.7 444.1 2240.9 2685.1 1.3734 5.9113 7.2847
150.00 111.4 0.001053 1.15612 1.1572 466.74 2052.9 2519.6 466.9 2226.3 2693.2 1.4330 5.7904 7.2234
175.00 116.1 0.001057 1.00248 1.0035 486.58 2037.9 2524.5 486.8 2213.4 2700.1 1.4844 5.6873 7.1717
200.00 120.2 0.001060 0.88498 0.8860 504.25 2024.7 2529.0 504.5 2201.7 2706.2 1.5295 5.5974 7.1269
225.00 124.0 0.001064 0.79229 0.7934 520.22 2012.8 2533.0 520.5 2191.1 2711.5 1.5700 5.5175 7.0874
250.00 127.4 0.001067 0.71751 0.7186 534.82 2001.9 2536.7 535.1 2181.2 2716.3 1.6066 5.4455 7.0521
275.00 130.6 0.001070 0.65602 0.6571 548.30 1991.7 2540.0 548.6 2172.1 2720.7 1.6401 5.3800 7.0201
300.00 133.5 0.001073 0.60457 0.6056 560.83 1982.1 2542.9 561.2 2163.5 2724.6 1.6710 5.3199 6.9910
325.00 136.3 0.001076 0.56082 0.5619 572.57 1973.1 2545.7 572.9 2155.4 2728.3 1.6998 5.2643 6.9641
350.00 138.9 0.001079 0.52305 0.5241 583.60 1964.6 2548.2 584.0 2147.7 2731.6 1.7266 5.2126 6.9392
375.00 141.3 0.001081 0.49007 0.4911 594.03 1956.6 2550.6 594.4 2140.3 2734.8 1.7519 5.1642 6.9161
400.00 143.6 0.001084 0.46105 0.4621 603.93 1948.9 2552.8 604.4 2133.3 2737.7 1.7757 5.1187 6.8944
425.00 145.8 0.001086 0.43534 0.4364 613.36 1941.6 2554.9 613.8 2126.6 2740.4 1.7982 5.0758 6.8740
450.00 147.9 0.001088 0.41242 0.4135 622.35 1934.5 2556.9 622.8 2120.1 2743.0 1.8196 5.0352 6.8548
475.00 149.9 0.001091 0.39188 0.3930 630.97 1927.7 2558.7 631.5 2113.9 2745.4 1.8400 4.9966 6.8366
500.00 151.8 0.001093 0.37336 0.3745 639.24 1921.2 2560.4 639.8 2107.9 2747.6 1.8595 4.9598 6.8193

Thermodynamic Properties of Steam: Pressure Table (SI Units)
Tsat
Psat (kPa) (°C) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
550.00 155.5 0.001097 0.34129 0.3424 654.86 1908.7 2563.5 655.5 2096.4 2751.8 1.8960 4.8910 6.7871
600.00 158.8 0.001101 0.31443 0.3155 669.42 1896.9 2566.3 670.1 2085.5 2755.6 1.9298 4.8278 6.7576
650.00 162.0 0.001104 0.29151 0.2926 683.06 1885.8 2568.8 683.8 2075.2 2759.0 1.9613 4.7692 6.7305
700.00 164.9 0.001108 0.27168 0.2728 695.93 1875.3 2571.2 696.7 2065.4 2762.1 1.9907 4.7146 6.7053
750.00 167.7 0.001111 0.25439 0.2555 708.11 1865.3 2573.4 708.9 2056.0 2765.0 2.0183 4.6634 6.6817
800.00 170.4 0.001115 0.23919 0.2403 719.68 1855.7 2575.4 720.6 2047.0 2767.6 2.0445 4.6152 6.6597
850.00 172.9 0.001118 0.22572 0.2268 730.71 1846.5 2577.2 731.7 2038.4 2770.0 2.0692 4.5696 6.6388
900.00 175.3 0.001121 0.21372 0.2148 741.27 1837.7 2578.9 742.3 20300 2.772.3 2.0928 4.5264 6.6192
950.00 177.7 0.001124 0.20295 0.2041 751.39 1829.1 2580.5 752.5 2021.9 2774.3 2.1152 4.4853 6.6005
1,000 179.9 0.001127 0.19322 0.1943 761.11 1820.8 2581.9 762.2 2014.0 2776.3 2.1367 4.4460 6.5827
1,100 184.1 0.001133 0.17631 0.1774 779.52 1805.1 2584.6 780.8 1999.0 2779.8 2.1771 4.3725 6.5495
1,200 187.9 0.001138 0.16209 0.1632 796.71 1790.2 2586.9 798.1 1984.7 2782.8 2.2145 4.3047 6.5191
1,300 191.6 0.001143 0.14998 0.1511 812.87 1776.1 2589.0 814.4 1971.1 2785.4 2.2493 4.2417 6.4910
1,400 195.0 0.001148 0.13956 0.1407 828.13 1762.7 2590.8 829.7 1958.0 2787.8 2.2820 4.1829 6.4649
1,500 198.3 0.001153 0.13050 0.1317 842.61 1749.8 2592.4 844.3 1945.5 2789.8 2.3127 4.1277 6.4405
1,600 201.4 0.001158 0.12254 0.1237 856.39 1737.3 2593.7 858.2 1933.4 2791.7 2.3418 4.0757 6.4176
1,700 204.3 0.001163 0.11549 0.1167 869.56 1725.4 2595.0 871.5 1921.7 2793.3 2.3695 4.0265 6.3960
1,800 207.1 0.001167 0.10918 0.1104 882.18 1713.9 2596.1 884.3 1910.4 2794.7 2.3958 3.9797 6.3755
1,900 209.8 0.001172 0.10351 0.1047 894.30 1702.7 2597.0 896.5 1899.4 2795.9 2.4210 3.9350 6.3559
2,000 212.4 0.001176 0.09839 0.0996 905.97 1691.9 2597.9 908.3 1888.7 2797.0 2.4450 3.8922 6.3372
2,250 218.4 0.001186 0.08751 0.0887 933.41 1666.2 2599.6 936.1 1863.1 2799.2 2.5012 3.7925 6.2936
2,500 223.9 0.001196 0.07873 0.0799 958.75 1642.1 2600.9 961.7 1839.0 2800.7 2.5525 3.7015 6.2540
2,750 229.1 0.001206 0.07147 0.0727 982.37 1619.4 2601.8 985.7 1816.0 28017 2.5997 3.6178 6.2176
3,000 233.8 0.001215 0.06538 0.0666 1004.5 1597.9 2602.4 1008.2 1794.0 2802.2 2.6437 3.5402 6.1839
3,250 238.3 0.001225 0.06028 0.0615 1025.4 1577.0 2602.5 1029.4 1772.9 2802.3 2.6848 3.4676 6.1524
3,500 242.5 0.001234 0.05594 0.0572 1045.3 1556.8 2602.1 1049.6 1752.6 2802.2 2.7234 3.3995 6.1229
3,750 246.5 0.001242 0.05208 0.0533 1064.2 1537.6 2601.8 1068.8 1732.9 2801.7 2.7600 3.3350 6.0950
4,000 250.3 0.001251 0.04864 0.0499 1082.2 1519.3 2601.5 1087.2 1713.9 2801.1 2.7947 3.2739 6.0686
5,000 263.9 0.001285 0.03811 0.0394 1147.9 1451.9 2599.8 1154.3 1642.4 2796.7 2.9186 3.0548 5.9734
6,000 275.5 0.001319 0.03109 0.0324 1205.6 1390.4 2596.0 1213.5 1576.9 2790.5 3.0251 2.8655 5.8906
7,000 285.8 0.001352 0.02603 0.0274 1257.8 1333.5 2591.3 1267.2 1515.7 2782.9 3.1194 2.6965 5.8159
8,000 295.0 0.001385 0.02214 0.0235 1305.9 1265.8 2571.7 1317.0 1442.9 2759.9 3.2050 2.5421 5.7471
9,000 303.3 0.001418 0.01907 0.0205 1351.0 1209.3 2560.3 1363.7 1380.9 2744.7 3.2840 2.3981 5.6821
10,000 311.0 0.001452 0.01658 0.0180 1393.7 1153.8 2547.5 1408.2 1319.5 2727.7 3.3580 2.2616 5.6196
11,000 318.1 0.001489 0.01450 0.0160 1434.6 1098.6 2533.1 1451.0 1258.0 2709.0 3.4283 2.1305 5.5588
12,000 324.7 0.001527 0.01273 0.0143 1474.0 1043.3 2517.3 1492.4 1196.0 2688.4 3.4958 2.0028 5.4986
13,000 331.0 0.001568 0.01119 0.0128 1512.4 987.6 2499.9 1532.8 1133.1 2665.8 3.5611 1.8771 5.4382
14,000 336.9 0.001612 0.00984 0.0114 1549.8 931.1 2480.9 1572.4 1068.8 2641.2 3.6249 1.7520 5.3769
15,000 342.4 0.001661 0.00862 0.0103 1586.6 873.4 2460.0 1611.5 1002.8 2614.3 3.6877 1.6265 5.3142
16,000 347.7 0.001715 0.00752 0.0092 1626.1 810.1 2436.2 1653.6 930.4 2584.0 3.7498 1.4996 5.2494
17,000 352.3 0.001769 0.00660 0.0084 1660.2 750.3 2410.5 1690.3 862.5 2552.8 3.8054 1.3819 5.1872
18,000 357.0 0.001839 0.00566 0.0075 1698.6 680.8 2379.4 1731.7 782.6 2514.3 3.8652 1.2481 5.1134
19,000 361.4 0.001926 0.00475 0.0067 1740.1 603.6 2343.6 1776.7 693.9 2470.5 3.9249 1.1063 5.0312
20,000 365.7 0.002037 0.00384 0.0059 1785.8 515.1 2300.9 1826.6 591.9 2418.5 3.9846 0.9568 4.9414
21,000 369.8 0.002208 0.00281 0.0050 1839.7 401.7 2241.4 1886.0 460.8 2346.8 4.0443 0.7681 4.8124
22,000 373.7 0.002623 0.00114 0.0038 1944.6 174.0 2118.6 2002.3 199.0 2201.3 4.1042 0.4563 4.5605
22,090 374.4 0.00315 0.00000 0.00315 2029.6 0.0 2029.6 2099.3 0.0 2099.3 4.4298 0.0 4.4298
Source: Kraus, A.D., Welty, J.R., and Aziz, A., Introduction to Thermal and Fluid Engineering, CRC Press, Boca Raton, FL, 2012.

Thermodynamic Properties of Steam: Superheated Vapor Thermodynamic Properties of Steam: Superheated Vapor
Table (SI Units) Table (SI Units)
T v u h s T v u h s
P (kPa) (°C) (m3/kg) (kJ/kg) (kJ/kg) (kJ/kg k) P (kPa) (°C) (m3/kg) (kJ/kg) (kJ/kg) (kJ/kg k)
10 (Tsat = 45.8°C) 750 4.7208 3564.8 4036.9 9.4471
100 17.196 2516.2 2688.1 8.4498 800 4.9518 3657.3 4152.4 9.5574
150 19.513 2588.2 2783.3 8.6893 850 5.1827 3751.5 4269.8 9.6643
200 21.826 2661.2 2879.5 8.9040 101.32
250 24.136 2735.5 2976.8 9.0996 (Tsat = 100.0°C)
300 26.446 2811.2 3075.6 9.2799 150 1.9109 2583.1 2776.7 7.6084
350 28.754 2888.3 3175.9 9.4476 200 2.1439 2658.0 2875.3 7.8286
400 31.063 2967.1 3277.7 9.6048 250 2.3747 2733.3 2973.9 8.0268
450 33.371 3047.4 3381.1 9.7530 300 2.6043 2809.6 3073.4 8.2085
500 35.679 3129.4 3486.2 9.8935 350 2.8334 2887.1 3174.2 8.3770
550 37.987 3213.1 3593.0 10.027 400 3.0621 2966.1 3276.3 8.5347
600 40.295 3298.5 3701.5 10.155 450 3.2905 3046.6 3380.0 8.6832
650 42.603 3385.7 3811.7 10.278 500 3.5189 3128.7 3485.3 8.8240
700 44.911 3474.5 3923.7 10.396 550 3.7471 3212.5 3592.2 8.9580
750 47.219 3565.2 4037.4 10.510 600 3.9752 3298.0 3700.8 9.0860
800 49.526 3657.6 4152.9 10.620 650 4.2033 3385.2 3811.1 9.2089
850 51.834 3751.8 4270.2 10.727 700 4.4313 3474.1 3923.1 9.3271
50 (Tsat = 81.3°C) 750 4.6593 3564.8 4036.9 9.4411
100 3.4182 2512.0 2682.9 7.6959 800 4.8872 3657.3 4152.4 9.5513
150 3.8894 2586.0 2780.5 7.9413 850 5.1152 3751.5 4269.8 9.6582
200 4.3561 2659.8 2877.6 8.1583 200 (Tsat = 120.2°C)
250 4.8206 2734.5 2975.6 8.3551 150 0.9596 2577.2 2769.1 7.2804
300 5.2840 2810.5 3074.7 8.5360 200 1.0804 2654.5 2870.6 7.5072
350 5.7468 2887.8 3175.1 8.7040 250 1.1989 2730.9 2970.7 7.7084
400 6.2092 2966.6 3277.1 8.8614 300 1.3162 2807.8 3071.1 7.8916
450 6.6715 3047.1 3380.6 9.0098 350 1.4329 2885.8 3172.4 8.0610
500 7.1336 3129.1 3485.8 9.1504 400 1.5492 2965.0 3274.9 8.2192
550 7.5956 3212.9 3592.6 9.2843 450 1.6653 3045.7 3378.8 8.3682
600 8.0575 3298.3 3701.2 9.4123 500 1.7813 3128.0 3484.3 8.5092
650 8.5193 3385.5 3811.4 9.5351 550 1.8971 3211.9 3591.3 8.6434
700 8.9811 3474.4 3923.4 9.6532 600 2.0129 3297.4 3700.0 8.7716
750 9.4428 3565.0 4037.2 9.7672 650 2.1286 3384.7 3810.4 8.8945
800 9.9045 3657.5 4152.7 9.8775 700 2.2442 3473.7 3922.5 9.0128
850 10.366 3751.7 4270.0 9.9843 750 2.3598 3564.4 4036.4 9.1268
100 (Tsat = 99.6°C) 800 2.4754 3656.9 4152.0 9.2372
100 1.6956 2506.4 2676.0 7.3610 850 2.5909 3751.1 4269.3 9.3441
150 1.9363 2583.1 2776.8 7.6146
200 2.1724 2658.1 2875.3 7.8347 300 (Tsat = 133.5°C)
250 2.4062 2733.3 2974.0 8.0329 150 0.6338 2570.8 2760.9 7.0779
300 2.6388 2809.6 3073.5 8.2146 200 0.7164 2650.8 2865.7 7.3122
350 2.8708 2887.1 3174.2 8.3831 250 0.7965 2728.5 2967.4 7.5165
400 3.1025 2966.1 3276.4 8.5407 300 0.8753 2806.1 3068.7 7.7014
450 3.3340 3046.6 3380.0 8.6893 350 0.9535 2884.4 3170.5 7.8717
500 3.5654 3128.8 3485.3 8.8300 400 1.0314 2963.9 3273.4 8.0305
550 3.7966 3212.5 3592.2 8.9640 450 1.1091 3044.8 3377.6 8.1798
600 4.0277 3298.0 3700.8 9.0921 500 1.1866 3127.2 3483.2 8.3211
650 4.2588 3385.2 3811.1 9.2149 550 1.2639 3211.2 3590.4 8.4554
700 4.4898 3474.1 3923.1 9.3331 600 1.3413 3296.9 3699.2 8.5838

T v u h s T v u h s
650 1.4185 3384.2 3809.7 8.7068 750 0.5891 3561.9 4033.1 8.4846
700 1.4957 3473.2 3921.9 8.8252 800 0.6181 3654.6 4149.1 8.5953
750 1.5728 3564.0 4035.8 8.9393 850 0.6471 3749.0 4266.7 8.7024
800 1.6499 3656.5 4151.5 9.0497 1000 (Tsat = 179.9°C)
850 1.7270 3750.8 4268.9 9.1566 200 0.2059 2621.4 2827.3 6.6930
400 (Tsat = 143.6°C) 250 0.2328 2710.0 2942.8 6.9251
150 0.4707 2563.9 2752.2 6.9287 300 0.2580 2793.1 3051.1 7.1229
200 0.5343 2647.0 2860.7 7.1712 350 0.2824 2874.7 3157.1 7.3003
250 0.5952 2726.0 2964.1 7.3789 400 0.3065 2956.3 3262.7 7.4633
300 0.6549 2804.3 3066.2 7.5655 450 0.3303 3038.5 3368.8 7.6154
350 0.7139 2883.1 3168.6 7.7367 500 0.3540 3121.9 3475.9 7.7585
400 0.7725 2962.9 3271.9 7.8961 550 0.3775 3206.6 3584.2 7.8942
450 0.8309 3044.0 3376.3 8.0458 600 0.4010 3292.8 3693.8 8.0235
500 0.8892 3126.5 3482.2 8.1873 650 0.4244 3380.6 3805.0 8.1473
550 0.9474 3210.6 3589.5 8.3219 700 0.4477 3470.0 3917.7 8.2662
600 1.0054 3296.3 3698.5 8.4504 750 0.4710 3561.0 4032.0 8.3808
650 1.0635 3383.7 3809.0 8.5735 800 0.4943 3653.8 4148.1 8.4916
700 1.1214 3472.7 3921.3 8.6919 850 0.5175 3748.3 4265.8 8.5988
750 1.1793 3563.5 4035.3 8.8062
800 1.2372 3656.1 4151.0 8.9166 1500 (Tsat = 198.3°C)
850 1.2951 3750.4 4268.4 9.0236 250 0.15199 2695.4 2923.4 6.7093
600 (Tsat = 158.8°C) 300 0.16971 2783.3 3037.8 6.9183
200 0.3521 2639.0 2850.2 6.9669 350 0.18654 2867.4 3147.2 7.1014
250 0.3939 2720.8 2957.2 7.1819 400 0.20292 2950.6 3255.0 7.2677
300 0.4344 2800.6 3061.3 7.3719 450 0.21906 3034.0 3362.5 7.4219
350 0.4742 2880.3 3164.8 7.5451 500 0.23503 3118.1 3470.6 7.5664
400 0.5136 2960.7 3268.9 7.7057 550 0.25089 3203.3 3579.7 7.7030
450 0.5528 3042.2 3373.8 7.8562 600 0.26666 3289.9 3689.9 7.8331
500 0.5919 3125.0 3480.1 7.9982 650 0.28237 3378.0 3801.5 7.9574
550 0.6308 3209.3 3587.7 8.1332 700 0.29803 3467.6 3914.7 8.0767
600 0.6696 3295.1 3696.9 8.2619 750 0.31364 3558.9 4029.4 8.1917
650 0.7084 3382.6 3807.7 8.3853 800 0.32921 3651.8 4145.7 8.3027
700 0.7471 3471.8 3920.1 8.5039 850 0.34475 3746.5 4263.6 8.4101
750 0.7858 3562.7 4034.2 8.6182
800 0.8245 3655.3 4150.0 8.7287 2000 (Tsat = 212.4°C)
850 0.8631 3749.7 4267.6 8.8358 250 0.11145 2679.5 2902.4 6.5451
800 (Tsat = 170.4°C) 300 0.12550 2772.9 3023.9 6.7671
200 0.2608 2630.4 2839.1 6.8156 350 0.13856 2860.0 3137.1 6.9565
250 0.2932 2715.5 2950.1 7.0388 400 0.15113 2944.8 3247.1 7.1263
300 0.3241 2796.9 3056,2 7.2326 450 0.16343 3029.3 3356.1 7.2826
350 0.3544 2877.5 3161.0 7.4079 500 0.17556 3114.2 3465.3 7.4286
400 0.3842 2958.5 3265.8 7.5697 550 0.18757 3200.0 3575.1 7.5662
450 0.4137 3040.4 3371.4 7.7209 600 0.19950 3287.0 3686.0 7.6970
500 0.4432 3123.5 3478.0 7.8635 650 0.21137 3375.4 3798.1 7.8218
550 0.4725 3208.0 3586.0 7.9988 700 0.22318 3465.3 3911.6 7.9416
600 0.5017 3294.0 3695.4 8.1278 750 0.23494 3556.8 4026.7 8.0569
650 0.5309 3381.6 3806.3 8.2514 800 0.24667 3649.9 4143.2 8.1681
700 0.5600 3470.9 3918.9 8.3702 850 0.25836 3744.7 4261.5 8.2758
(continued)

T v u h s T v u h s
2500 (Tsat = 223.9°C) 5000 (Tsat = 263.9°C)
250 0.08699 2662.2 2879.7 6.4076 300 0.045302 2698.2 2924.7 6.2085
300 0.09893 2762.0 3009.3 6.6446 350 0.051943 2809.9 3069.6 6.4512
350 0.10975 2852.2 3126.6 6.8409 400 0.057792 2907.7 3196.6 6.6474
400 0.12004 2938.9 3239.1 7.0145 450 0.063252 3000.0 3316.2 6.8188
450 0.13005 3024.6 3349.7 7.1730 500 0.068495 3090.0 3432.5 6.9743
500 0.13987 3110.3 3459.9 7.3205 550 0.073603 3179.5 3547.5 7.1184
550 0.14958 3196.6 3570.6 7.4592 600 0.078617 3269.2 3662.3 7.2538
600 0.15921 3284.1 3682.1 7.5906 650 0.083560 3359.6 3777.4 7.3820
650 0.16876 3372.8 3794.7 7.7161 700 0.088447 3451.1 3893.4 7.5044
700 0.17827 3462.9 3908.6 7.8362 750 0.093289 3543.9 4010.4 7.6216
750 0.18772 3554.6 4024.0 7.9518 800 0.098094 3638.2 4128.7 7.7345
800 0.19714 3648.0 4140.8 8.0633 850 0.102867 3734.0 4248.3 7.8435
850 0.20653 3742.9 4259.3 8.1712
3000 (Tsat = 233.8°C) 6000 (Tsat = 275.6°C)
250 0.07055 2643.2 2854.9 6.2855 300 0.036146 2667.1 2884.0 6.0669
300 0.08116 2750.6 2994.1 6.5399 350 0.042223 2790.9 3044.2 6.3354
350 0.09053 2844.3 3115.9 6.7437 400 0.047380 2894.3 3178.6 6.5429
400 0.09931 2932.9 3230.9 6.9213 450 0.052104 2989.7 3302.3 6.7202
450 0.10779 3019.8 3343.1 7.0822 500 0.056592 3081.7 3421.3 6.8793
500 0.11608 3106.3 3454.5 7.2312 550 0.060937 3172.5 3538.1 7.0258
550 0.12426 3193.2 3566.0 7.3709 600 0.065185 3263.1 3654.2 7.1627
600 0.13234 3281.1 3678.1 7.5031 650 0.069360 3354.3 3770.4 7.2922
650 0.14036 3370.2 3791.3 7.6291 700 0.073479 3446.4 3887.2 7.4154
700 0.14832 3460.6 3905.6 7.7497 750 0.077552 3539.6 4004.9 7.5333
750 0.15624 3552.5 4021.2 7.8656 800 0.081588 3634.3 4123.8 7.6467
800 0.16412 3646.0 4138.4 7.9774 850 0.085592 3730.4 4243.9 7.7562
850 0.17197 3741.1 4257.1 8.0855
4000 (Tsat = 250.3°C) 7000 (Tsat = 285.8°C)
300 0.058835 2725.8 2961.2 6.3622 300 0.029459 2631.9 2838.1 5.9299
350 0.066448 2827.6 3093.4 6.5835 350 0.035234 2770.6 3017.2 6.2301
400 0.073377 2920.6 3214.1 6.7699 400 0.039922 2880.4 3159.8 6.4504
450 0.079959 3010.0 3329.8 6.9358 450 0.044132 2979.1 3288.1 6.6343
500 0.086343 3098.2 3443.6 7.0879 500 0.048087 3073.2 3409.8 6.7971
550 0.092599 3186.4 3556.8 7.2298 550 0.051890 3165.4 3528.6 6.9460
600 0.098764 3275.2 3670.2 7.3636 600 0.055591 3257.0 3646.1 7.0846
650 0.10486 3364.9 3784.3 7.4907 650 0.059218 3348.9 3763.4 7.2153
700 0.11090 3455.9 3899.5 7.6121 700 0.062788 3441.6 3881.1 7.3394
750 0.11690 3548.2 4015.8 7.7287 750 0.066312 3535.3 3999.5 7.4580
800 0.12285 3642.1 4133.5 7.8411 800 0.069798 3630.3 4118.9 7.5720
300 0.045302 2698.2 2924.7 6.2085 850 0.073253 3726.7 4239.5 7.6818
350 0.051943 2809.9 3069.6 6.4512 Source: Kraus, A.D., Welty, J.R., and Aziz, A., Introduction to Thermal
400 0.057792 2907.7 3196.6 6.6474 and Fluid Engineering, CRC Press, Boca Raton, FL, 2012.
850 0.12878 3737.6 4252.7 7.9496
TABLE C.17
Thermodynamic Properties of Steam: Temperature Table (English Units)
Specific Volume (ft3/lbm) Internal Energy (Btu/lbm) Enthalpy (Btu/lbm) Entropy (Btu/lbm · °R)
Sat. Sat. Sat. Sat. Sat. Sat. Sat. Sat. Sat.
Temp., Press., Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor,
T (°F) Psat (psia) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
32.018 0.08871 0.01602 3299.88 3299.9 0.000 1021.0 1021.0 0.000 1075.2 1075.2 0.00000 2.18672 2.1867
35 0.09998 0.01602 2945.68 2945.7 3.004 1019.0 1022.0 3.004 1073.5 1076.5 0.00609 2.17011 2.1762
40 0.12173 0.01602 2443.58 2443.6 8.032 1015.6 1023.7 8.032 1070.7 1078.7 0.01620 2.14271 2.1589
45 0.14756 0.01602 2035.78 2035.8 13.05 1012.2 1025.3 13.05 1067.8 1080.9 0.02620 2.11587 2.1421
50 0.17812 0.01602 1703.08 1703.1 18.07 1008.9 1026.9 18.07 1065.0 1083.1 0.03609 2.08956 2.1256
55 0.21413 0.01603 1430.38 1430.4 23.07 1005.5 1028.6 23.07 1062.2 1085.3 0.04586 2.06377 2.1096
60 0.25638 0.01604 1206.08 1206.1 28.08 1002.1 1030.2 28.08 1059.4 1087.4 0.05554 2.03847 2.0940
65 0.30578 0.01604 1020.78 1020.8 33.08 998.76 1031.8 33.08 1056.5 1089.6 0.06511 2.01366 2.0788
70 0.36334 0.01605 867.16 867.18 38.08 995.39 1033.5 38.08 1053.7 1091.8 0.07459 1.98931 2.0639
75 0.43016 0.01606 739.25 739.27 43.07 992.02 1035.1 43.07 1050.9 1093.9 0.08398 1.96541 2.0494
80 0.50745 0.01607 632.39 632.41 48.06 988.65 1036.7 48.07 1048.0 1096.1 0.09328 1.94196 2.0352
85 0.59659 0.01609 542.78 542.80 53.06 985.28 1038.3 53.06 1045.2 1098.3 0.10248 1.91892 2.0214
90 0.69904 0.01610 467.38 467.40 58.05 981.90 1040.0 58.05 1042.4 1100.4 0.11161 1.89630 2.0079
95 0.81643 0.01612 403.72 403.74 63.04 978.52 1041.6 63.04 1039.5 1102.6 0.12065 1.87408 1.9947
100 0.95052 0.01613 349.81 349.83 68.03 975.14 1043.2 68.03 1036.7 1104.7 0.12961 1.85225 1.9819
110 1.2767 0.01617 264.94 264.96 78.01 968.36 1046.4 78.02 1031.0 1109.0 0.14728 1.80970 1.9570
120 1.6951 0.01620 202.92 202.94 88.00 961.56 1049.6 88.00 1025.2 1113.2 0.16466 1.76856 1.9332
130 2.2260 0.01625 157.07 157.09 97.99 954.73 1052.7 97.99 1019.4 1117.4 0.18174 1.72877 1.9105
140 2.8931 0.01629 122.79 122.81 107.98 947.87 1055.9 107.99 1013.6 1121.6 0.19855 1.69024 1.8888
150 3.7234 0.01634 96.91 96.929 117.98 940.98 1059.0 117.99 1007.8 1125.7 0.21508 1.65291 1.8680
160 4.7474 0.01639 77.17 77.185 127.98 934.05 1062.0 128.00 1001.8 1129.8 0.23136 1.61670 1.8481
170 5.9999 0.01645 61.97 61.982 138.00 927.08 1065.1 138.02 995.88 1133.9 0.24739 1.58155 1.8289
180 7.5197 0.01651 50.16 50.172 148.02 920.06 1068.1 148.04 989.85 1137.9 0.26318 1.54741 1.8106
190 9.3497 0.01657 40.90 40.920 158.05 912.99 1071.0 158.08 983.76 1141.8 0.27874 1.51421 1.7930
200 11.538 0.01663 33.60 33.613 168.10 905.87 1074.0 168.13 977.60 1145.7 0.29409 1.48191 1.7760
210 14.136 0.01670 27.78 27.798 178.15 898.68 1076.8 178.20 971.35 1149.5 0.30922 1.45046 1.7597
212 14.709 0.01671 26.77 26.782 180.16 897.24 1077.4 180.21 970.09 1150.3 0.31222 1.44427 1.7565
220 17.201 0.01677 23.12 23.136 188.22 891.43 1079.6 188.28 965.02 1153.3 0.32414 1.41980 1.7439
230 20.795 0.01684 19.36 19.374 198.31 884.10 1082.4 198.37 958.59 1157.0 0.33887 1.38989 1.7288
240 24.985 0.01692 16.30 16.316 208.41 876.70 1085.1 208.49 952.06 1160.5 0.35342 1.36069 1.7141
250 29.844 0.01700 13.80 13.816 218.54 869.21 1087.7 218.63 945.41 1164.0 0.36779 1.33216 1.6999
260 35.447 0.01708 11.74 11.760 228.68 861.62 1090.3 228.79 938.65 1167.4 0.38198 1.30425 1.6862
270 41.877 0.01717 10.04 10.059 238.85 853.94 1092.8 238.98 931.76 1170.7 0.39601 1.27694 1.6730
280 49.222 0.01726 8.63 8.6439 249.04 846.16 1095.2 249.20 924.74 1173.9 0.40989 1.25018 1.6601
290 57.573 0.01735 7.44 7.4607 259.26 838.27 1097.5 259.45 917.57 1177.0 0.42361 1.22393 1.6475
300 67.028 0.01745 6.45 6.4663 269.51 830.25 1099.8 269.73 910.24 1180.0 0.43720 1.19818 1.6354
310 77.691 0.01755 5.61 5.6266 279.79 822.11 1101.9 280.05 902.75 1182.8 0.45065 1.17289 1.6235
320 89.667 0.01765 4.90 4.9144 290.11 813.84 1104.0 290.40 895.09 1185.5 0.46396 1.14802 1.6120
330 103.07 0.01776 4.29 4.3076 300.46 805.43 1105.9 300.80 887.25 1188.1 0.47716 1.12355 1.6007
340 118.02 0.01787 3.77 3.7885 310.85 796.87 1107.7 311.24 879.22 1190.5 0.49024 1.09945 1.5897
350 134.63 0.01799 3.32 3.3425 321.29 788.16 1109.4 321.73 870.98 1192.7 0.50321 1.07570 1.5789
360 153.03 0.01811 2.94 2.9580 331.76 779.28 1111.0 332.28 862.53 1194.8 0.51607 1.05227 1.5683
370 173.36 0.01823 2.61 2.6252 342.29 770.23 1112.5 342.88 853.86 1196.7 0.52884 1.02914 1.5580
380 195.74 0.01836 2.32 2.3361 352.87 761.00 1113.9 353.53 844.96 1198.5 0.54152 1.00628 1.5478
390 220.33 0.01850 2.07 2.0842 363.50 751.58 1115.1 364.25 835.81 1200.1 0.55411 0.98366 1.5378
400 247.26 0.01864 1.85 1.8639 374.19 741.97 1116.2 375.04 826.39 1201.4 0.56663 0.96127 1.5279
(continued)

Thermodynamic Properties of Steam: Temperature Table (English Units)
Temp., Press., Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor,
T (°F) Psat (psia) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
410 276.69 0.01878 1.65 1.6706 384.94 732.14 1117.1 385.90 816.71 1202.6 0.57907 0.93908 1.5182
420 308.76 0.01894 1.48 1.5006 395.76 722.08 1117.8 396.84 806.74 1203.6 0.59145 0.91707 1.5085
430 343.64 0.01910 1.33 1.3505 406.65 711.80 1118.4 407.86 796.46 1204.3 0.60377 0.89522 1.4990
440 381.49 0.01926 1.20 1.2178 417.61 701.26 1118.9 418.97 785.87 1204.8 0.61603 0.87349 1.4895
450 422.47 0.01944 1.08 1.0999 428.66 690.47 1119.1 430.18 774.94 1205.1 0.62826 0.85187 1.4801
460 466.75 0.01962 0.98 0.99510 439.79 679.39 1119.2 441.48 763.65 1205.1 0.64044 0.83033 1.4708
470 514.52 0.01981 0.88 0.90158 451.01 668.02 1119.0 452.90 751.98 1204.9 0.65260 0.80885 1.4615
480 565.96 0.02001 0.80 0.81794 462.34 656.34 1118.7 464.43 739.91 1204.3 0.66474 0.78739 1.4521
490 621.24 0.02022 0.72 0.74296 473.77 644.32 1118.1 476.09 727.40 1203.5 0.67686 0.76594 1.4428
500 680.56 0.02044 0.66 0.67558 485.32 631.94 1117.3 487.89 714.44 1202.3 0.68899 0.74445 1.4334
510 744.11 0.02067 0.59 0.61489 496.99 619.17 1116.2 499.84 700.99 1200.8 0.70112 0.72290 1.4240
520 812.11 0.02092 0.54 0.56009 508.80 605.99 1114.8 511.94 687.01 1199.0 0.71327 0.70126 1.4145
530 884.74 0.02118 0.49 0.51051 520.76 592.35 1113.1 524.23 672.47 1196.7 0.72546 0.67947 1.4049
540 962.24 0.02146 0.44 0.46553 532.88 578.23 1111.1 536.70 657.31 1194.0 0.73770 0.65751 1.3952
550 1044.8 0.02176 0.40 0.42465 545.18 563.58 1108.8 549.39 641.47 1190.9 0.75000 0.63532 1.3853
560 1132.7 0.02207 0.37 0.38740 557.68 548.33 1106.0 562.31 624.91 1187.2 0.76238 0.61284 1.3752
570 1226.2 0.02242 0.33 0.35339 570.40 532.45 1102.8 575.49 607.55 1183.0 0.77486 0.59003 1.3649
580 1325.5 0.02279 0.30 0.32225 583.37 515.84 1099.2 588.95 589.29 1178.2 0.78748 0.56679 1.3543
590 1430.8 0.02319 0.27 0.29367 596.61 498.43 1095.0 602.75 570.04 1172.8 0.80026 0.54306 1.3433
600 1542.5 0.02362 0.24 0.26737 610.18 480.10 1090.3 616.92 549.67 1166.6 0.81323 0.51871 1.3319
610 1660.9 0.02411 0.22 0.24309 624.11 460.73 1084.8 631.52 528.03 1159.5 0.82645 0.49363 1.3201
620 1786.2 0.02464 0.20 0.22061 638.47 440.14 1078.6 646.62 504.92 1151.5 0.83998 0.46765 1.3076
630 1918.9 0.02524 0.17 0.19972 653.35 418.12 1071.5 662.32 480.07 1142.4 0.85389 0.44056 1.2944
640 2059.3 0.02593 0.15 0.18019 668.86 394.36 1063.2 678.74 453.14 1131.9 0.86828 0.41206 1.2803
650 2207.8 0.02673 0.14 0.16184 685.16 368.44 1053.6 696.08 423.65 1119.7 0.88332 0.38177 1.2651
660 2364.9 0.02767 0.12 0.14444 702.48 339.74 1042.2 714.59 390.84 1105.4 0.89922 0.34906 1.2483
670 2531.2 0.02884 0.10 0.12774 721.23 307.22 1028.5 734.74 353.54 1088.3 0.91636 0.31296 1.2293
680 2707.3 0.03035 0.08 0.11134 742.11 269.00 1011.1 757.32 309.57 1066.9 0.93541 0.27163 1.2070
690 2894.1 0.03255 0.06 0.09451 766.81 220.77 987.6 784.24 253.96 1038.2 0.95797 0.22089 1.1789
700 3093.0 0.03670 0.04 0.07482 801.75 146.50 948.3 822.76 168.32 991.1 0.99023 0.14514 1.1354
705.10 3200.1 0.04975 0 0.04975 866.61 0 866.6 896.07 0 896.1 1.05257 0 1.0526
Source: Cengel, Y. and Boles, M., Thermodynamics: An Engineering Approach, 7th edn., McGraw-Hill, New York, 2011.
TABLE C.18
Thermodynamic Properties of Steam: Pressure Table (English Units)
Press., Temp., Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor,
P (psia) Tsat (°F) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
1 101.69 0.01614 333.47 333.49 69.72 973.99 1043.7 69.72 1035.7 1105.4 0.13262 1.84495 1.9776
2 126.02 0.01623 173.69 173.71 94.02 957.45 1051.5 94.02 1021.7 1115.8 0.17499 1.74444 1.9194
3 141.41 0.01630 118.68 118.70 109.39 946.90 1056.3 109.40 1012.8 1122.2 0.20090 1.68489 1.8858
4 152.91 0.01636 90.61 90.629 120.89 938.97 1059.9 120.90 1006.0 1126.9 0.21985 1.64225 1.8621
5 162.18 0.01641 73.51 73.525 130.17 932.53 1062.7 130.18 1000.5 1130.7 0.23488 1.60894 1.8438
6 170.00 0.01645 61.97 61.982 138.00 927.08 1065.1 138.02 995.88 1133.9 0.24739 1.58155 1.8289
8 182.81 0.01652 47.33 47.347 150.83 918.08 1068.9 150.86 988.15 1139.0 0.26757 1.53800 1.8056
10 193.16 0.01659 38.41 38.425 161.22 910.75 1072.0 161.25 981.82 1143.1 0.28362 1.50391 1.7875
14.696 211.95 0.01671 26.79 26.805 180.12 897.27 1077.4 180.16 970.12 1150.3 0.31215 1.44441 1.7566
15 212.99 0.01672 26.28 26.297 181.16 896.52 1077.7 181.21 969.47 1150.7 0.31370 1.44441 1.7549
20 227.92 0.01683 20.08 20.093 196.21 885.63 1081.8 196.27 959.93 1156.2 0.33582 1.39606 1.7319
25 240.03 0.01692 16.29 16.307 208.45 876.67 1085.1 208.52 952.03 1160.6 0.35347 1.36060 1.7141
30 250.30 0.01700 13.73 13.749 218.84 868.98 1087.8 218.93 945.21 1164.1 0.36821 1.33132 1.6995
35 259.25 0.01708 11.88 11.901 227.92 862.19 1090.1 228.03 939.16 1167.2 0.38093 1.30632 1.6872
40 267.22 0.01715 10.48 10.501 236.02 856.09 1092.1 236.14 933.69 1169.8 0.39213 1.28448 1.6766
45 274.41 0.01721 9.39 9.4028 243.34 850.52 1093.9 243.49 928.68 1172.2 0.40216 1.26506 1.6672
50 280.99 0.01727 8.50 8.5175 250.05 845.39 1095.4 250.21 924.03 1174.2 0.41125 1.24756 1.6588
55 287.05 0.01732 7.77 7.7882 256.25 840.61 1096.9 256.42 919.70 1176.1 0.41958 1.23162 1.6512
60 292.69 0.01738 7.16 7.1766 262.01 836.13 1098.1 262.20 915.61 1177.8 0.42728 1.21697 1.6442
65 297.95 0.01743 6.64 6.6560 267.41 831.90 1099.3 267.62 911.75 1179.4 0.43443 1.20341 1.6378
70 302.91 0.01748 6.19 6.2075 272.50 827.90 1100.4 272.72 908.08 1180.8 0.44112 1.19078 1.6319
75 307.59 0.01752 5.80 5.8167 277.31 824.09 1101.4 277.55 904.58 1182.1 0.44741 1.17895 1.6264
80 312.02 0.01757 5.46 5.4733 281.87 820.45 1102.3 282.13 901.22 1183.4 0.45335 1.16783 1.6212
85 316.24 0.01761 5.15 5.1689 286.22 816.97 1103.2 286.50 898.00 1184.5 0.45897 1.15732 1.6163
90 320.26 0.01765 4.88 4.8972 290.38 813.62 1104.0 290.67 894.89 1185.6 0.46431 1.14737 1.6117
95 324.11 0.01770 4.64 4.6532 294.36 810.40 1104.8 294.67 891.89 1186.6 0.46941 1.13791 1.6073
100 327.81 0.01774 4.41 4.4327 298.19 807.29 1105.5 298.51 888.99 1187.5 0.47427 1.12888 1.6032
110 334.77 0.01781 4.02 4.0410 305.41 801.37 1106.8 305.78 883.44 1189.2 0.48341 1.11201 1.5954
120 341.25 0.01789 3.71 3.7289 312.16 795.79 1107.9 312.55 878.20 1190.8 0.49187 1.09646 1.5883
130 347.32 0.01796 3.44 3.4557 318.48 790.51 1109.0 318.92 873.21 1192.1 0.49974 1.08204 1.5818
140 353.03 0.01802 3.20 3.2202 324.45 785.49 1109.9 324.92 868.45 1193.4 0.50711 1.06858 1.5757
150 358.42 0.01809 3.00 3.0150 330.11 780.69 1110.8 330.61 863.88 1194.5 0.51405 1.05595 1.5700
160 363.54 0.01815 2.82 2.8347 335.49 776.10 1111.6 336.02 859.49 1195.5 0.52061 1.04405 1.5647
170 368.41 0.01821 2.66 2.6749 340.62 771.68 1112.3 341.19 855.25 1196.4 0.52682 1.03279 1.5596
180 373.07 0.01827 2.51 2.5322 345.53 767.42 1113.0 346.14 851.16 1197.3 0.53274 1.02210 1.5548
190 377.52 0.01833 2.39 2.4040 350.24 763.31 1113.6 350.89 847.19 1198.1 0.53839 1.01191 1.5503
200 381.80 0.01839 2.27 2.2882 354.78 759.32 1114.1 355.46 843.33 1198.8 0.54379 1.00219 1.5460
250 400.97 0.01865 1.83 1.8440 375.23 741.02 1116.3 376.09 825.47 1201.6 0.56784 0.95912 1.5270
300 417.35 0.01890 1.52 1.5435 392.89 724.77 1117.7 393.94 809.41 1203.3 0.58818 0.92289 1.5111
350 431.74 0.01912 1.31 1.3263 408.55 709.98 1118.5 409.79 794.65 1204.4 0.60590 0.89143 1.4973
400 444.62 0.01934 1.14 1.1617 422.70 696.31 1119.0 424.13 780.87 1205.0 0.62168 0.86350 1.4852
(continued)

Thermodynamic Properties of Steam: Pressure Table (English Units)
Press., Temp., Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor, Liquid, Evap., Vapor,
P (psia) Tsat (°F) vf vfg vg uf ufg ug hf hfg hg sf sfg sg
450 456.31 0.01955 1.01 1.0324 435.67 683.52 1119.2 437.30 767.86 1205.2 0.63595 0.83828 1.4742
500 467.04 0.01975 0.91 0.92819 447.68 671.42 1119.1 449.51 755.48 1205.0 0.64900 0.81521 1.4642
550 476.97 0.01995 0.82 0.84228 458.90 659.91 1118.8 460.93 743.60 1204.5 0.66107 0.79388 1.4550
600 486.24 0.02014 0.75 0.77020 469.46 648.88 1118.3 471.70 732.15 1203.9 0.67231 0.77400 1.4463
700 503.13 0.02051 0.64 0.65589 488.96 627.98 1116.9 491.62 710.29 1201.9 0.69279 0.73771 1.4305
800 518.27 0.02087 0.55 0.56920 506.74 608.30 1115.0 509.83 689.48 1199.3 0.71117 0.70502 1.4162
900 532.02 0.02124 0.48 0.50107 523.19 589.54 1112.7 526.73 669.46 1196.2 0.72793 0.67505 1.4030
1000 544.65 0.02159 0.42 0.44604 538.58 571.49 1110.1 542.57 650.03 1192.6 0.74341 0.64722 1.3906
1200 567.26 0.02232 0.34 0.36241 566.89 536.87 1103.8 571.85 612.39 1184.2 0.77143 0.59632 1.3677
1400 587.14 0.02307 0.28 0.30161 592.79 503.50 1096.3 598.76 575.66 1174.4 0.79658 0.54991 1.3465
1600 604.93 0.02386 0.23 0.25516 616.99 470.69 1087.7 624.06 539.18 1163.2 0.81972 0.50645 1.3262
1800 621.07 0.02470 0.19 0.21831 640.03 437.86 1077.9 648.26 502.35 1150.6 0.84144 0.46482 1.3063
2000 635.85 0.02563 0.16 0.18815 662.33 404.46 1066.8 671.82 464.60 1136.4 0.86224 0.42409 1.2863
2500 668.17 0.02860 0.10 0.13076 717.67 313.53 1031.2 730.90 360.79 1091.7 0.91311 0.31988 1.2330
3000 695.41 0.03433 0.05 0.08460 783.39 186.41 969.8 802.45 214.32 1016.8 0.97321 0.18554 1.1587
3200.1 705.10 0.04975 0 0.04975 866.61 0 866.6 896.07 0 896.1 1.05257 0 1.0526
TABLE C.19
Thermodynamic Properties of Steam: Superheated Vapor Table (English Units)
v u h s v u h s v u h s
T (°F) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu /lbm °R) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu/lbm °R) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu/lbm °R)
P = 1.0 psia (101.69°F)a P = 5.0 psia (162.18°F) P = 10 psia (193.16°F)
b
Sat. 333.49 1043.7 1105.4 1.9776 73.525 1062.7 1130.7 1.8438 38.425 1072.0 1143.1 1.7875
200 392.53 1077.5 1150.1 2.0509 78.153 1076.2 1148.5 1.8716 38.849 1074.5 1146.4 1.7926
240 416.44 1091.2 1168.3 2.0777 83.009 1090.3 1167.1 1.8989 41.326 1089.1 1165.5 1.8207
280 440.33 1105.0 1186.5 2.1030 87.838 1104.3 1185.6 1.9246 43.774 1103.4 1184.4 1.8469
320 464.20 1118.9 1204.8 2.1271 92.650 1118.4 1204.1 1.9490 46.205 1117.6 1203.1 1.8716
360 488.07 1132.9 1223.3 2.1502 97.452 1132.5 1222.6 1.9722 48.624 1131.9 1221.8 1.8950
400 511.92 1147.1 1241.8 2.1722 102.25 1146.7 1241.3 1.9944 51.035 1146.2 1240.6 1.9174
440 535.77 1161.3 1260.4 2.1934 107.03 1160.9 1260.0 2.0156 53.441 1160.5 1259.4 1.9388
500 571.54 1182.8 1288.6 2.2237 114.21 1182.6 1288.2 2.0461 57.041 1182.2 1287.8 1.9693
600 631.14 1219.4 1336.2 2.2709 126.15 1219.2 1335.9 2.0933 63.029 1219.0 1335.6 2.0167
700 690.73 1256.8 1384.6 2.3146 138.09 1256.7 1384.4 2.1371 69.007 1256.5 1384.2 2.0605
800 750.31 1295.1 1433.9 2.3553 150.02 1294.9 1433.7 2.1778 74.980 1294.8 1433.5 2.1013
1000 869.47 1374.2 1535.1 2.4299 173.86 1374.2 1535.0 2.2524 86.913 1374.1 1534.9 2.1760
1200 988.62 1457.1 1640.0 2.4972 197.70 1457.0 1640.0 2.3198 98.840 1457.0 1639.9 2.2433
1400 1107.8 1543.7 1748.7 2.5590 221.54 1543.7 1748.7 2.3816 110.762 1543.6 1748.6 2.3052
P = 15 psia (212.99°F) P = 20 psia (227.92°F) P = 40 psia (267.22°F)
Sat. 26.297 1077.7 1150.7 1.7549 20.093 1081.8 1156.2 1.7319 10.501 1092.1 1169.8 1.6766
240 27.429 1087.8 1163.9 1.7742 20.478 1086.5 1162.3 1.7406
280 29.085 1102.4 1183.2 1.8010 21.739 1101.4 1181.9 1.7679 10.713 1097.3 1176.6 1.6858
320 30.722 1116.9 1202.2 1.8260 22.980 1116.1 1201.2 1.7933 11.363 1112.9 1197.1 1.7128
360 32.348 1131.3 1221.1 1.8496 24.209 1130.7 1220.2 1.8171 11.999 1128.1 1216.9 1.7376
400 33.965 1145.7 1239.9 1.8721 25.429 1145.1 1239.3 1.8398 12.625 1143.1 1236.5 1.7610
440 35.576 1160.1 1258.8 1.8936 26.644 1159.7 1258.3 1.8614 13.244 1157.9 1256.0 1.7831
500 37.986 1181.9 1287.3 1.9243 28.458 1181.6 1286.9 1.8922 14.165 1180.2 1285.0 1.8143
600 41.988 1218.7 1335.3 1.9718 31.467 1218.5 1334.9 1.9398 15.686 1217.5 1333.6 1.8625
700 45.981 1256.3 1383.9 2.0156 34.467 1256.1 1383.7 1.9837 17.197 1255.3 1382.6 1.9067
800 49.967 1294.6 1433.3 2.0565 37.461 1294.5 1433.1 2.0247 18.702 1293.9 1432.3 1.9478
1000 57.930 1374.0 1534.8 2.1312 43.438 1373.8 1534.6 2.0994 21.700 1373.4 1534.1 2.0227
1200 65.885 1456.9 1639.8 2.1986 49.407 1456.8 1639.7 2.1668 24.691 1456.5 1639.3 2.0902
1400 73.836 1543.6 1748.5 2.2604 55.373 1543.5 1748.4 2.2287 27.678 1543.3 1748.1 2.1522
1600 81.784 1634.0 1861.0 2.3178 61.335 1633.9 1860.9 2.2861 30.662 1633.7 1860.7 2.2096
(continued)
493
494

Thermodynamic Properties of Steam: Superheated Vapor Table (English Units)
v u h s v u h s v u h s
T (°F) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu /lbm °R) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu/lbm °R) (ft3/lbm) (Btu/lbm) (Btu/lbm) (Btu/lbm °R)
Sat. 7.1766 1098.1 1177.8 1.6442 5.4733 1102.3 1183.4 1.6212 4.4327 1105.5 1187.5 1.6032
320 7.4863 1109.6 1192.7 1.6636 5.5440 1105.9 1187.9 1.6271
360 7.9259 1125.5 1213.5 1.6897 5.8876 1122.7 1209.9 1.6545 4.6628 1119.8 1206.1 1.6263
400 8.3548 1140.9 1233.7 1.7138 6.2187 1138.7 1230.8 1.6794 4.9359 1136.4 1227.8 1.6521
440 8.7766 1156.1 1253.6 1.7364 6.5420 1154.3 1251.2 1.7026 5.2006 1152.4 1248.7 1.6759
500 9.4005 1178.8 1283.1 1.7682 7.0177 1177.3 1281.2 1.7350 5.5876 1175.9 1279.3 1.7088
600 10.4256 1216.5 1332.2 1.8168 7.7951 1215.4 1330.8 1.7841 6.2167 1214.4 1329.4 1.7586
700 11.4401 1254.5 1381.6 1.8613 8.5616 1253.8 1380.5 1.8289 6.8344 1253.0 1379.5 1.8037
800 12.4484 1293.3 1431.5 1.9026 9.3218 1292.6 1430.6 1.8704 7.4457 1292.0 1429.8 1.8453
1000 14.4543 1373.0 1533.5 1.9777 10.8313 1372.6 1532.9 1.9457 8.6575 1372.2 1532.4 1.9208
1200 16.4525 1456.2 1638.9 2.0454 12.3331 1455.9 1638.5 2.0135 9.8615 1455.6 1638.1 1.9887
1400 18.4464 1543.0 1747.8 2.1073 13.8306 1542.8 1747.5 2.0755 11.0612 1542.6 1747.2 2.0508
1600 20.438 1633.5 1860.5 2.1648 15.3257 1633.3 1860.2 2.1330 12.2584 1633.2 1860.0 2.1083
1800 22.428 1727.6 1976.6 2.2187 16.8192 1727.5 1976.5 2.1869 13.4541 1727.3 1976.3 2.1622
2000 24.417 1825.2 2096.3 2.2694 18.3117 1825.0 2096.1 2.2376 14.6487 1824.9 2096.0 2.2130
Sat. 3.7289 1107.9 1190.8 1.5883 3.2202 1109.9 1193.4 1.5757 2.8347 1111.6 1195.5 1.5647
360 3.8446 1116.7 1202.1 1.6023 3.2584 1113.4 1197.8 1.5811
400 4.0799 1134.0 1224.6 1.6292 3.4676 1131.5 1221.4 1.6092 3.0076 1129.0 1218.0 1.5914
450 4.3613 1154.5 1251.4 1.6594 3.7147 1152.6 1248.9 1.6403 3.2293 1150.7 1246.3 1.6234
500 4.6340 1174.4 1277.3 1.6872 3.9525 1172.9 1275.3 1.6686 3.4412 1171.4 1273.2 1.6522
550 4.9010 1193.9 1302.8 1.7131 4.1845 1192.7 1301.1 1.6948 3.6469 1191.4 1299.4 1.6788
600 5.1642 1213.4 1328.0 1.7375 4.4124 1212.3 1326.6 1.7195 3.8484 1211.3 1325.2 1.7037
700 5.6829 1252.2 1378.4 1.7829 4.8604 1251.4 1377.3 1.7652 4.2434 1250.6 1376.3 1.7498
800 6.1950 1291.4 1429.0 1.8247 5.3017 1290.8 1428.1 1.8072 4.6316 1290.2 1427.3 1.7920
1000 7.2083 1371.7 1531.8 1.9005 6.1732 1371.3 1531.3 1.8832 5.3968 1370.9 1530.7 1.8682
1200 8.2137 1455.3 1637.7 1.9684 7.0367 1455.0 1637.3 1.9512 6.1540 1454.7 1636.9 1.9363
1400 9.2149 1542.3 1746.9 2.0305 7.8961 1542.1 1746.6 2.0134 6.9070 1541.8 1746.3 1.9986
1600 10.2135 1633.0 1859.8 2.0881 8.7529 1632.8 1859.5 2.0711 7.6574 1632.6 1859.3 2.0563
1800 11.2106 1727.2 1976.1 2.1420 9.6082 1727.0 1975.9 2.1250 8.4063 1726.9 1975.7 2.1102
2000 12.2067 1824.8 2095.8 2.1928 10.4624 1824.6 2095.7 2.1758 9.1542 1824.5 2095.5 2.1610
Sat. 2.5322 1113.0 1197.3 1.5548 2.2882 1114.1 1198.8 1.5460 2.0423 1115.3 1200.3 1.5360
400 2.6490 1126.3 1214.5 1.5752 2.3615 1123.5 1210.9 1.5602 2.0728 1119.7 1206.0 1.5427
450 2.8514 1148.7 1243.7 1.6082 2.5488 1146.7 1241.0 1.5943 2.2457 1144.1 1237.6 1.5783
500 3.0433 1169.8 1271.2 1.6376 2.7247 1168.2 1269.0 1.6243 2.4059 1166.2 1266.3 1.6091
550 3.2286 1190.2 1297.7 1.6646 2.8939 1188.9 1296.0 1.6516 2.5590 1187.2 1293.8 1.6370
600 3.4097 1210.2 1323.8 1.6897 3.0586 1209.1 1322.3 1.6771 2.7075 1207.7 1320.5 1.6628
700 3.7635 1249.8 1375.2 1.7361 3.3796 1249.0 1374.1 1.7238 2.9956 1248.0 1372.7 1.7099
800 4.1104 1289.5 1426.5 1.7785 3.6934 1288.9 1425.6 1.7664 3.2765 1288.1 1424.5 1.7528
900 4.4531 1329.7 1478.0 1.8179 4.0031 1329.2 1477.3 1.8059 3.5530 1328.5 1476.5 1.7925
1000 4.7929 1370.5 1530.1 1.8549 4.3099 1370.1 1529.6 1.8430 3.8268 1369.5 1528.9 1.8296
1200 5.4674 1454.3 1636.5 1.9231 4.9182 1454.0 1636.1 1.9113 4.3689 1453.6 1635.6 1.8981
1400 6.1377 1541.6 1746.0 1.9855 5.5222 1541.4 1745.7 1.9737 4.9068 1541.1 1745.4 1.9606
1600 6.8054 1632.4 1859.1 2.0432 6.1238 1632.2 1858.8 2.0315 5.4422 1632.0 1858.6 2.0184
1800 7.4716 1726.7 1975.6 2.0971 6.7238 1726.5 1975.4 2.0855 5.9760 1726.4 1975.2 2.0724
2000 8.1367 1824.4 2095.4 2.1479 7.3227 1824.3 2095.3 2.1363 6.5087 1824.1 2095.1 2.1232
Sat. 1.8440 1116.3 1201.6 1.5270 1.6806 1117.0 1202.6 1.5187 1.5435 1117.7 1203.3 1.5111
450 2.0027 1141.3 1234.0 1.5636 1.8034 1138.5 1230.3 1.5499 1.6369 1135.6 1226.4 1.5369
500 2.1506 1164.1 1263.6 1.5953 1.9415 1162.0 1260.8 1.5825 1.7670 1159.8 1257.9 1.5706
550 2.2910 1185.6 1291.5 1.6237 2.0715 1183.9 1289.3 1.6115 1.8885 1182.1 1287.0 1.6001
600 2.4264 1206.3 1318.6 1.6499 2.1964 1204.9 1316.7 1.6380 2.0046 1203.5 1314.8 1.6270
650 2.5586 1226.8 1345.1 1.6743 2.3179 1225.6 1343.5 1.6627 2.1172 1224.4 1341.9 1.6520
700 2.6883 1247.0 1371.4 1.6974 2.4369 1246.0 1370.0 1.6860 2.2273 1244.9 1368.6 1.6755
800 2.9429 1287.3 1423.5 1.7406 2.6699 1286.5 1422.4 1.7294 2.4424 1285.7 1421.3 1.7192
900 3.1930 1327.9 1475.6 1.7804 2.8984 1327.3 1474.8 1.7694 2.6529 1326.6 1473.9 1.7593
1000 3.4403 1369.0 1528.2 1.8177 3.1241 1368.5 1527.4 1.8068 2.8605 1367.9 1526.7 1.7968
1200 3.9295 1453.3 1635.0 1.8863 3.5700 1452.9 1634.5 1.8755 3.2704 1452.5 1634.0 1.8657
1400 4.4144 1540.8 1745.0 1.9488 4.0116 1540.5 1744.6 1.9381 3.6759 1540.2 1744.2 1.9284
1600 4.8969 1631.7 1858.3 2.0066 4.4507 1631.5 1858.0 1.9960 4.0789 1631.3 1857.7 1.9863
1800 5.3777 1726.2 1974.9 2.0607 4.8882 1726.0 1974.7 2.0501 4.4803 1725.8 1974.5 2.0404
2000 5.8575 1823.9 2094.9 2.1116 5.3247 1823.8 2094.7 2.1010 4.8807 1823.6 2094.6 2.0913
a The temperature in parentheses is the saturation temperature at the specified pressure.
b Properties of saturated vapor at the specified pressure.

495
TABLE C.20
Enthalpy of Formation, Gibbs of Formation, and Absolute Entropy of Various Substances
at 25°C (77°F) and 1 atm
hfo g fo so
Substance Formula kJ/kmol Btu/lbmol kJ/kmol Btu/lbmol kJ/kmol K Btu/lbmol °R

Carbon C(s) 0 0 0 0 5.74 1.36
Hydrogen H2(g) 0 0 0 0 130.57 31.21
Nitrogen N2(g) 0 0 0 0 191.5 45.77
Oxygen O2(g) 0 0 0 0 205.03 49
Carbon monoxide CO(g) −110,530 −47,540 −137,150 −59,010 197.54 47.21
Carbon dioxide CO2(g) −393,520 −169,300 −394,380 −169,680 213.69 51.07
Water H2O(g) −241,820 −104,040 −228,590 −98,350 188.72 45.11
Water H2O(l) −285,830 −122,970 −237,180 −102,040 69.95 16.71
Hydrogen peroxide H2O2(g) −136,310 −58,640 −105,600 −45,430 232.63 55.6
Ammonia NH3(g) −46,190 −19,750 −16,590 −7,140 192.33 45.97
Oxygen O(g) 249,170 107,210 231,770 99,710 160.95 38.47
Hydrogen H(g) 218,000 93,780 203,290 87,460 114.61 27.39
Nitrogen N(g) 472,680 203,340 455,510 195,970 153.19 36.61
Hydroxyl OH(g) 39,460 16,790 34,280 14,750 183.75 43.92
Methane CH4(g) −74,850 −32,210 −50,790 −21,860 186.16 44.49
Acetylene C2H2(g) 226,730 97,540 209,170 87,990 200.85 48
Ethylene C2H4(g) 52,280 22,490 68,120 29,306 219.83 52.54
Ethane C2H6(g) −84,680 −36,420 −32,890 −14150 229.49 54.85
Propylene C3H6(g) 20,410 8,790 62,720 26,980 266.94 63.8
Propane C3H8(g) −103,850 −44,680 −23,490 −10,105 269.91 64.51
Butane C4H10(g) −126,150 −54,270 −15,710 −6,760 310.03 74.11
Octane C8H18(g) −208,450 −89,680 17,320 7,110 463.67 111.55
Octane C8H18(l) −249,910 −107,530 6,610 2,840 360.79 86.23
Dodecane C12H26(g) −291,010 −125,190 50,150 21,570 622.83 148.86
Benzene C6H6(g) 82,930 35,680 129,660 55,780 269.2 64.34
Methyl alcohol CH3OH(g) −200,890 −86,540 −162,140 −69,700 239.7 57.29
Methyl alcohol CH3OH(l) −238,810 −102,670 −166,290 −71,570 126.8 30.3
Ethyl alcohol C2H5OH(g) −235,310 −101,230 −168,570 −72,520 282.59 67.54
Ethyl alcohol C2H5OH(l) −277,690 −119,470 174,890 −75,240 160.7 38.4
Sources: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001, Table B.2, p. 716;
Cengel, Y. and Boles, M., Thermodynamics: An Engineering Approach, 7th edn., McGraw-Hill, New York, 2011,
Table A-26E, p. 997.
TABLE C.21
Combustion Data for Hydrocarbons (Metric and English Units)
Flammability
Higher Heating Max Flame Adiabatic Flame Limits (in Air)
Value (Vapor) Speed Temp (in Air) Ignition Temp (in Air) Flash Point (% by volume)
Theor. Air/Fuel
Hydrocarbon Formula kJ/kg Btu/lbm Ratio (by mass) m/s ft/s °C °F °C °F °C °F LFL UFL
Paraffins or alkanes
Methane CH4 55,533 23,875 17.195 0.34 1.1 1,918 3,484 705 1,301 Gas Gas 5.0 15.0
Ethane C2H6 51,923 22,323 15.899 0.40 1.3 1,949 3,540 520–630 968–1,166 Gas Gas 3.0 12.5
Propane C3H8 50,402 21,669 15.246 0.40 1.3 1,967 3,573 466 871 Gas Gas 2.1 10.1
n-Butane C4H10 49,593 21,321 14.984 0.37 1.2 1,973 3,583 405 761 −60 −76 1.86 8.41
iso-Butane C4H10 49,476 21,271 14.984 0.37 1.2 1,973 3,583 462 864 −83 −117 1.80 8.44
n-Pentane C5H12 49,067 21,095 15.323 0.40 1.3 2,232 4,050 309 588 <−40 <−40 1.40 7.80
iso-Pentane C5H12 48,955 21,047 15.323 0.37 1.2 2,235 4,055 420 788 <−51 <−60 1.32 9.16
Neopentane C5H12 48,795 20,978 15.323 0.34 1.1 2,238 4,060 450 842 Gas Gas 1.38 7.22
n-Hexane C6H14 48,767 20,966 15.238 0.40 1.3 2,221 4,030 248 478 −22 −7 1.25 7.00
Neohexane C6H14 48,686 20,931 15.238 0.37 1.2 2,235 4,055 425 797 −48 −54 1.19 7.58
n-Heptane C7H16 48,506 20,854 15.141 0.40 1.3 2,196 3,985 223 433 −4 25 1.00 6.00
Triptane C7H16 48,437 20,824 15.141 0.37 1.2 2,224 4,035 454 849 — — 1.08 6.69
n-Octane C8H18 48,371 20,796 15.093 — — — — 220 428 13 56 0.95 3.20
iso-Octane C8H18 48,311 20,770 15.093 0.34 1.1 — — 447 837 −12 10 0.76 5.94
Olefins or alkenes
Ethylene C2H4 50,325 21,636 14.807 0.67 2.2 2,343 4,250 490 914 Gas Gas 2.75 28.6
Propylene C3H6 48,958 21,048 14.807 0.43 1.4 2,254 4,090 458 856 gas gas 2.00 11.1
Butylene C4H8 48,506 20,854 14.807 0.43 1.4 2,221 4,030 443 829 Gas Gas 1.98 9.65
iso-Butene C4H8 48,234 20,737 14.807 0.37 1.2 — — 465 869 Gas Gas 1.80 9.00
n-Pentene C5H10 48,195 20,720 14.807 0.43 1.4 2,296 4,165 298 569 — — 1.65 7.70
Aromatics
Benzene C6H6 42,296 18,184 13.297 0.40 1.3 2,266 4,110 562 1,044 −11 12 1.35 6.65
Toluene C7H8 43,033 18,501 13.503 0.37 1.2 2,232 4,050 536 997 4 40 1.27 6.75
p-Xylene C8H10 43,410 18,663 13.663 — — 2,210 4,010 464 867 17 63 1.00 6.00
Other hydrocarbons
Acetylene C2H2 50,014 21,502 13.297 1.40 4.6 2,632 4,770 406–440 763–824 Gas Gas 2.50 81.0
Naphthalene C10H8 40,247 17,303 12.932 — — 2,260 4,100 515 959 79 174 0.90 5.9
Source: Adapted from Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001, Table B.1, p. 715.
497
TABLE C.22
Chemical, Physical, and Thermal Properties of Gases: Gases and Vapors, Including Fuels and Refrigerants,
English and Metric Units
Common Name(s) Acetylene (Ethyne) Butadiene n-Butane Isobutane (2-Methyl
Propane)
Chemical Formula C2H2 C4H6 C4H10 C4H10
Refrigerant Number — — 600 600a
Chemical and physical properties
Molecular weight 26.04 54.09 58.12 58.12
Specific gravity, air = 1 0.90 1.87 2.07 2.07
Specific volume, ft3/lb 14.9 7.1 6.5 6.5
Specific volume, m3/kg 0.93 0.44 0.405 0.418
Density of liquid (at atm bp), lb/ft3 43.0 37.5 37.2
Density of liquid (at atm bp), kg/m3 693.0 604.0 599.0
Vapor pressure at 25°C, psia 35.4 50.4
Vapor pressure at 25°C, MN/m2 0.0244 0.347
Viscosity (abs), lbm/ft · s 6.72 × 10−6 4.8 × 10−6
Viscosity (abs), centipoisesa 0.01 0.007
Sound velocity in gas, m/s 343 226 216 216
Thermal and thermodynamic properties
Specific heat, cp, Btu/lb · °F or cal/g · °C 0.40 0.341 0.39 0.39
Specific heat, cp, J/kg · K 1,674.0 1,427.0 1,675.0 1,630.0
Specific heat ratio, cp/cv 1.25 1.12 1.096 1.10
Gas constant R, ft lb/lb · °R 59.3 28.55 26.56 26.56
Gas constant R, J/kg · °C 319 154.0 143.0 143.0
Thermal conductivity, Btu/h · ft · °F 0.014 0.01 0.01
Thermal conductivity, W/m °C 0.024 0.017 0.017
Boiling point (sat 14.7 psia), °F −103 24.1 31.2 10.8
Boiling point (sat 760 mm), °C −75 −4.5 −0.4 −11.8
Latent heat of evap. (at bp), Btu/lb 264 165.6 157.5
Latent heat of evap. (at bp), J/kg 614 000 386 000 366 000
Freezing (melting) point, °F (1 atm) −116 −164.0 −217.0 −229
Freezing (melting) point, °C (1 atm) −82.2 −109.0 −138 −145
Latent heat of fusion, Btu/lb 23.0 19.2
Latent heat of fusion, J/kg 53 500 44 700
Critical temperature, °F 97.1 306 273.0
Critical temperature, °C 36.2 171.0 152.0 134.0
Critical pressure, psia 907.0 652.0 550.0 537.0
Critical pressure, MN/m2 6.25 3.8 3.7
Critical volume, ft3/lb 0.070
Critical volume, m3/kg 0.004 3
Flammable (yes or no) Yes Yes Yes Yes
Heat of combustion, Btu/ft3 1,450 2,950 3,300 3,300
Heat of combustion, Btu/lb 21,600 20,900 21,400 21,400
Heat of combustion, kJ/kg 50,200 48,600 49,700 49,700

Common Name(s) 1-Butene (Butylene) cis-2-Butene trans-2-Butene Isobutene
Chemical Formula C4H8 C4H8 C4H8 C4H8
Refrigerant Number — — — —
Density of liquid (at atm bp), lb/ft3
Density of liquid (at atm bp), kg/m3
Vapor pressure at 25°C, psia
Vapor pressure at 25°C, MN/m2
Viscosity (abs), lbm/ft · s
Viscosity (abs), centipoisesa
Sound velocity in gas, m/s 222 223.0 221.0 221.0
Specific heat, cp, Btu/lb · °F or cal/g °C 0.36 0.327 0.365 0.37
Specific heat, cp, J/kg K 1,505.0 1,368.0 1,527.0 1,548.0
Gas constant R, ft lb/lb · °F 27.52
Gas constant R, J/kg · °C 148.0
Thermal conductivity, Btu/h ft °F
Thermal conductivity, W/m °C
Boiling point (sat 14.7 psia), °F 20.6 38.6 33.6 19.2
Boiling point (sat 760 mm), °C −6.3 3.7 0.9 −7.1
Latent heat of evap. (at bp), Btu/lb 167.9 178.9 174.4 169.0
Latent heat of evap. (at bp), J/kg 391,000 416, 000.0 406, 000.0 393, 000.0
Freezing (melting) point, °F (1 atm) −301.6 −218.0 −158.0
Freezing (melting) point, °C (1 atm) −185.3 −138.9 −105.5
Latent heat of fusion, Btu/lb 16.4 31.2 41.6 25.3
Latent heat of fusion, J/kg 38, 100 72, 600.0 96, 800.0 58, 800.0
Critical temperature, °F 291.0
Critical temperature, °C 144.0 160.0 155.0
Critical pressure, psia 621.0 595.0 610.0
Critical pressure, MN/m2 4.28 4.10 4.20
Critical volume, ft3/lb 0.068
Critical volume, m3/kg 0.0042
Flammable (yes or no) Yes Yes Yes Yes
Heat of combustion, Btu/ft3 3,150.0 3,150.0 3,150.0 3,150.0
Heat of combustion, Btu/lb 21, 000.0 21, 000.0 21, 000.0 21, 000.0
Heat of combustion, kJ/kg 48, 800.0 48, 800.0 48, 800.0 48, 800.0
(continued)

Common Name(s) Carbon Dioxide Carbon Monoxide Ethane Ethylene (Ethene)
Chemical Formula CO2 CO C2H6 C2H4
Refrigerant Number 744 — 170 1150
Density of liquid (at atm bp), lb/ft3 — 28.0 35.5
Density of liquid (at atm bp), kg/m3 — 449.0 569.0
Vapor pressure at 25°C, psia 931.0
Vapor pressure at 25°C, MN/m2 6.42
Viscosity (abs), lbm/ft · s 9.4 × 10−6 12.1 × 10−6 64. × 10−6 6.72 × 10−6
Viscosity (abs), centipoisesa 0.014 0.018 0.095 0.010
Sound velocity in gas, m/s 270.0 352.0 316.0 331.0
Specific heat, cp, Btu/lb °F or cal/g °C 0.205 0.25 0.41 0.37
Specific heat, cp, J/kg K 876.0 1,046.0 1,715.0 1,548.0
Gas constant R, ft lb/lb · °F 35.1 55.2 51.4 55.1
Gas constant R, J/kg · °C 189.0 297.0 276.0 296.0
Thermal conductivity, Btu/h ft °F 0.01 0.014 0.010 0.010
Thermal conductivity, W/m °C 0.017 0.024 0.017 0.017
Boiling point (sat 14.7 psia), °F −109.4b −312.7 −127.0 −155.0
Boiling point (sat 760 mm), °C −78.5 −191.5 −88.3 −103.8
Latent heat of evap. (at bp), J/kg 572, 000.0 216, 000.0 488, 000.0 484, 000.0
Freezing (melting) point, °F (1 atm) −337.0 −278.0 −272.0
Freezing (melting) point, °C (1 atm) −205.0 −172.2 −169.0
Latent heat of fusion, Btu/lb — 12.8 41.0 51.5
Latent heat of fusion, J/kg — 95, 300.0 120, 000.0
Critical temperature, °F 88.0 −220.0 90.1 49.0
Critical temperature, °C 31.0 −140.0 32.2 9.5
Critical pressure, psia 1,072.0 507.0 709.0 741.0
Critical pressure, MN/m2 7.4 3.49 4.89 5.11
Critical volume, ft3/lb 0.053 0.076 0.073
Critical volume, m3/kg 0.0033 0.0047 0.0046
Flammable (yes or no) No Yes Yes Yes
Heat of combustion, Btu/ft3 — 310.0 1,480.0
Heat of combustion, Btu/lb — 4,340.0 22,300.0 20, 600.0
Heat of combustion, kJ/kg — 10, 100.0 51, 800.0 47, 800.0

Common Name(s) Hydrogen Methane Nitric Oxide Nitrogen
Chemical Formula H2 CH4 NO N2
Refrigerant Number 702 50 — 728
Density of liquid (at atm bp), lb/ft3 4.43 26.3 50.46
Density of liquid (at atm bp), kg/m3 71.0 421.0 808.4
Vapor pressure at 25°C, psia
Vapor pressure at 25°C, MN/m2
Viscosity (abs), lbm/ft · s 6.05 × 10−6 7.39 × 10−6 12.8 × 10−6 12.1 × 10−6
Sound velocity in gas, m/s 1 315.0 446.0 341.0 353.0
Specific heat, cp, J/kg · K 14, 310.0 2,260.0 983.0 1040.0
Gas constant R, J/kg · °C 4,126.0 518.0 277.0 297.0
Thermal conductivity, Btu/h · ft · °F 0.105 0.02 0.015 0.015
Thermal conductivity, W/m · °C 0.0182 0.035 0.026 0.026
Boiling point (sat 14.7 psia), °F −423.0 −259.0 −240.0 −320.4
Boiling point (sat 760 mm), °C 20.4 K −434.2 −151.5 −195.8
Latent heat of evap. (at bp), Btu/lb 192.0 219.2 85.5
Latent heat of evap. (at bp), J/kg 447, 000.0 510, 000.0 199, 000.0
Freezing (melting) point, °F (1 atm) −434.6 −296.6 −258.0 −346.0
Freezing (melting) point, °C (1 atm) −259.1 −182.6 −161.0 −210.0
Latent heat of fusion, Btu/lb 25.0 14.0 32.9 11.1
Latent heat of fusion, J/kg 58, 000.0 32, 600.0 76, 500.0 25, 800.0
Critical temperature, °F −399.8 −116.0 −136.0 −232.6
Critical temperature, °C −240.0 −82.3 −93.3 −147.0
Critical pressure, psia 189.0 673.0 945.0 493.0
Critical volume, ft3/lb 0.53 0.099 0.0332 0.051
Critical volume, m3/kg 0.033 0.0062 0.00207 0.00318
Flammable (yes or no) Yes Yes No No
Heat of combustion, Btu/ft3 320.0 985.0 — —
Heat of combustion, Btu/lb 62, 050.0 2 2,900.0 — —
Heat of combustion, kJ/kg 144, 000.0 — —
(continued)

Common Name(s) Nitrous Oxide Oxygen Propane Propylene (Propene)
Chemical Formula N2O O2 C3H6
Refrigerant Number 744A 732 290 1270
Molecular weight 44.012 31.998 8 44.097 42.08
Density of liquid (at atm bp), lb/ft3 76.6 71.27 36.2 37.5
Density of liquid (at atm bp), kg/m3 1,227.0 1,142.0 580.0 601.0
Vapor pressure at 25°C, psia 135.7 166.4
Vapor pressure at 25°C, MN/m2 0.936 1.147
Viscosity (abs), lbm/ft · s 10.1 × 10−6 13.4 × 10−6 53.8×10−6 57.1 × 10−6
Sound velocity in gas, m/s 268.0 329.0 253.0 261.0
Specific heat, cp, J/kg · K 879.0 920.0 1,630.0 1,506.0
Gas constant R, J/kg · °C 189.0 260.0 188.0 197.0
Thermal conductivity, Btu/h · ft · °F 0.010 0.015 0.010 0.010
Thermal conductivity, W/m °C 0.017 0.026 0.017 0.017
Boiling point (sat 14.7 psia), °F −127.3 −297.3 −44.0 −54.0
Boiling point (sat 760 mm), °C −88.5 −182.97 −42.2 −48.3
Latent heat of evap. (at bp), J/kg 376, 000.0 213, 000.0 428, 000.0 438, 000.0
Freezing (melting) point, °F (1 atm) −131.5 −361.1 −309.8 −301.0
Freezing (melting) point, °C (1 atm) −90.8 −218.4 −189.9 −185.0
Latent heat of fusion, Btu/lb 63.9 5.9 19.1
Latent heat of fusion, J/kg 149, 000.0 13, 700.0 44, 400.0
Critical temperature, °F 97.7 −181.5 205.0 197.0
Critical temperature, °C 36.5 −118.6 96.0 91.7
Critical pressure, psia 1,052.0 726.0 618.0 668.0
Critical volume, ft3/lb 0.036 0.040 0.073 0.069
Critical volume, m3/kg 0.0022 0.0025 0.0045 0.0043
Flammable (yes or no) No No Yes Yes
Heat of combustion, Btu/ft3 — — 2,450.0 2,310.0
Heat of combustion, Btu/lb — — 21, 660.0 21, 500.0
Heat of combustion, kJ/kg — — 50, 340.0 50, 000.0
Sources: Baukal, C.E., The CRC Press Handbook of Thermal Engineering, CRC Press, Boca Raton, FL, 2000; Baukal, C.E. (ed.), The John Zink
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
Note: The properties of pure gases are given at 25°C (77°F, 298 K) and atmospheric pressure (expect as stated).
a For N · s / m2 divide by 1000.
TABLE C.23
Burning Velocities of Various Fuels
ϕ = 0.7 ϕ = 0.8 ϕ = 0.9 ϕ = 1.0 ϕ = 1.1 ϕ = 1.2 ϕ = 1.3 ϕ = 1.4 Smax ϕ at Smax
Saturated hydrocarbons
Ethane 30.6 36.0 40.6 44.5 47.3 47.3 44.4 37.4 47.6 1.14
Propane 42.3 45.6 46.2 42.4 34.3 46.4 1.06
n-Butane 38.0 42.6 44.8 44.2 41.2 34.4 25.0 44.9 1.03
Methane 30.0 38.3 43.4 44.7 39.8 31.2 44.8 1.08
n-Pentane 35.0 40.5 42.7 42.7 39.3 33.9 43.0 1.05
n-Heptane 37.0 39.8 42.2 42.0 35.5 29.4 42.8 1.05
2,2,4-Trimethylpentane 37.5 40.2 41.0 37.2 31.0 23.5 41.0 0.98
2,2,3-Trimethylpentane 37.8 39.5 40.1 39.5 36.2 40.1 1.00
2,2-Dimethylbutane 33.5 38.3 39.9 37.0 33.5 40.0 0.98
Isopentane 33.0 37.6 39.8 38.4 33.4 24.8 39.9 1.01
2,2-Dimethylpropane 31.0 34.8 36.0 35.2 33.5 31.2 36.0 1.10
Unsaturated hydrocarbons
Acetylene 107 130 144 151 154 154 152 155 1.25
Ethylene 37.0 50.0 60.0 68.0 73.0 72.0 66.5 60.0 73.5 1.13
Propyne 62.0 66.6 70.2 72.2 71.2 61.0 72.5 1.14
1,3-Butadiene 42.6 49.6 55.0 57.0 56.9 55.4 57.2 1.23
n-1-Heptyne 46.8 50.7 52.3 50.9 47.4 41.6 52.3 1.00
Propylene 48.4 51.2 49.9 46.4 40.8 51.2 1.00
n-2-Pentene 35.1 42.6 47.8 46.9 42.6 34.9 48.0 1.03
2,2,4-Trimethyl-3-pentene 34.6 41.3 42.2 37.4 33.0 42.5 0.98
Substituted alkyls
Methanol 34.5 42.0 48.0 50.2 47.5 44.4 42.2 50.4 1.08
Isopropyl alcohol 34.4 39.2 41.3 40.6 38.2 36.0 34.2 41.4 1.04
Triethylamine 32.5 36.7 38.5 38.7 36.2 28.6 38.8 1 06
n-Butyl chloride 24.0 30.7 33.8 34.5 32.5 26.9 20.0 34.5 1.00
Allyl chloride 30.6 33.0 33.7 32.4 29.6 33.8 0.89
Isopropyl mercaptan 30.0 33.5 33.0 26.6 33.8 0.44
Ethylamine 28.7 31.4 32.4 31.8 29.4 25.3 32.4 1.00
Isopropylamine 27.0 29.5 30.6 29.8 27.7 30.6 1 01
n-Propyl chloride 24.7 28.3 27.5 24.1 28.5 0.93
Isopropyl chloride 24.8 27.0 27.4 25.3 27.6 0.97
n-Propyl bromide No ignition
Silanes
Tetramethylsilane 39.5 49.5 57.3 58.2 57.7 54.5 47.5 58.2 1.01
Trimethylethoxysilane 34.7 41.0 47.4 50.3 46.5 41.0 35.0 50.3 1.00
Aldehydes
Acrolein 47.0 58.0 66.6 65.9 56.5 67.2 0.95
Propionaldehyde 37.5 44.3 49.0 49.5 46.0 41.6 37.2 50.0 1.06
Acetaldehyde 26.6 35.0 41.4 41.4 36.0 30.0 42.2 1.05
Ketones
Acetone 40.4 44.2 42.6 38.2 44.4 0.93
Methyl ethyl ketone 36.0 42.0 43.3 41.5 37.7 33.2 43.4 0.99
Esters
Vinyl acetate 29.0 36.6 39.8 41.4 42.1 41.6 35.2 42.2 1.13
Ethyl acetate 30.7 35.2 37.0 35.6 30.0 37.0 1.00
Ethers
Dimethyl ether 44.8 47.6 48.4 47.5 45.4 42.6 48.6 0.99
Diethyl ether 30.6 37.0 43.4 48.0 47.6 40.4 32.0 48.2 1.05
Dimethoxymethane 32.5 38.2 43.2 46.6 48.0 46.6 43.3 48.0 1.10
Diisopropyl ether 30.7 35.5 38.3 38.6 36.0 31.2 38.9 1.06
(continued)

Burning Velocities of Various Fuels
ϕ = 0.7 ϕ = 0.8 ϕ = 0.9 ϕ = 1.0 ϕ = 1.1 ϕ = 1.2 ϕ = 1.3 ϕ = 1.4 Smax ϕ at Smax
Thio ethers
Dimethyl sulfide 29.9 31.9 33.0 30.1 24.8 33.0 1.00
Peroxides
Di-t-butyl peroxide 41.0 46.8 50.0 49.6 46.5 42.0 35.5 50.4 1.04
Aromatic compounds
Furan 48.0 55.0 60.0 62.5 62.4 60.0 62.9 1.05
Benzene 39.4 45.6 47.6 44.8 40.2 35.6 47.6 1.00
Thiophane 33.8 37.4 40.6 43.0 42.2 37.2 24.6 43.2 1.03
Cyclic compounds
Ethylene oxide 57.2 70.7 83.0 88.8 89.5 87.2 81.0 73.0 89.5 1.07
Butadiene monoxide 36.6 47.4 57.8 64.0 66.9 66.8 64.5 67.1 1.24
Propylene oxide 41.6 53.3 62.6 66.5 66.4 62.5 53.8 67.0 1.05
Dihydropyran 39.0 45.7 51.0 54.5 55.6 52.6 44.3 32.0 55.7 1.08
Cyclopropane 40.6 49.0 54.2 55.6 53.5 44.0 55.6 1.10
Tetrahydropyran 44.8 51.0 53.6 51.5 42.3 53.7 0.93
Cyclic compounds
Tetrahydrofuran 43.2 48.0 50.8 51.6 49.2 44.0 51.6 1.19
Cyclopentadiene 36.0 41.8 45.7 47.2 45.5 40.6 32.0 47.2 1.00
Ethylenimine 37.6 43.4 46.0 45.8 43.4 38.9 46.4 1.04
Cyclopentane 31.0 38.4 43.2 45.3 44.6 41.0 34.0 45.4 1.03
Cyclohexane 41.3 43.5 43.9 38.0 44.0 1.08
Inorganic compounds
Hydrogen 102 120 145 170 204 245 213 290 325 1.80
Carbon disulfide 50.6 58.0 59.4 58.8 57.0 55.0 52.8 51.6 59.4 0.91
Carbon monoxide 28.5 32.0 34.8 38.0 52.0 2.05
Hydrogen sulfide 34.8 39.2 40.9 39.1 32.3 40.9 0.90
Propylene oxide 74.0 86.2 93.0 96.6 97.8 94.0 84.0 71.5 97.9 1.09
Hydrazine 87.3 90.5 93.2 94.3 93.0 90.7 87.4 83.7 94.4 0.98
Furfural 62.0 73.0 83.3 87.0 87.0 84.0 77.0 65.5 87.3 1.05
Ethyl nitrate 70.2 77.3 84.0 86.4 83.0 72.3 86.4 1.00
Butadiene monoxide 51.4 57.0 64.5 73.0 79.3 81.0 80.4 76.7 81.1 1.23
Carbon disulfide 64.0 72.5 76.8 78.4 75.5 71.0 66.0 62.2 78.4 1.00
n-Butyl ether 67.0 72.6 70.3 65.0 72.7 0.91
Methanol 50.0 58.5 66.9 71.2 72.0 66.4 58.0 48.8 72.2 1.08
Diethyl cellosolve 49.5 56.0 63.0 69.0 69.7 65.2 70.4 1.05
Cyclohexene
Monoxide 54.5 59.0 63.5 67.7 70.0 64.0 70.0 1.10
Epichlorohydrin 53.0 59.5 65.0 68.6 70.0 66.0 58.2 70.0 1.10
n-Pentane 50.0 55.0 61.0 62.0 57.0 49.3 42.4 62.9 1.05
n-Propyl alcohol 49.0 56.6 62.0 64.6 63.0 50.0 37.4 64.8 1.03
n-Heptane 41.5 50.0 58.5 63.8 59.5 53.8 46.2 38.8 63.8 1.00
Ethyl nitrite 54.0 58.8 62.6 63.5 59.0 49.5 42.0 36.7 63.5 1.00
Pinene 48.5 58.3 62.5 62.1 56.6 50.0 63.0 0.95
Nitroethane 51.5 57.8 61.4 57.2 46.0 28.0 61.4 0.92
Isooctane 50.2 56.8 57.8 53.3 50.5 58.2 0.98
Pyrrole 52.0 55.6 56.6 56.1 52.8 48.0 43.1 56.7 1.00
Aniline 41.5 45.4 46.6 42.9 37.7 32.0 46.8 0.98
Dimethyl formamide 40.0 43.6 45.8 45.5 40.7 36.7 46.1 1.04
Source: Annamalai, K. and Puri, I.K., Combustion Science and Engineering, CRC Press, Boca Raton, FL, 2007.
Note: T = 25°C (air−fuel temperature); P = 1 atm (0.31 mol % H2O in air); burning velocity S as a function of equivalence ratio ϕ in cm/s. The data
are for premixed fuel−air mixtures at 100°C and 1 atm pressure; 0.31 mol % H2O in air; burning velocity S as a function of ϕ in cm/s.
0.20 0.25 0.30 0.35 0.40
Sensible heat ratio = Qs/Qt
34.0
10
0
33.0
28
30 0.45
30 32.0
0.94
Psychrometric chart 31.0
90 26
Normal temperature 30.0
SI units 29.0
Sea level 0.50 13
30 24 28.0 0
we
Barometric pressure: 101.325 kPa 80 tb
ul 27.0
b
te
m
1.0 – pe
• 1.0 • ra 22 26.0
tu
0.92
re
-° 0.55
C 25.0
11.7 0.8 70
Sensible heat Qs 25 24.0 12
2.0 –4.6 25 0
Total heat Qt 20
4 –2.3 23.0
7.0 0.6 8 .0 0.60
–8•– .0 22.0
0.5 .0
–
.0
60 18
0.4
–4 2.0
21.0
–1
0.0
4.
.0
25
0.65
7
0.3
–0
air
0.2
–00..4
20.0
– .3
–0.2
0.1
–0.1
0.90 v
0
dry
o
11
1.2
0
of
16 19.0
3.5
0.70
lume
50
gra
18.0
2.3
20
ilo
- °C
- cu
20
rk
re
Enthalpy Dh 17.0
0.75
atu
j pe
14
bic m
Humidity ratio DW
-k
per
y
16.0
em
eter p
alp
0.80
nt
40
th
0.88
15.0 10
tio
0
En
90%
Dew point temperature - °C
er kg
20
d
12
ura
0.85
Vapor pressure - mm of mercury
14.0
Sat
%
ry air
80
15 0.90 13.0
15
30
10 0.95 12.0
%
%
70
25
0.86 1.00 11.0

Enthalpy - kj per kilogram of dry air
2005 90
15
%
Summer
10.0
60
10 comfort 8
20 2005 10
Winter 9.0
% comfort
50 8.0
0.84
10 6
10 5 % 7.0
4 0% 15 5
6.0 80
0
0.82
5 % 5.0
0 30 idity 4
–1 hum 0
0 tive
Humidity ratio - grams moisture per kilogram dry air
rela 4.0
8%
0.80
–5 20%
–2 6% –5 3.0
0 y
–10 umidit 2
0.78
–5 lative h 4% –10 2.0
10% re
–3 –15
0 –1 70
0.76
–20 0 2% 1.0
–25 –1
0.74
5
–30 –25 –20 –15 –10 –5 0 5 10 15 20 25 30 35 40 45 50 55
Chart by: Hands down software, www.handsdownsoftware.com Dry bulb temperature - °C
–3 –2 –1 0 10 20 30 40 50 60
(a) 0 0 0
FIGURE C.1
Psychrometric chart. (a) SI units.
505
(continued)
506
0.20 0.25 0.30 0.35 0.40

55 60
1.35
200
50 1.30
90
190 0.45 1.25

85
85
15
Psychrometric chart 1.20
180
Normal temperature
.0 volu
45
1.15
me
I-P units
170
- cu.F
Sea level 0.50
1.10
Barometric pressure: 29.921 in. of mercury
t. Per
85
we
tb 160 1.05
40 80 ul 80
1.0 1.0 –
lb. dry
• • bt
em
air
pe
ra 150 1.00
tu 0.55 60
0.8 re
-°
F
5000
Sensible heat Qs .95
2.0 –200 140
Total heat Qt 0
80 .90
0.6 4.0 –10
00
35
75 0.60
300
0 – 8. 75
–8• 0 130
0.5 .0 .85
.0
–2
–4 .0
0.4
0
–1
.0
0.3
20
120
–0
0.65 .80
00
0.2
–0.5.4
–0.3
0.1
–0.2
55
–0.1
0
air
30
50
70 75
.75
0
110 70
dry
150
0.70
0
of
°F
d
.70
e-
un
1000
o
tur
100
rp
Enthalpy Dh 0.75
era
pe
.65
p
Humidity ratio 65
tu
DW 70 65
tem
25
-B
n
y
90 0.80 .60
tio
alp
90%
Dew point temperature - °F
th
50
ura
En
0.85
Sat
.55
14.0
80
%
Vapor pressure - inches of mercury
60
60
80
65
0.90
.50
20
70 0.95
%
%
70
25
55 .45
Enthalpy - Btu per pound of dry air
55 1.00
60 2005
%
Summer 60 .40
60
comfort
50 45
2005 50
15 55 Winter 50 .35
5 0% comfort
45
45 .30
50
% % 40
40 40 15
10 40 .25
45
35 30
0% 35 .20
40 3 idity
5 30 hum
13.0
lative 30
Humidity ratio - grains of moisture per pound of dry air
40
35 8 % re .15
25 20
30 20% 25
0 20 6%
y 20 .10
15 25 umidit
10 lative h 4% 10
20 10% re 10
5 15
0 .05
–10 –5 5 10 2% 0
–15
12.0
0
–20 –15 –10 –5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130
Chart by: Hands down software, www.handsdownsoftware.com Dry bulb temperature - °F 35
(b) 0 5 10 15 20 25 30
FIGURE C.1 (continued)

Psychrometric chart. (b) English units. (From ASHRAE, Coolerado, or hands down software.)
5. Keenan, J.H. and Kaye, J., Gas Tables, Wiley, New York,
1945.
References 6. Kraus, A.D., Welty, J.R., and Aziz, A., Introduction to
1. Annamalai, K. and Puri, I.K., Combustion Science and Thermal and Fluid Engineering, CRC Press, Boca Raton,
Engineering, CRC Press, Boca Raton, FL, 2007. FL, 2012.
2. Baukal, C.E. (ed.), The John Zink Combustion Handbook, 7. Kreith, F. (ed.), The CRC Handbook of Thermal Engineering,
CRC Press, Boca Raton, FL, 2001. CRC Press, Boca Raton, FL, 2000.
3. Cengel, Y. and Boles, M., Thermodynamics: An Engineering 8. Moran, M.J. and Shapiro, H.N., Fundamentals of
Approach, 7th edn., McGraw-Hill, New York, 2011. Engineering Thermodynamics, 3rd edn., Wiley, New York,
4. Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC 1995.
Appendix D: Properties of Solids
TABLE D.1
Thermal Properties of Selected Metallic Elements at 293 K (20°C) or 528°R (65°F)
Melting
Specific Heat, cp Thermal Conductivity, k Diffusivity, α Temperature
Specific
Element Gravity J/(kg K) Btu/(lbm °R) W/(m K) Btu/(h ft °R) m2/s × 106 ft2/s × 103 K °R
Aluminum 2.702 896 0.214 236 136 97.5 1.05 933 1680
Beryllium 1.850 1750 0.418 205 118 63.3 0.681 1550 2790
Chromium 7.160 440 0.105 91.4 52.8 29.0 0.312 2118 3812
Copper 8.933 383 0.0915 399 231 116.6 1.26 1356 2441
Gold 19.300 129 0.0308 316 183 126.9 1.37 1336 2405
Iron 7.870 452 0.108 31.1 18.0 22.8 0.245 1810 3258
Lead 11.340 129 0.0308 35.3 20.4 24.1 0.259 601 1082
Magnesium 1.740 1017 0.243 156 90.1 88.2 0.949 923 1661
Manganese 7.290 486 0.116 7.78 4.50 2.2 0.0236 1517 2731
Molybdenum 10.240 251 0.0600 138 79.7 53.7 0.578 2883 5189
Nickel 8.900 446 0.107 91 52.6 22.9 0.246 1726 3107
Platinum 21.450 133 0.0318 71.4 41.2 25.0 0.269 2042 3676
Potassium 0.860 741 0.177 103 59.6 161.6 1.74 337 607
Silicon 2.330 703 0.168 153 88.4 93.4 1.01 1685 3033
Silver 10.500 234 0.0559 427 247 173.8 1.87 1234 2221
Tin 5.750 227 0.0542 67.0 38.7 51.3 0.552 505 909
Titanium 4.500 611 0.146 22.0 12.7 8.0 0.0861 1953 3515
Tungsten 19.300 134 0.0320 179 103 69.2 0.745 3653 6575
Uranium 19.070 113 0.0270 27.4 15.8 12.7 0.137 1407 2533
Vanadium 6.100 502 0.120 31.4 18.1 10.3 0.111 2192 3946
Zinc 7.140 385 0.0920 121 69.9 44.0 0.474 693 1247
Notes: Density = ρ = specific gravity × 62.4 lbm/ft3 = specific gravity × 1000 kg/m3.
Diffusivity = α; for aluminum, α m2/s × 106 = 97.5; so α = 97.5 × 10−6 m2/s.
Also, α = k/ρcp.
509
510
TABLE D.2
Thermal Properties of Selected Alloys
Coeff. of Linear Approximate
Specific Heat, cp Thermal Conductivity, k Diffusivity, α Expansion Melting point
Specific
Alloy Composition Gravity J/(kg K) Btu/(lbm °R) W/(m K) Btu/(h ft °R) m2/s × 105 ft2/s × 104 μ m/m K μ in./in. °F °C °F
Aluminum
Aluminum alloy ASTM B221 2.73 155.7 90 23.2 12.9 649 1200
3003, rolled
Aluminum alloy ASTM B221 2.8 164.4 95 22.9 12.7 641 1185
2017, annealed
Aluminum ASTM SC84B 2.7 96.9 56 20.9 11.6 566 1050
alloy 380
Duralumin 95 Al, 5 Cu 2.787 833 0.199 164 94.7 6.676 7.187
Silumin 87 Al, 13 Si 2.659 871 0.208 164 94.7 7.099 7.642
Copper
Copper ASTM B152, B124, 8.91 389.3 225 16.7 9.3 1082 1980
B133, B1, B2, B3
Red brass (cast) ASTM B30, No. 4A 8.7 72.7 42 18.0 10.0 996 1825
Yellow brass ASTM B36, B134, 8.47 119.4 69 18.9 10.5 932 1710
(high brass) B135
Aluminum ASTM B169, Alloy A; 7.8 70.9 41 16.6 9.2 1038 1900
bronze ASTM B124, B150
Beryllium ASTM B194 8.25 12.1 7 16.7 9.3 927 1700
copper 25
A-bronze 95 Cu, 5 Al 8.666 410 0.0979 83 47.9 2.330 2.508
Bronze 75 Cu, 25 Sn 8.666 343 0.0819 26 15.0 0.859 0.925
Red brass 85 Cu, 9 Sn, 6 Zn 8.714 385 0.0920 61 35.2 1.804 1.942
Brass 70 Cu, 30 Zn 8.522 385 0.0920 111 64.1 3.412 3.673
German silver 62 Cu, 15 Ni, 22 Zn 8.618 394 0.0941 24.9 14.4 0.733 0.789
Constantan 60 Cu, 40 Ni 8.922 410 0.0979 22.7 13.1 0.612 0.659
Cupronickel 30% 8.95 29.4 17 15.3 8.5 1227 2240
Cupronickel 55-45 (Constantan) 8.9 22.5 13 14.6 8.1 1260 2300
Iron
Ingot iron 7.86 72.7 42 12.2 6.8 1538 2800
Cast gray iron ASTM A48-48, Class 7.2 45.0 26 12.1 6.7 1177 2150
25
Malleable iron ASTM A47 7.32 11.9 6.6 1232 2250
Ductile cast iron ASTM A339, A395 7.2 32.9 19 13.5 7.5 1149 2100
Ni-resist cast iron type 2 7.3 39.8 23 17.3 9.6 1232 2250
Cast iron 4C 7.272 420 0.100 52 30.0 1.702 1.832
Wrought iron 0.5 CH 7.849 460 0.110 59 34.1 1.626 1.750
Steel
Plain carbon steel AISI-SAE 1020 7.86 51.9 30 12.1 6.7 1516 2760
Carbon steel 1C 7.801 473 0.113 43 24.8 1.172 1.262
1.5 C 7.753 486 0.113 36 20.8 0.970 1.040
Chrome steel 1 Cr 7.865 460 0.110 61 35.2 1.665 1.792
5 Cr 7.833 460 0.110 40 23.1 1.110 1.195
10 Cr 7.785 460 0.110 31 17.9 0.867 0.933
Chrome-nickel 15 Cr, 10 Ni 7.865 460 0.110 19 11.0 0.526 0.577

steel 20 Cr, 15 Ni 7.833 460 0.110 15.1 8.72 0.415 0.447
Nickel steel 10 Ni 7.945 460 0.110 26 15.0 0.720 0.775
20 Ni 7.993 460 0.110 19 11.0 0.526 0.566
40 Ni 8.169 460 0.110 10 5.78 0.279 0.300
60 Ni 8.378 460 0.110 19 11.0 0.493 0.531
Nickel-chrome 80 Ni, 15 C 8.522 460 0.110 17 9.82 0.444 0.478
steel 40 Ni, 15 C 8.073 460 0.110 11.6 6.70 0.305 0.328
Manganese steel 1 Mn 7.865 460 0.110 50 28.9 1.388 1.494
5 Mn 7.849 460 0.110 22 12.7 0.637 0.686
Silicon steel 1 Si 7.769 460 0.110 42 24.3 1.164 1.164
5 Si 7.417 460 0.110 19 11.0 0.555 0.597
Stainless steel Type 304 8.02 461 0.110 14.4 8.32 0.387 0.417 17.3 9.6 1427 2600
Type 347 7.97 461 0.110 14.3 8.26 0.387 0.417
Tungsten steel 1W 7.913 448 0.107 66 31.1 1.858 2.000
5W 8.073 435 0.104 54 31.2 1.525 1.642
Other
Chemical lead 11.35 34.6 20 29.5 16.4 327 621
Antimonial lead 10.9 29.4 17 27.2 15.1 290 554
(hard lead)
Magnesium alloy AZ31B 1.77 77.9 45 26.1 14.5 627 1160
Nickel ASTM B160, B161, 8.89 60.6 35 11.9 6.6 1441 2625
B162
Nickel silver 18% ASTM B122, No. 2 8.8 32.9 19 16.2 9.0 1110 2030
alloy A (wrought)
Commercial 5 17.3 10 8.8 4.9 1816 3300
titanium
Zinc ASTM B69 7.14 107.3 62 32.4 18.0 418 785
(continued)
511
512
TABLE D.2 (continued)

Thermal Properties of Selected Alloys
Coeff. of Linear Approximate
Specific Heat, cp Thermal Conductivity, k Diffusivity, α Expansion Melting point
Specific
Alloy Composition Gravity J/(kg K) Btu/(lbm °R) W/(m K) Btu/(h ft °R) m2/s × 105 ft2/s × 104 μ m/m K μ in./in. °F °C °F
Zirconium, 6.5 17.3 10 5.2 2.9 1843 3350
commercial
Cast 28-7 alloy ASTM A297-63T 7.6 2.6 1.5 16.6 9.2 1482 2700
(HD)
Hastelloy C 3.94 8.7 5 11.3 6.3 1288 2350
Inconel X, annealed 8.25 15.6 9 12.1 6.7 1399 2550
Haynes Stellite alloy 25 (L605) 9.15 9.5 5.5 13.7 7.6 1371 2500
K Monel 8.47 19.0 11 13.3 7.4 1349 2460
Solder 50-50 8.89 45.0 26 23.6 13.1 216 420
Sources: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000, Table B.2, p. 644; Kreith, F. (ed.), The CRC Handbook of Thermal Engineering, CRC Press, Boca Raton,
FL, 2000, Table C.6.
Appendix D: Properties of Solids 513
TABLE D.3
Thermal Properties of Selected Building Materials and Insulations at 293 K (20°C) or 528°R (65°F)
Specific Heat, cp Thermal Conductivity, k Diffusivity, α
Specific
Material Gravity J/(kg K) Btu/(lbm °R) w/(m K) Btu/(h ft °R) m /s × 105
2 ft2/s × 106
Asbestos 0.383 816 0.195 0.113 0.0653 0.036 3.88
Asphalt 2.120 0.698 0.403
Bakelite 1.270 0.233 0.135
Brick
Carborundum (50%SiC) 2.200 5.82 3.36
Common 1.800 840 0.201 0.38–0.52 0.22–0.30 0.028–0.034 3.0–3.66
Magnesite (50%MgO) 2.000 2.68 1.55
Masonry 1.700 837 0.200 0.658 0.38 0.046 5.0
Silica (95%SiO2) 1.900 1.07 0.618
Cardboard 0.14–0.35 0.08–0.2
Cement (hard) 1.047 0.605
Clay (48.7%moist) 1.545 880 0.210 1.26 0.728 0.101 10.9
Coal (anthracite) 1.370 1260 0.301 0.238 0.137 0.013–0.015 1.4–1.6
Concrete (dry) 0.500 837 0.200 0.128 0.074 0.049 5.3
Cork board 0.150 1880 0.449 0.042 0.0243 0.015–0.044 1.6–4.7
Cork (expanded) 0.120 0.036 0.0208
Earth (diatomaceous) 0.466 879 0.210 0.126 0.072 0.031 3.3
Earth (clay with 28% moist) 1.500 1.51 0.872
Earth (sandy with 8% moist) 1.500 1.05 0.607
Glass fiber 0.220 0.035 0.02
Glass (window pane) 2.800 800 0.191 0.81 0.47 0.034 3.66
Glass (wool) 0.200 670 0.160 0.040 0.023 0.028 3.0
Granite 2.750 3.0 1.73
Ice at 0°C 0.913 1830 0.437 2.22 1.28 0.124 13.3
Kapok 0.025 0.035 0.02
Linoleum 0.535 0.081 0.047
Mica 2.900 0.523 0.302
Pine bark 0.342 0.080 0.046
Plaster 1.800 0.814 0.47
Notes: Density = ρ = specify gravity × 62.4 lbm/ft3 = specific gravity × 1000 kg/m3.
Diffusivity = α; for asbestos, α × 103 = 0.036 m2/s; so α = 0.036 × 10−3 m2/s also, α = k/ρcp.
References
1. Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC
2. Kreith, F. (ed.), The CRC Handbook of Thermal Engineering,
The Coen & Hamworthy Combustion Handbook,
Fundamentals for Power, Marine &
Industrial Applications
Edited by Stephen B. Londerville & Charles E. Baukal, Jr.,
John Zink Company LLC, Tulsa, Oklahoma, USA
The rigorous treatment of combustion can be so complex that the kinetic variables,
fluid turbulence factors, luminosity, and other factors cannot be defined well enough to
find realistic solutions. Simplifying the processes, The Coen & Hamworthy Combustion
Handbook provides practical guidance to help readers make informed choices about fuels,
burners, and associated combustion equipment—and to clearly understand the impacts
of the many variables. Editors Stephen B. Londerville and Charles E. Baukal, Jr., top
combustion experts from John Zink Hamworthy Combustion and the Coen Company, supply
a thorough, state-of-the-art overview of boiler burners that covers Coen, Hamworthy, and
Todd brand boiler burners.
Roughly divided into two parts, the book first reviews combustion engineering fundamentals. It
then uses a building-block approach to present specific computations and applications in
industrial and utility combustion systems, including those for
• Transport and introduction of fuel and air to a system
• Safe monitoring of the combustion system
• Control of flows and operational parameters
• Design of a burner/combustion chamber to achieve performance levels for emissions

and heat transfer
• Avoidance of excessive noise and vibration and the extension of equipment life under
adverse conditions
Coverage includes units, fluids, chemistry, and heat transfer, as well as atomization,
computational fluid dynamics (CFD), noise, auxiliary support equipment, and the combustion
of gaseous, liquid, and solid fuels. Significant attention is also given to the formation,
reduction, and prediction of emissions from combustion systems. Each chapter builds from
the simple to the more complex and contains a wealth of practical examples and full-color
photographs and illustrations.
A ready reference and refresher, this unique handbook is designed for anyone involved in
combustion equipment selection, sizing, and emissions control. It will help readers make
calculations and decisions on design features, fuel choices, emissions, controls, burner
selection, and burner/furnace combinations with more confidence.
K13399
ISBN: 978-1-4398-7333-5
90000
9 781439 873335
w w w.crcpress.com

Combustion Handbook - Fundamentals For Power, Marine & Industrial Applications - The Coen & Hamworthy

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Combustion Handbook - Fundamentals For Power, Marine & Industrial Applications - The Coen & Hamworthy

Hochgeladen von

Copyright:

Verfügbare Formate

THE COEN & HAMWORTHY

Fundamentals for Power, Marine & Industrial Applications

The Coen & Hamworthy Combustion Handbook:

No claim to original U.S. Government works

International Standard Book Number-13: 978-1-4398-7334-2 (eBook - PDF)

6. Solid Fuel Combustion in Suspension........................................................................................................................115

7. Heat Transfer................................................................................................................................................................... 127

8. Fundamentals of Fluid Dynamics............................................................................................................................... 153

9. CFD-Based Combustion Modeling............................................................................................................................ 183

10. Pollutant Emissions.........................................................................................................................................................211

11. Noise.................................................................................................................................................................................. 267

12. Combustion Controls, Burner Management, and Safety Systems....................................................................... 301

13. Blowers for Combustion Systems............................................................................................................................... 335

15. Duct Burners.................................................................................................................................................................... 397

16. Air Heaters....................................................................................................................................................................... 421

17. Marine and Offshore Applications............................................................................................................................. 431

Appendix A: Units and Conversions................................................................................................................................. 443

Appendix B: Physical Properties of Materials................................................................................................................. 449

Appendix D: Properties of Solids...................................................................................................................................... 509

Figure 4.3 Burning velocity for various gaseous fuels................................................................................................. 84

Figure 5.27 CFI versus CCR, RCR and asphaltenes...................................................................................................... 112

Figure 8.19 Photograph of a gas exiting a nozzle......................................................................................................... 181

Figure 10.9 Schematic of fuel NOx formation pathways............................................................................................ 221

Figure 10.37 Particulate sampling train......................................................................................................................... 241

Figure 14.4 Coen QLA burner performance at high fire........................................................................................... 367

Figure 14.7 Hamworthy DFL® burner.......................................................................................................................... 371

Table 2.1 Common English Units.................................................................................................................................... 16

Table 7.1 Specific Heat Curve Fit Equation.................................................................................................................. 128

Table 13.4 Fan Bearing Vibration Limits........................................................................................................................ 350

Stephen B. Londerville, Timothy Webster, and Charles E. Baukal, Jr.

type of heating process. While this book is concerned

a complete line of coal-fired burners and its fuel feed

Flame Radiant section

1.3 History of Hamworthy Combustion

hydrocarbon fuels. This section covers a variety of different

1.4.2 Package Boilers

A “D” style boiler has two drums located directly

• High reliability during long-term operation

Advanced Electrostatic Electrostatic

firebox. By controlling the location of the fireball rela-

1.8 Air Heaters

used are also called ribbon burners, linear burners, or

was slowly modernized and CFD was first utilized for a

the incident cost NASA well over $600 million.1 Despite

and universal measurement system and thus ushered TABLE 2.1

Thus in English units yields TABLE 2.3

• Kilogram-force As discussed earlier in this section, all units can be derived

miles 5, 280 ft 0.3048 m 1 km km

From this A list of conversion factors can be found in Table 2.6 at

196.12 N = 20 kg f 2.2.5 Some Simple Derived Units

2.2.5.1 Velocity Example 2.5

Common units of specific weight are

mass Common units of pressure include

Example 2.8 6 Btu 1J 1h

Temperature is commonly measured in Fahrenheit Example 2.10

y = mx + b 2.5 Enthalpy and Specific Heat

Kelvin : †K = °C + 273.15 (2.12) energy

Common units of specific heat

0.454 kg 1° C∆ 1.00 Btu

It is important to note here that In liquids, additional intermolecular forces exist

Example 2.13 TABLE 2.5

Commonly used units include

Stephen B. Londerville, Joseph Colannino, and Charles E. Baukal, Jr.

Figure 4.3 Burning velocity for various gaseous fuels................................................................................................. 84

Figure 5.27 CFI versus CCR, RCR and asphaltenes...................................................................................................... 112

Figure 8.19 Photograph of a gas exiting a nozzle......................................................................................................... 181

Figure 10.9 Schematic of fuel NOx formation pathways............................................................................................ 221

Figure 10.37 Particulate sampling train......................................................................................................................... 241

Figure 14.4 Coen QLA burner performance at high fire........................................................................................... 367

Figure 14.7 Hamworthy DFL® burner.......................................................................................................................... 371

Example 3.2 60°F (15.6°C) = Air @ 14.7 psia (1 atm) at 60°F