Beruflich Dokumente
Kultur Dokumente
COMBUSTION HANDBOOK
STEPHEN B. LONDERVILLE
CHARLES E. BAUKAL, JR.
Editors
THE COEN & HAMWORTHY
COMBUSTION HANDBOOK
Fundamentals for Power, Marine & Industrial Applications
IndustrIal combustIon serIes
Series Editors:
Charles E. Baukal, Jr.
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
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
14. Burners and Combustion Systems for Industrial and Utility Boilers................................................................. 359
Vladimir Lifshits
vii
viii Contents
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
diabatic 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............................................................................................ 63
Figure 3.21 A
diabatic 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.................................................................................................... 64
Figure 3.22 A
diabatic equilibrium stoichiometric calculations for the predicted gas composition of the
minor species as a function of the fuel preheat temperature for air/CH4 flames, where the air
is at ambient temperature and pressure.................................................................................................... 64
Figure 3.23 A
diabatic 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......................................................... 65
Figure 3.24 A
diabatic equilibrium stoichiometric calculations for the predicted gas composition of the
minor 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 3.25 A
diabatic equilibrium stoichiometric calculations for the predicted gas composition of the
major species as a function of the fuel blend (N2 + CH4) composition for air/fuel flames,
where the air and fuel are at ambient temperature and pressure......................................................... 66
Figure 3.26 A
diabatic equilibrium stoichiometric calculations for the predicted gas composition of the
minor species as a function of the fuel blend (N2 + CH4) composition for air/fuel flames,
where the air and fuel are at ambient temperature and pressure......................................................... 66
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
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 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 14.26 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
xxiii
xxiv List of Tables
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
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
1
2 The Coen & Hamworthy Combustion Handbook
FIGURE 1.4
Front elevation of “O” type boiler—the furnace and drum locations
form the letter “O,” hence the name.
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.)
6 The Coen & Hamworthy Combustion Handbook
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.
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
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
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
12 The Coen & Hamworthy Combustion Handbook
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
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
15
16 The Coen & Hamworthy Combustion Handbook
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
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
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
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
Engineering Fundamentals 19
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
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.
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
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.
22 The Coen & Hamworthy Combustion Handbook
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
Engineering Fundamentals 23
The units for bulk modulus are Measures, this notation is used mostly by continental
European countries (see Table 2.5).
Unit Area
Ev =
Force
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
26 The Coen & Hamworthy Combustion Handbook
FIGURE 3.1
“O” type package boiler. (Courtesy of Victory Energy Operations, LLC.)
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.
2H 2 + O 2 → 2H 2O
–
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.)
30 The Coen & Hamworthy Combustion Handbook
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
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
32 The Coen & Hamworthy Combustion Handbook
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:
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
Combustion Fundamentals 35
B B B
εA+ εA+ + 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.
36 The Coen & Hamworthy Combustion Handbook
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%
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%
(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.)
Combustion Fundamentals 37
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%
20% 90%
18%
Dry N2
85%
16%
H2O
14%
80%
Vol.% species (except N2)
CO2
12%
Wet N2
Vol.% N2
10% 75%
8%
70%
6%
O2
4%
65%
2%
20% 90%
18%
85%
16%
Dry N2
14%
CO2 80%
Vol.% species (except N2)
12%
Wet N2
Vol.% N2
10% 75%
H2O
8%
O2 70%
6%
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%
(e) Percent excess air
20% 90%
18%
85%
16%
Dry N2
14%
CO2 80%
Vol.% species (except N2)
12%
Wet N2 Vol.% N2
10% 75%
H2O
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%
(f ) Percent excess air
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.
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
40 The Coen & Hamworthy Combustion Handbook
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:
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:
= 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.
42 The Coen & Hamworthy Combustion Handbook
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.
∫
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
44 The Coen & Hamworthy Combustion Handbook
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
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
Combustion Fundamentals 45
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
Guess
Heating value TempAD LHV
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
46 The Coen & Hamworthy Combustion Handbook
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.
Combustion Fundamentals 47
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
48 The Coen & Hamworthy Combustion Handbook
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
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.
Combustion Fundamentals 49
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
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.)
0 100 200 300 400 500 600 700 800 900 1000
4600
2500
4400
2400
Adiabatic flame temperature (°F)
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.)
Combustion Fundamentals 51
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
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.)
52 The Coen & Hamworthy Combustion Handbook
2450
4400 77°F (25°C)
2400
500°F (260°C)
4300
1000°F (538°C) 2350
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.)
80 80
70 70
60 60
Available heat (%)
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
54 The Coen & Hamworthy Combustion Handbook
85 85
H2
80 CH4 80
C3H8
75 75
Available heat (%)
65 65
60 60
55 55
50 50
45 45
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Air preheat temperature (°F)
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.)
5 5
H2
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
Fuel preheat temperature (°F)
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.)
Combustion Fundamentals 55
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
56 The Coen & Hamworthy Combustion Handbook
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
Combustion Fundamentals 57
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:
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 )
58 The Coen & Hamworthy Combustion Handbook
20% 90%
88%
18%
86%
16% 84%
H2O Dry N2
82%
14%
80%
Vol.% species (except N2)
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%
Vol.% species (except N2)
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
(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.)
60 The Coen & Hamworthy Combustion Handbook
20% 90%
88%
18%
86%
16% 84%
Dry N2
H2O 82%
14%
80%
Vol.% species (except N2)
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
(c) Φ
20% 90%
Dry N2 88%
18%
86%
16% 84%
H2O 82%
14%
80%
Vol.% species (except N2)
CO2
12% 78%
Vol.% N2
76%
10%
CO 74%
8% 72%
Wet N2 70%
6%
68%
O2
4% 66%
64%
2% H2
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
(d) Φ
20% 90%
88%
18%
Dry N2 86%
16% 84%
82%
14%
CO2 80%
Vol.% species (except N2)
12% 78%
Vol.% N2
CO 76%
10%
74%
H2O
8% 72%
Wet N2
70%
6% O2
68%
4% H2 66%
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
(e) Φ
20% 90%
88%
18%
86%
16% 84%
Dry N2
CO 82%
14%
CO2 80%
Vol.% species (except N2)
12% 78%
76% Vol.% N2
10%
74%
H2O
8% 72%
70%
6%
Wet N2 68%
4% O2
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
(f) Φ
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.)
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
Combustion Fundamentals 63
55 55
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
CO
3.0 H 3.0
H2
NO
2.5 2.5
Species concentration (vol.%)
O
Species concentration (vol.%)
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
Air preheat temperature (°F)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
64 The Coen & Hamworthy Combustion Handbook
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
1.0 1.0
0.9 CO 0.9
H2
0.8 NO 0.8
Species concentration (vol.%)
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
Fuel preheat temperature (°F)
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.)
Combustion Fundamentals 65
70 70
60 60
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
1.6 1.6
CO
H
1.4 H2 1.4
NO
1.2 OH 1.2
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)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
66 The Coen & Hamworthy Combustion Handbook
100 100
90 90
80 80
Species concentration (vol.%)
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)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
0.9 0.9
0.8 CO 0.8
H2
0.7 NO 0.7
OH
Species concentration (vol.%)
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)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
Combustion Fundamentals 67
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.)
0.8 CO 0.8
H2
0.7 NO 0.7
OH
Species concentration (vol.%)
Species concentration (vol.%)
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
Gas temperature (°F)
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.)
68 The Coen & Hamworthy Combustion Handbook
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
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
Combustion Fundamentals 69
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
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.
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
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
72 The Coen & Hamworthy Combustion Handbook
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
74 The Coen & Hamworthy Combustion Handbook
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.)
76 The Coen & Hamworthy Combustion Handbook
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
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
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
Note: All values calculated using 60°F fuel gas and 60°F, 50% relative humidity combustion air.
79
80 The Coen & Hamworthy Combustion Handbook
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)
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
82 The Coen & Hamworthy Combustion Handbook
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:
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
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
84 The Coen & Hamworthy Combustion Handbook
Hydrogen
3.2 8.0
Burning velocity, ft/s (except hydrogen and acetylene)
2.4 6.0
Ethylene
Acetylene
1.6 4.0
Coke oven gas
Carbon Illuminants
Butane
1.2 monoxide 3.0
Ethane
0.8 2.0
Producer
gas
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.
Crude
Typical refinery
products
Regular
gasoline
ATMOS
TWR Middle distillates Hydro Solvents
section treating
Aviation fuels
Treatingandblending
Lube oils
Vacuum Solvent
section extraction
Greases
Gasoline, naphtha,
Lube base stocks
Hydro and middle distillates
Asphalts
cracking
Refinery
fuel oil
Gasoline, naphtha, and middle distillates
Visbreaker
Fuel oil
Asphalt
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
88 The Coen & Hamworthy Combustion Handbook
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
Main-column bottoms
Heavy recycle
(optional)
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) — —
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
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.
TABLE 4.13
Naphtha Elemental Analysis
Component Vol%
n-Heptane 1.610
Methylcyclohexane 2.433
2-Methylheptane 5.618
4-Methylheptane 1.824
3-Methylheptane 4.841
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
2-Methyloctane 4.491
3-Methyloctane 4.576
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
300
Temperature,°F
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
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 ]
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
94 The Coen & Hamworthy Combustion Handbook
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
95
96 The Coen & Hamworthy Combustion Handbook
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
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
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.
98 The Coen & Hamworthy Combustion Handbook
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
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
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:
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
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.
∑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
∑ ND
i
i i
2
FIGURE 5.6
A gun with a 90° machine angle. Its spray angle actually is about 30°.
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
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
Liquid port
FIGURE 5.15
WDH waste aqueous gun design with one liquid exit port sur-
rounded with eight atomizing ports.
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)
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
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-
Radial position (cm)
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)
8
Mass ratio
0
0 20 40 60 80 100 120
Oil pressure (Psig)
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
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
−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
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
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
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.
114 The Coen & Hamworthy Combustion Handbook
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
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
115
116 The Coen & Hamworthy Combustion Handbook
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.
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
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.
Solid Fuel Combustion in Suspension 121
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
FIGURE 6.4
Fuel introduction for conveying options.
The Coen & Hamworthy Combustion Handbook
Solid Fuel Combustion in Suspension 123
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)
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
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
128 The Coen & Hamworthy Combustion Handbook
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%
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.
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
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
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
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:
∑hp − ∑hr
g g = ∆hc(LHV) heat input ignoring changes in composition to
determine error in this method.
136 The Coen & Hamworthy Combustion Handbook
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,
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
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
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
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:
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
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.
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)
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.)
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%
146 The Coen & Hamworthy Combustion Handbook
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
L (pw + pc) = 5.0 ft-atm
1.5
0.04
1.0
L (pw + pc) = 5.0 ft-atm
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.)
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
( )
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.)
150 The Coen & Hamworthy Combustion Handbook
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
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
153
154 The Coen & Hamworthy Combustion Handbook
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
ft 2
FIGURE 8.1 1 cSt = 10.76 †
Velocity profile of fluid along a solid surface. s
156 The Coen & Hamworthy Combustion Handbook
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.
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.
158 The Coen & Hamworthy Combustion Handbook
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)
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
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.)
160 The Coen & Hamworthy Combustion Handbook
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
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
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.
162 The Coen & Hamworthy Combustion Handbook
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.
Fundamentals of Fluid Dynamics 163
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:
Therefore,
F ⋅ ds Work
Power = ∑ ∆T
=
Unit†time
(8.29)
hpActual =
hpTheoretical 69.3
η
=
0.75
= 92.4 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
Fundamentals of Fluid Dynamics 165
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
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.)
Fundamentals of Fluid Dynamics 167
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
168 The Coen & Hamworthy Combustion Handbook
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
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°
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.)
Fundamentals of Fluid Dynamics 169
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:
∆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.
170 The Coen & Hamworthy Combustion Handbook
−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.
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)
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.
172 The Coen & Hamworthy Combustion Handbook
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
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
Fundamentals of Fluid Dynamics 173
KA
AA AA KA
P1 P2 P1 P2
AB
AB KB
KB
FIGURE 8.12
Common arrangement for parallel flow problems.
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
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
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.
Fundamentals of Fluid Dynamics 175
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
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
176 The Coen & Hamworthy Combustion Handbook
FIGURE 8.17
Discharge coefficients for different internal nozzle designs. (Adapted from Baukal, C.E., Industrial Combustion Testing, CRC Press, Boca Raton,
FL, 2010.)
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
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
in. WC lbf fL γV 2
0.00875psi × 27.7 = 0.24 in. WC 2.69 = 1 + 2
psi ft 2
D 2g
178 The Coen & Hamworthy Combustion Handbook
−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
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 = ρ1 + 0.912 = 0.04121
Pt = P 1 + M (8.72) 2 ft 3
2
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 = ρ1 + 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
Fundamentals of Fluid Dynamics 181
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.)
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.)
182 The Coen & Hamworthy Combustion Handbook
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
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
183
184 The Coen & Hamworthy Combustion Handbook
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).
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
186 The Coen & Hamworthy Combustion Handbook
(
∂ 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
CFD-Based Combustion Modeling 187
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)
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
TABLE 9.2
Cartesian Differential Equation Set
∂(ρu
φ ) ∂(ρvφ) ∂(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
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
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,
190 The Coen & Hamworthy Combustion Handbook
TABLE 9.3
Cylindrical Differential Equation Set
∂(ρu
φ ) ∂( rρvφ) ∂(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
2vµ 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 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
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
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
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.
192 The Coen & Hamworthy Combustion Handbook
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)
1
The difficulty at hand for the formulation of a global Y F , actual = (9.42)
(1/τYkin ) + (1/Yeddy )
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 Yi = 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 YF = 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) YO2 = A O2 (9.45)
RT rO2 k
CFD-Based Combustion Modeling 195
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)
m1 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
CFD-Based Combustion Modeling 197
∫
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
198 The Coen & Hamworthy Combustion Handbook
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.
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
200 The Coen & Hamworthy Combustion Handbook
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
CFD-Based Combustion Modeling 201
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.
FIGURE 9.10
Discretized geometry of a typical boiler burner.
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
CFD-Based Combustion Modeling 205
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.
206 The Coen & Hamworthy Combustion Handbook
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
CFD-Based Combustion Modeling 207
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
FIGURE 9.18
Temperature distribution in flame for a tangential fired boiler.
(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.
CFD-Based Combustion Modeling 209
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
212 The Coen & Hamworthy Combustion Handbook
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
PM10 14.9
Lead 4.8
NO2 0.0
CO 0.0
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.)
Direct PM2.5
Direct PM10
NH3
Source category
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
Pollutant Emissions 215
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
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
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
Pollutant Emissions 217
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
[ NO ] = K1t e K /T [ N 2 ][ O 2 ]1/2 (10.15)
2
220 The Coen & Hamworthy Combustion Handbook
3000
2500
NOx emissions (short tons)
2000
1500
1000
500
0
1970 1975 1980 1985 1990 1995 2000
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.)
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,
CRC Press, Boca Raton, FL, 2001.)
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
Gas temperature (°F)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
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
Air preheat temperature (°F)
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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.
224 The Coen & Hamworthy Combustion Handbook
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
Fuel preheat temperature (°F)
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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
Pollutant Emissions 225
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
Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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
45
40
35
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
Pollutant Emissions 227
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
Burner A
Burner B
Insulation
FIGURE 10.19
Schematic of HiTAC furnace. (Adapted from Tsuji, H. et al., High Temperature Air Combustion: From Energy Conservation to Pollution Reduction,
CRC Press, Boca Raton, FL, 2003.)
228 The Coen & Hamworthy Combustion Handbook
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.
Pollutant Emissions 229
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
230 The Coen & Hamworthy Combustion Handbook
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
232 The Coen & Hamworthy Combustion Handbook
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 com- ing 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 pro- of 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
Pollutant Emissions 233
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
25–40 ppm
FIGURE 10.27
History of low NO burner development for (a) round flame burners and (b) radiant wall burners, firing on gaseous fuels.
The Coen & Hamworthy Combustion Handbook
Pollutant Emissions 235
(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
236 The Coen & Hamworthy Combustion Handbook
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
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.)
8000 8000
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
Gas temperature (°F)
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.
Pollutant Emissions 237
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
Air preheat temperature (°F)
FIGURE 10.32
Adiabatic equilibrium CO as a function of air preheat temperature for stoichiometric air/fuel flames. (From Baukal, C.E., Industrial Combustion
Pollution and Control, Marcel Dekker, New York, 2004.)
CH4
14,000 14,000
C3H8
13,000 13,000
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
Fuel preheat temperature (°F)
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
Pollution and Control, Marcel Dekker, New York, 2004.)
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
238 The Coen & Hamworthy Combustion Handbook
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.)
Pollutant Emissions 239
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.
240 The Coen & Hamworthy Combustion Handbook
Temperature sensor
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
242 The Coen & Hamworthy Combustion Handbook
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
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.
244 The Coen & Hamworthy Combustion Handbook
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)
Pollutant Emissions 245
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
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.
Pollutant Emissions 247
Exhaust
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
Sampling ports
Quartz
step and liner
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.)
Pollutant Emissions 249
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.)
250 The Coen & Hamworthy Combustion Handbook
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
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
By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
section simulator)
10–6
10–7
10–8
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.)
252 The Coen & Hamworthy Combustion Handbook
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
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.)
Pollutant Emissions 253
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.)
254 The Coen & Hamworthy Combustion Handbook
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
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
International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
Pollutant Emissions 255
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
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,
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.)
10–3
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
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
Pollutant Emissions 257
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
258 The Coen & Hamworthy Combustion Handbook
108
107
106
105 b(a)p
b(a)p b(a)p
104
ppt
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
Stoichiometric ratio
FIGURE 10.59
Benzene (blue) and PAH (red) emissions versus stoichiometric ratio 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 International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
Pollutant Emissions 259
1011
109
CO
107
ppt
105
103
PAH
101
10–1
Stoichiometric ratio
FIGURE 10.60
CO (green) and PAH (red) emissions versus stoichiometric ratio 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.)
1011
109
HC
107
ppt
105
103
PAH
101
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 10.61
HC (yellow) and PAH (red) emissions versus stoichiometric ratio 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.)
260 The Coen & Hamworthy Combustion Handbook
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
Stoichiometric ratio
FIGURE 10.62
HC, aldehyde, VOC, and PAH emissions versus stoichiometric ratio 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 International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
108
107
106
105
104
ppt
103
102
101
100
10–1
3.0 1.25 0.80 0.67 0.60
Stoichiometric ratio
FIGURE 10.63
Total heavy VOC emissions versus stoichiometric ratio 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
International Congress on Combustion By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
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,
Pollutant Emissions 261
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
By-Products and Their Health Effects, Ischia, Italy, June 17–20, 2007.)
262 The Coen & Hamworthy Combustion Handbook
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.
264 The Coen & Hamworthy Combustion Handbook
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.
Pollutant Emissions 265
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
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
268 The Coen & Hamworthy Combustion Handbook
FIGURE 11.1
Community located close to an industrial plant. (Photograph from Dreamstime.)
Noise 269
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
270 The Coen & Hamworthy Combustion Handbook
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
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.)
272 The Coen & Hamworthy Combustion Handbook
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.
Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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.)
274 The Coen & Hamworthy Combustion Handbook
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
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
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook,
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
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
Atmospheric attenuation (dB/1000 ft)
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
Atmospheric attenuation (dB/1000 ft)
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)
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
280 The Coen & Hamworthy Combustion Handbook
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.
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
282 The Coen & Hamworthy Combustion Handbook
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.
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
W(watts)
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)
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
Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
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
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
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
286 The Coen & Hamworthy Combustion Handbook
110
Over-steamed flare
105
100
90
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
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.)
288 The Coen & Hamworthy Combustion Handbook
Region of
maximum jet
mixing noise 30°
15°
Jet centerline
Nozzle 0°
1 2 3 4
Nozzle exit
Noise level
Broadband
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.)
290 The Coen & Hamworthy Combustion Handbook
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
292 The Coen & Hamworthy Combustion Handbook
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
110
105
100
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
Press, Boca Raton, FL, 2001.)
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
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
294 The Coen & Hamworthy Combustion Handbook
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.
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.
11.5.2 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
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)
FIGURE 11.45
Effect of distance on flare noise.
298 The Coen & Hamworthy Combustion Handbook
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.
300 The Coen & Hamworthy Combustion Handbook
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
302 The Coen & Hamworthy Combustion Handbook
• 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.
304 The Coen & Hamworthy Combustion Handbook
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 usu- ally 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.
Combustion Controls, Burner Management, and Safety Systems 305
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
FIGURE 12.2 E=
(CV − BIAS) (12.3)
Steam pressure control logic. GAIN
306 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 307
Master
demand
f (x)
Power unit
Jackshaft
Mechanical
linkage Adjustable
cam
FIGURE 12.3
Schematic of SPP.
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%
T A T A
HMI HMI
FIGURE 12.5
Parallel positioning of fuel and air.
Combustion Controls, Burner Management, and Safety Systems 309
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
310 The Coen & Hamworthy Combustion Handbook
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 sub- systems 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
Combustion Controls, Burner Management, and Safety Systems 311
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.
312 The Coen & Hamworthy Combustion Handbook
Air Fuel
Master flow flow
demand
FT FT
A
÷ ÷
Δ ÷
f3(x)
Σ
>HI >LO
1.0 >HI ÷
× 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.
60
Flammability limit in a well stirred reactor
50
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
50
Over 50% firing rate
40
2
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
Air to fuel stoichiometric ratio
FIGURE 12.9
Operating ranges and NOx performance of a lean premixed burner with auxiliary fuel zones.
314 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 315
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
316 The Coen & Hamworthy Combustion Handbook
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 burn- ing 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.
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.
318 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 319
FIGURE 12.11
COEN i-scan™ flame scanner.
320 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 321
0.15 0.3
0.125
–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
322 The Coen & Hamworthy Combustion Handbook
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.
Combustion Controls, Burner Management, and Safety Systems 323
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.
324 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 325
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
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.
The Coen & Hamworthy Combustion Handbook
Combustion Controls, Burner Management, and Safety Systems 327
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
Water inlet
Thermocouple
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)
Combustion Controls, Burner Management, and Safety Systems 329
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).
330 The Coen & Hamworthy Combustion Handbook
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
Combustion Controls, Burner Management, and Safety Systems 331
P
E ~ T × ln 1 (12.7)
P2
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.
Combustion Controls, Burner Management, and Safety Systems 333
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
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
335
336 The Coen & Hamworthy Combustion Handbook
• 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
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.)
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%
Source: Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.
a Size is based on fans at the same speed, volume, and pressure.
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
Blowers for Combustion Systems 339
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
340 The Coen & Hamworthy Combustion Handbook
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
Blowers for Combustion Systems 341
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
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
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.)
Blowers for Combustion Systems 343
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.
344 The Coen & Hamworthy Combustion Handbook
3
RPM new
BHPnew = × BHPold (13.14)
RPM old
Density old
RPM new BHPnew = × BHPold (13.20)
CFM new = × CFM old (13.12) Density new
RPM old
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
Speed reduced
Surge point
Design point
Static pressure
Static pressure
Volumetric flow
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
Blowers for Combustion Systems 349
FIGURE 13.21
Close-up of a flexible coupling.
FIGURE 13.23
Oil lubricated bearings with reservoir.
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.
Fan wheel
Driver
Inlet
box
Outboard
concrete
pedestal Fan
HSG
Soil Soil
Bedrock
Bedrock
FIGURE 13.25
Fan foundation.
FIGURE 13.26
Inlet and outlet expansion joints for vibration isolation of ducting.
352 The Coen & Hamworthy Combustion Handbook
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
Blowers for Combustion Systems 353
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
50 System resistance
SP3 curve
10
Opposed
blade
With outlet outlet
Design
dampering damper
2000 fan BHP
BHP
follows base
HP curve BHP1 = 1257
1000
BHP2 = 1095
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
FIGURE 13.28
Inlet damper fan curve with horsepower.
Blowers for Combustion Systems 355
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.
356 The Coen & Hamworthy Combustion Handbook
TABLE 13.8
Blower Trouble Shooting Guide
Problem Probable Cause
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
359
360 The Coen & Hamworthy Combustion Handbook
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
362 The Coen & Hamworthy Combustion Handbook
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).
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.
Burners and Combustion Systems for Industrial and Utility Boilers 363
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
364 The Coen & Hamworthy Combustion Handbook
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.
Burners and Combustion Systems for Industrial and Utility Boilers 365
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.
366 The Coen & Hamworthy Combustion Handbook
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
Air to fuel stoichiometric ratio
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.
368 The Coen & Hamworthy Combustion Handbook
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
Burners and Combustion Systems for Industrial and Utility Boilers 369
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
370 The Coen & Hamworthy Combustion Handbook
Throat
Gas ring
Pilot
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.
Burners and Combustion Systems for Industrial and Utility Boilers 371
FIGURE 14.6
Coen Variflame™ burner. (From Baukal, C.E., Ed., The John Zink Combustion Handbook, CRC Press, Boca Raton, FL, 2001.)
FIGURE 14.8
Coen DAF™ burner.
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.
374 The Coen & Hamworthy Combustion Handbook
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
Burners and Combustion Systems for Industrial and Utility Boilers 375
FIGURE 14.14
Coen RMB™ burner equipped with an air isolation sliding barrel damper.
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.)
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).
378 The Coen & Hamworthy Combustion Handbook
FIGURE 14.20
Assembled Coen QLA burner (side).
Burners and Combustion Systems for Industrial and Utility Boilers 379
FIGURE 14.21
Assembled Coen QLA burner (front).
FIGURE 14.23
Peabody LVC™ burner for firing BFG.
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
Burners and Combustion Systems for Industrial and Utility Boilers 381
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).
20
18
16
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).
384 The Coen & Hamworthy Combustion Handbook
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.
Burners and Combustion Systems for Industrial and Utility Boilers 385
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
386 The Coen & Hamworthy Combustion Handbook
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.
Burners and Combustion Systems for Industrial and Utility Boilers 387
0.9
Original burner
0.8 Gas REACH
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
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.
388 The Coen & Hamworthy Combustion Handbook
90%
80%
70%
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
Burners and Combustion Systems for Industrial and Utility Boilers 389
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 direc- tion 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
390 The Coen & Hamworthy Combustion Handbook
(a) (b)
FIGURE 14.40
Flame stabilizers and buckets of tilting (a) gas- and (b) oil-fired burner.
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
392 The Coen & Hamworthy Combustion Handbook
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.
Burners and Combustion Systems for Industrial and Utility Boilers 393
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
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.
394 The Coen & Hamworthy Combustion Handbook
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
398 The Coen & Hamworthy Combustion Handbook
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
400 The Coen & Hamworthy Combustion Handbook
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.)
402 The Coen & Hamworthy Combustion Handbook
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.)
FIGURE 15.7
Inline burner. (From Baukal, C.E. (ed.), The John 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
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 differ- ing 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.)
408 The Coen & Hamworthy Combustion Handbook
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
410 The Coen & Hamworthy Combustion Handbook
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.)
412 The Coen & Hamworthy Combustion Handbook
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.
414 The Coen & Hamworthy Combustion Handbook
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
Handbook, CRC Press, Boca Raton, FL, 2001.)
Vent to Vent to
atmosphere atmosphere
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
FIGURE 15.21
Typical main gas fuel train: multiple elements with individual firing capability. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook,
CRC Press, Boca Raton, FL, 2001.)
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,
CRC Press, Boca Raton, FL, 2001.)
Vent to Vent to
atmosphere atmosphere
V4 V8
To
(Optional PI igniter
location) (typical)
V7
PSL
Gas
supply
V1 V2 V3 V3 To
other
igniters
PSH
(permanently
installed)
PSL
FIGURE 15.23
Typical pilot gas train: multiple elements with individual firing capability. (From Baukal, C.E. (ed.), The John Zink Combustion Handbook, CRC
Press, Boca Raton, FL, 2001.)
416 The Coen & Hamworthy Combustion Handbook
• 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
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
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
V6 = Igniter flow control 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.)
418 The Coen & Hamworthy Combustion Handbook
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
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.
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
421
422 The Coen & Hamworthy Combustion Handbook
Burner
Hot
gases
outlet
Fuel
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.
424 The Coen & Hamworthy Combustion Handbook
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
Burner
Hot
gases
outlet
Fuel
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
426 The Coen & Hamworthy Combustion Handbook
Air inlet
Oil gun
C D A1
Flexible
hoses
Peephole Ignitor
Ignitor Diffuser
Gas scroll Oil supply
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
428 The Coen & Hamworthy Combustion Handbook
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
FIGURE 16.8
Low NOx QL burner for air heaters.
430 The Coen & Hamworthy Combustion Handbook
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
431
432 The Coen & Hamworthy Combustion Handbook
(a) (b)
FIGURE 17.2
Hamworthy Combustion DF register burner. (a) Windbox view with fuel valve train and (b) burner throat view.
(a) (b)
FIGURE 17.4
Hamworthy Combustion HXG dual-fuel register burner. (a) Burner frontplate view and (b) burner throat view.
(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
Marine and Offshore Applications 437
Exhaust
Mixing
air jets
Dilution Dilution
air fan air fan
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
438 The Coen & Hamworthy Combustion Handbook
Temperature
(plane 2)
2.216e + 003
1.742e + 003
1.269e + 003
7.963e + 002
3.232e + 002
(K)
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
Marine and Offshore Applications 439
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
440 The Coen & Hamworthy Combustion Handbook
FIGURE 17.11
Hamworthy Combustion’s triple 120 m-ton/h (132 U.S. ton/h) steam boiler module for FPSO.
FIGURE 17.13
Hamworthy Combustion fuel gas knock-out pot on an FSO.
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
443
444 Appendix A: Units and Conversions
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
449
450 Appendix B: Physical Properties of Materials
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
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
TABLE B.2 (continued)
456
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
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)
460 Appendix B: Physical Properties of Materials
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.
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
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
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.
Appendix B: Physical Properties of Materials 463
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
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,
CRC Press, Boca Raton, FL, 2001.
2. Kreith, F., The CRC Handbook of Mechanical Engineering,
CRC Press, Boca Raton, FL, 1998.
3. Russel, Bursdall & Ward Corp., Helpful Hints, Mentor,
OH, Chapter 23.
Appendix C: Properties of Gases and Liquids
465
466 Appendix C: Properties of Gases and Liquids
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)
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 3.4808 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
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
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)
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 3.9118 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
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
Appendix C: Properties of Gases and Liquids
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)
Temp, T Density, p Specific Heat, cp Kinematic Viscosity, v Thermal Conductivity, k Thermal Diffusivity, α
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
Appendix C: Properties of Gases and Liquids
600 1080 0.8938 0.0558 1076 0.257 30.02 32.31 0.04313 0.02491 0.4483 1.737 0.668
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
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)
Temp, T Density, p Specific Heat, cp Kinematic Viscosity, v Thermal Conductivity, k Thermal Diffusivity, α
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
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
481
482 Appendix C: Properties of Gases and Liquids
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
Appendix C: Properties of Gases and Liquids 483
TABLE C.15
Thermodynamic Properties of Steam: Pressure Table (SI Units)
Specific Volume (m3/kg) Internal Energy (kJ/kg) Enthalpy (kJ/kg) Entropy (kJ/kg K)
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
Appendix C: Properties of Gases and Liquids 485
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)
490 Appendix C: Properties of Gases and Liquids
TABLE C.18
Thermodynamic Properties of Steam: Pressure 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.
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)
492 Appendix C: Properties of Gases and Liquids
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
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
P = 250 psia (400.97°F) P = 275 psia (409.45°F) P = 300 psia (417.35°F)
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
Source: Cengel, Y. and Boles, M., Thermodynamics: An Engineering Approach, 7th edn., McGraw-Hill, New York, 2011.
a The temperature in parentheses is the saturation temperature at the specified pressure.
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
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
498 Appendix C: Properties of Gases and Liquids
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
Appendix C: Properties of Gases and Liquids 499
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)
504 Appendix C: Properties of Gases and Liquids
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
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
Appendix C: Properties of Gases and Liquids
–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
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
Sensible heat ratio = Qs/Qt
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
1.35
200
50 1.30
90
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
Sensible heat ratio = Qs/Qt
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
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.
Press, Boca Raton, FL, 2000.
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
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
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
Appendix D: Properties of Solids
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
Appendix D: Properties of Solids
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
Source: Janna, W.S., Engineering Heat Transfer, 2nd edn., CRC Press, Boca Raton, FL, 2000.
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
Press, Boca Raton, FL, 2000.
2. Kreith, F. (ed.), The CRC Handbook of Thermal Engineering,
CRC Press, Boca Raton, FL, 2000.
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
• 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