Sie sind auf Seite 1von 153

HEAT RECOVERY STEAM

GENERATORS FOR POWER


GENERATION AND OTHER
INDUSTRIAL APPLICATIONS
Report No. COAL R232
DTI/Pub URN 03/804

by
D Blood, S Simpson and R Harries, Powergen UK plc
D Dillon, Mitsui Babcock Energy Ltd
A Weekes, ME Engineering Ltd

The work described in this report was carried out under contract as part of
the Department of Trade and Industrys Cleaner Coal Technology Transfer
Programme. The Programme is managed by Future Energy Solutions. The
views and judgements expressed in this report are those of the Contractor
and do not necessarily reflect those of Future Energy Solutions or the
Department of Trade and Industry.

Crown Copyright 2003


First published March 2003

HEAT RECOVERY STEAM GENERATORS FOR POWER


GENERATION AND OTHER INDUSTRIAL APPLICATIONS
by
D Blood, S Simpson and R Harries, Powergen UK plc
D Dillon, Mitsui Babcock Energy Ltd
A Weekes, ME Engineering Ltd

Executive Summary
A Technology Status Review of Heat Recovery Steam Generators for Power
Generation and Other Industrial Applications has been completed for the
Future Energy Solutions (FES), on behalf of the UK Department of Trade and
Industry (DTI). The aims of the review were to:

Assess objectively the current state of development and application of heat


recovery steam generator (HRSG) technologies world-wide, identify
trends in future developments and assess the market potential for
exploiting these technologies.

Critically assess the strengths and shortcomings of existing technologies in


relation to commercial or near-commercial needs and provide information
on manufacturers, suppliers, developers, consultants and major users,
quantifying the potential future demand for such technologies in the world
on a regional basis.

Review current activities and capabilities of companies/organisations


working in the HRSG field, with particular emphasis on the UK.

Identify priority areas in which UK research, development and


demonstration (RD&D) activities could/should be focused to enhance the
market opportunities for UK manufacturers, developers and consultants,
including any small- and medium-sized enterprises (SMEs) who are active
but perhaps lack the resources needed to succeed.

The review has been led by Power Technology (part of Powergen UK plc) in
partnership with Mitsui Babcock Energy Limited and ME Engineering
Limited.
The main conclusions of the review are:

Current Status
Current state of the art utility scale HRSGs operate at high pressure
(HP) steam conditions of up to 124 bar/565C allowing the associated
combined cycle gas turbine (CCGT) plant to deliver electrical power at
a claimed net efficiency of up to 60%. The CCGT is built at a cost of
around 425/kW, with the HRSG accounting for 10-15% of this total,
and delivers energy at around 2.2p/kWh.
Operational experience with HRSGs shows that the inclusion of
specific design features and attention to detail during fabrication are
(ii)

essential to reliability. Key areas for improvement include build


quality, access for in-service inspection, control & instrumentation and
capability for flexible operation. Overall cycle chemistry philosophy
also needs to be more thoroughly considered at the design stage.
The current challenge for operational HRSGs, particularly in the UK,
is the need to cycle plant that has been designed for and/or previously
operated at base load. Many users are currently carrying out
investigations/trials and plant modifications.

New & Developing Technologies


Future increases in HRSG operating conditions will largely be dictated
by increases in gas turbine (GT) exhaust temperature. Areas of
significant interest are once through design, with its advantages for
flexible operation and the use of HRSG steam for GT blade cooling,
which presents significant challenges to HRSG design. In addition, the
use of HRSGs within integrated gasification combined cycle (IGCC)
plant is now approaching the status of commercial operation, although
the costs still remain relatively high. Other development areas include
modular design to reduce build costs, improving reliability and
improving access.
Industrial scale HRSG technology is relatively mature, but tends to
benefit from the trickling down of technology from utility scale
HRSGs.

World Wide Activities


Over the ten years, 1992-2001, the biggest sales of utility scale HRSGs
have been in the USA (with 48% of the market), the United Kingdom
and Japan (4% each). Key manufacturers were Alstom Power (14.2%),
Nooter/Eriksen (12.6%), Deltak (9.5%), NEM (7.7%) and Aalborg
Industries (7.5%).
Sales of industrial scale HRSGs were biased more towards Europe
(33% of sales in each of the USA and Europe), with the other leading
market being Asia and Australasia (excluding China) with 19%.

Market Potential
Whilst the utility scale HRSG market has been healthy in recent years,
there is a predicted sharp downturn in the HRSG market in the shortmedium term due to plant over capacity. The situation is not expected
to pick up again until around 2007-2011. Key future HRSG markets
are seen as the USA and China (via IGCC).
For industrial scale HRSGs, the European market is depressed due to
falling electricity prices and rising gas prices. However potential
markets include Russia, Central and Eastern European countries,
Turkey and the Middle East. In the USA, despite problems on the
utility scale, there are still opportunities for development of combined
heat and power (CHP) schemes on industrial sites.

(iii)

The current surplus of generating capacity in the UK and fluctuations


in the price of natural gas have led to a low requirement to build large
scale power generation plant within the UK, although plant
performance upgrade opportunities are present.
The combination of the New Electricity Trading Arrangements
(NETA) and a high natural gas price has dramatically reduced the
market for new HRSGs / CHP schemes. Enhanced government support
for CHP is required if its target of 10GWe by 2010 is to be met.

UK Activities
There are a number of utility-scale HRSG turnkey contractors
operating within the UK (e.g. Alstom Power, Mitsui Babcock, Foster
Wheeler Energy Ltd, Mott MacDonald, Nooter/Eriksen-CCT Ltd and
Siemens KWU). Whilst Mitsui Babcock has its headquarters and
manufacturing facilities in the UK, the remainder have their
headquarters and manufacturing facilities overseas or subcontract the
manufacture. Thermal Engineering International Ltd - Greens (TEI
Greens) is the largest independent manufacturer of utility-scale
HRSGs in the UK and has manufactured utility-scale HRSGs for most
of the worlds leading boiler designers/makers on both domestic and
export projects.
Wellman Robey, BIB Cochran, ME Engineering and TEI Greens are
UK companies of UK origin with the capability to design and supply
industrial HRSGs. Wellman Robey and BIB Cochran manufacture
shell type boilers and have UK manufacturing facilities. ME
Engineering supply water tube or shell boilers manufactured outside
the UK. TEI Greens design and manufacture industrial HRSGs of
water tube or smoke tube design in the UK. The UK arm of
Nooter/Eriksen also supplies industrial HRSGs although its design
capability is based in the US.
In such a competitive HRSG market, licensing agreements and
collaborative partnerships have been necessary in order for companies
to maintain the ability to compete. Under such conditions the
requirement to be continually developing new technologies is vital.
Areas of current research include once through technology, cyclic
operation, new forms of gasification and novel low emission power
cycles.

(iv)

CONTENTS
Page
1 Introduction ................................................................................................ 1
1.1 Scope of Report.................................................................................... 1
1.2 The Technology Status Review Partners ............................................. 2
2 HRSG Technologies ................................................................................... 3
2.1 Introduction.......................................................................................... 3
2.2 Utility vs. Industrial HRSGs ................................................................ 3
2.3 Background to the Development and Use of HRSG Technology ....... 3
2.4 Utility Scale HRSG Technology.......................................................... 6
2.4.1 Justification for Using Gas Turbine HRSGs ................................. 6
2.4.2 Technical Considerations for Utility HRSG Design and their
Economic Implications ............................................................................... 8
2.4.3 Requirement for Finned Tube ..................................................... 10
2.4.4 HRSG Circulation and Configuration ......................................... 12
2.4.5 Significance of HRSG Pressure Levels and Reheat in Increasing
Cycle Efficiency ....................................................................................... 16
2.5 Alternative Modes of Operating HRSGs ........................................... 18
2.5.1 Supplementary Firing Mode........................................................ 18
2.5.2 Auxiliary Firing Mode ................................................................ 19
2.6 Industrial Scale HRSG Technology................................................... 19
2.6.1 Water Tube Designs .................................................................... 19
2.6.2 Smoke Tube Designs................................................................... 20
2.6.3 Once Through Steam Generators ................................................ 20
2.6.4 Design Aspects............................................................................ 22
3 Current Status of HRSG Technologies..................................................... 24
3.1 Introduction........................................................................................ 24
3.2 Specification and Design of HRSG Plant .......................................... 24
3.2.1 Design Code ................................................................................ 25
3.2.2 Quality of Supply ........................................................................ 26
3.2.3 Pressure Part Materials................................................................ 26
3.2.4 Seamless Pressure Part Components........................................... 27
3.2.5 Stub to Header Weld Detail ........................................................ 27
3.2.6 Tube Finning Procedure .............................................................. 29
3.2.7 Header End Cap Design .............................................................. 29
3.2.8 HRSG Access.............................................................................. 30
3.3 Operational Issues with HRSG Pressure Parts................................... 31
3.3.1 Preheaters and Economisers........................................................ 31
3.3.1.1 Steaming .................................................................................. 31
3.3.1.2 Thermal Fatigue ....................................................................... 31
3.3.1.3 External Dew Point Corrosion ................................................. 33
3.3.2 Evaporative Circuits.................................................................... 33
3.3.2.1 Drum Sizing & Design............................................................. 33
3.3.2.2 Case Study: Drum Sizing and Design...................................... 34
3.3.2.3 Drums and Flexible Operation................................................. 35
3.3.2.4 Evaporator Headers & Tubing ................................................. 36
3.3.2.5 Once-Through Evaporators...................................................... 36
3.3.3 HP Superheater and Reheater...................................................... 37
3.3.3.1 Condensate Formation and Quenching.................................... 37
3.3.3.2 The Superheater and Reheater on Cold Starts ......................... 37
(v)

3.3.3.3 Case Study: Excessive Ramp Rates and Through-Wall


Differentials on Superheater and Reheater Headers ............................. 38
3.3.3.4 Reducing the Risks of Reheater and Superheater Fatigue ....... 41
3.3.4 Attemperators .............................................................................. 41
3.3.5 Tube Banks.................................................................................. 42
3.3.6 Pipework and Pipework Support Systems .................................. 44
3.3.6.1 Case Study: Problems with Pipework and Pipework Support . 44
3.3.7 Valves & Fittings ........................................................................ 45
3.4 Operational Issues with HRSG Non-Pressure Parts .......................... 46
3.4.1 Ducting & Casing........................................................................ 46
3.4.1.1 Hot Casing Design ................................................................... 46
3.4.1.2 Case Study: Hot Casing Design............................................... 47
3.4.1.3 Cold Casing Design ................................................................. 48
3.4.1.4 Case Study: Cold Casing Design ............................................. 49
3.4.1.5 Warm Casing Design ............................................................... 49
3.4.1.6 Case Study: Warm Casing Design........................................... 49
3.4.2 Expansion Joints.......................................................................... 50
3.4.2.1 Case Study: Expansion Joints .................................................. 51
3.4.3 Bypass Duct and Damper............................................................ 53
3.4.4 Burners ........................................................................................ 54
3.4.5 Silencers ...................................................................................... 55
3.4.6 Case Study: Silencers .................................................................. 55
3.4.7 Stack Damper .............................................................................. 56
3.5 Cycle Chemistry Issues in HRSG Plant............................................. 57
3.5.1 Introduction ................................................................................. 57
3.5.2 Steam Droplet Carryover ............................................................ 59
3.5.3 Flow Accelerated Corrosion........................................................ 59
3.5.4 Design Features Promoting FAC. ............................................... 60
3.5.4.1 Horizontal Tubed HRSGs ........................................................ 60
3.5.4.2 Vertical Tubed HRSGs ............................................................ 62
3.5.4.3 Combined LP Drum / Deaerator or LP Drum / HP Feedwater
Tank 62
3.5.5 Phosphate Hideout in Drum Evaporator Circuits........................ 63
3.5.6 Organic Amines for Feedwater pH Control ................................ 65
3.5.7 Effects of Organic Matter in the Steam / Water Cycle ............... 66
3.5.8 Cycle Chemistry Selection .......................................................... 66
3.5.9 Steam Purity from HRSGs .......................................................... 68
3.5.10 Once Through HRSG Chemistry ................................................ 69
3.5.11 Dew Point Corrosion on the Gas Side of HRSGs. ...................... 69
3.6 Control and Instrumentation Issues on HRSG Plant.......................... 70
3.7 Flexible Operation of HRSG Plant .................................................... 71
3.8 HRSG Costs, Reliability and Maintenance........................................ 73
3.8.1 Capital Cost ................................................................................. 73
3.8.2 Operating Costs ........................................................................... 74
3.8.3 Reliability .................................................................................... 74
3.8.4 Maintenance ................................................................................ 74
3.9 Industrial HRSG Applications ........................................................... 75
3.9.1 Industrial Gas Turbine HRSGs ................................................... 75
3.9.2 Reciprocating Engine Exhaust Gas Boilers ................................ 75
3.9.3 Heat Recovery from Other Industrial Exhaust Gases ................. 76
(vi)

3.9.4 Process Integrated HRSGs .......................................................... 76


3.10 Conclusions........................................................................................ 77
4 New and Developing Technologies.......................................................... 78
4.1 Introduction........................................................................................ 78
4.2 Developments in HRSG Design ........................................................ 78
4.2.1 Utility Scale Once Through HRSG Designs ............................... 78
4.2.2 Industrial Scale Once Through HRSG Designs .......................... 79
4.2.3 Reliability Improvements............................................................ 79
4.2.4 Modularity and Improved Maintenance Features ....................... 81
4.2.5 Control and Instrumentation........................................................ 81
4.2.6 Highly Fired HRSG Designs....................................................... 81
4.3 Improvements to Cycle and Other Plant Components....................... 82
4.3.1 Steam Cooled Turbine Blades..................................................... 82
4.3.2 Fuel Heating ................................................................................ 84
4.3.3 Gas Turbine Steam Injection for Power Augmentation .............. 85
4.3.4 Gas Turbine Inlet Air Chilling .................................................... 86
4.3.5 Increases in Gas Turbine Exhaust Temperature.......................... 86
4.3.6 Supercritical Technology ............................................................ 88
4.4 New Applications for HRSGs............................................................ 88
4.4.1 The Role of HRSGs in IGCC Plant............................................. 88
4.4.1.1 IGCC Plant Description ........................................................... 88
4.4.1.2 IGCC Plant Performance ......................................................... 94
4.4.2 Biomass Integrated Gasification Combined Cycle .................. 95
4.4.3 Microturbines .............................................................................. 96
4.5 Conclusions........................................................................................ 96
5 World-wide Activities .............................................................................. 98
5.1 Introduction........................................................................................ 98
5.2 Survey Responses .............................................................................. 98
5.3 Utility Scale Market Published Information ................................. 100
5.3.1 Source of Market Information................................................... 100
5.3.2 The HRSG Buyers..................................................................... 100
5.3.3 The HRSG Manufacturers......................................................... 100
5.4 Conclusions...................................................................................... 101
6 Market Potential ..................................................................................... 102
6.1 Introduction...................................................................................... 102
6.2 Market Survey.................................................................................. 102
6.3 Market Perception amongst Consultees........................................... 104
6.4 UK Market ....................................................................................... 105
6.4.1 Non-Technical Barriers in the UK Utility HRSG Market......... 106
6.4.1.1 Current Surplus of Generating Capacity in the UK ............... 106
6.4.1.2 Fluctuations in the Price of Natural Gas ................................ 107
6.4.1.3 Current Unpredictability of the UK Retrofit Market ............. 107
6.4.2 UK Industrial CHP Market ....................................................... 107
6.4.2.1 Stricter Consents Policy......................................................... 108
6.4.2.2 Natural Gas Prices.................................................................. 109
6.4.2.3 New Electricity Trading Arrangements ................................. 109
6.4.3 Other Barriers for UK Firms ..................................................... 109
6.4.4 Future Industrial CHP Market Potential in the UK and Mainland
Europe 110
6.4.5 Action to Stimulate the UK Market / Support the UK Industry 111
(vii)

6.5 North American Market................................................................... 112


6.5.1 Possible Non-Technical Barriers to Development of Combined
Cycle Technology within the USA......................................................... 113
6.5.2 The US CHP Market ................................................................. 113
6.6 Chinese Market ................................................................................ 113
6.6.1 Possible Non-Technical Barriers to Development of Combined
Cycle Technology within China ............................................................. 115
6.6.2 Complex Administrative Procedures ........................................ 115
6.6.3 Low Institutional Capability...................................................... 115
6.6.4 Environmental Emission Controls ............................................ 116
6.6.5 Financial Issues ......................................................................... 116
6.6.6 Maturity of the Technology....................................................... 116
6.6.7 Issue of Intellectual Property..................................................... 116
6.6.8 Long-Term Collaboration ......................................................... 116
6.7 Conclusions...................................................................................... 117
7 UK Activities.......................................................................................... 118
7.1 Prospects of UK Suppliers and Manufacturers in the Global Market
118
7.2 UK Capabilities in HRSG Design, Manufacture and Supply UtilityScale 118
7.3 UK Capabilities in HRSG Design, Manufacture and Supply
Industrial-Scale .......................................................................................... 120
7.4 Research, Development & Demonstration Activities undertaken in the
UK 123
7.4.1 HRSG Once-Through Technology............................................ 123
7.4.2 Effects of Cyclic Loading on HRSGs ....................................... 124
7.4.3 Various Novel HRSG Designs.................................................. 124
7.4.4 Ongoing Gasification Development.......................................... 125
7.5 Conclusions...................................................................................... 127
8 Overall Conclusions ............................................................................... 128
9 References .............................................................................................. 131
10
Acknowledgements ............................................................................. 137
11
Appendix A ......................................................................................... 138

(viii)

Figures
Figure 1: Schematic of a CCGT plant (Courtesy of Mitsui Babcock Energy
Ltd). ............................................................................................................ 5
Figure 2: Gas turbine combined cycle (Courtesy of Innogy plc)..................... 7
Figure 3: Temperature profile of a single pressure HRSG (Courtesy of Mitsui
Babcock Energy Ltd).................................................................................. 9
Figure 4: Finned tube manufacture (Courtesy of Mitsui Babcock Energy Ltd).
.................................................................................................................. 11
Figure 5: Finned tube bundle (Courtesy of Mitsui Babcock Energy Ltd). .... 11
Figure 6: High frequency welded segmented fin (Courtesy of Mitsui Babcock
Energy Ltd)............................................................................................... 11
Figure 7: Typical modern horizontal gas-flow, natural circulation HRSG
(Courtesy of Mitsui Babcock Energy Ltd). .............................................. 13
Figure 8: Typical modern vertical gas-flow, assisted circulation HRSG
(Courtesy of Mitsui Babcock Energy Ltd). .............................................. 13
Figure 9: Full (a) and partial (b) penetration welds (Courtesy of Power
Technology).............................................................................................. 28
Figure 10: Finning defect in an L type finned tube (courtesy of Mitsui
Babcock Energy Ltd)................................................................................ 29
Figure 11: Forged (a) and flat plate (b) type end caps (Courtesy of Power
Technology).............................................................................................. 30
Figure 12: Drum working volume vs. heated evaporator volume for HP, LP
and IP circuits on several utility-scale HRSGs (Courtesy of Power
Technology).............................................................................................. 33
Figure 13: Fluctuation in drum level during start-up - shaded area represents
the drum level (courtesy of Power Technology). ..................................... 35
Figure 14: The HP circuit of the once-through HRSG at Cottam Development
Centre (Courtesy of Siemens KWU). ....................................................... 36
Figure 15: Temperature of the HP Superheater outlet header shell and tube
stubs on a warm start (Courtesy of Power Technology). ......................... 39
Figure 16: Temperature differential between the HP Superheater outlet header
shell and tube stubs during a cold start (Courtesy of Power Technology).
.................................................................................................................. 39
Figure 17: Reheater header through-wall temperature differentials during a
cold start (Courtesy of Power Technology).............................................. 40
Figure 18: Comparison of start-up and shut-down temperature differentials
(Courtesy of Power Technology). ............................................................ 41
Figure 19: Tube distortion due to quenching caused by poor attemperator
control (Courtesy of Power Technology). ................................................ 42
Figure 20: Fretting of a tube where it passes through a tube sheet (Courtesy of
Power Technology)................................................................................... 43
Figure 21: Pipe support movements - predicted and measured (Courtesy of
Power Technology)................................................................................... 45
Figure 22: Internal bracing and external stiffeners on a hot casing design
(Courtesy of Powergen UK plc). .............................................................. 47
Figure 23: Destruction of the internal insulation system in a GT diffuser duct
(Courtesy of Power Technology). ............................................................ 50
Figure 24: Expansion joint flange distortion (Courtesy of Power Technology).
.................................................................................................................. 52
(ix)

Figure 25: Thermal image of expansion joint (Courtesy of Powergen UK plc).


.................................................................................................................. 53
Figure 26: Silencer damage downstream of the GT exhaust in a CHP plant
(Courtesy of Power Technology). ............................................................ 56
Figure 27: Example of flow accelerated corrosion damage to tube (Courtesy
of Power Technology). ............................................................................. 61
Figure 28: Diagram illustrating point at which worst FAC damage occurs
(Courtesy of Power Technology). ............................................................ 61
Figure 29: Decrease in wall thickness at bends associated with flow
accelerated corrosion (Courtesy of Power Technology). ......................... 62
Figure 30: Temperature profile immediately above LP evaporator (Courtesy
of Power Technology). ............................................................................. 65
Figure 31: Condensate preheater deposits (Courtesy of Power Technology).70
Figure 32: Effect of gas turbine cooling methods on efficiency (Courtsey of
Innogy plc) [6]............................................................................................ 83
Figure 33: Diagram of a typical IGCC plant with dry feed gasifier (Courtesy
of Mitsui Babcock Energy Ltd) ................................................................ 93
Figure 34: Average percentage of business by geographical market........... 103
Figure 35: Percentage of enquiries coming from geographical market
(Courtesy of ME Engineering Ltd)......................................................... 103
Figure 36: Average percentage of business by HRSG size.......................... 104
Figure 37: UK generation capacity available for the 2002/3 winter............ 106
Figure 38: Installed CHP capacity in the UK 1997-2001. ........................... 108
Figure 39: GAS ZEP concept (courtesy of Mitsui Babcock Energy Ltd).... 125

(x)

1
1.1

INTRODUCTION
Scope of Report

Towards the end of 2001, FES, on behalf of the UK Department of Trade and
Industry (DTI), invited proposals for the completion of a Technology Status
Review of Heat Recovery Steam Generators for Power Generation and Other
Industrial Applications. The aims of this review were to:

Assess objectively the current state of development and application of


HRSG technologies world-wide, identify trends in future developments
and assess the market potential for exploiting these technologies.
Critically assess the strengths and shortcomings of existing technologies in
relation to commercial or near-commercial needs and provide information
on manufacturers, suppliers, developers, consultants and major users,
quantifying the potential future demand for such technologies in the world
on a regional basis.
Review current activities and capabilities of companies/organisations
working in the HRSG field, with particular emphasis on the UK.
Identify priority areas in which UK research, development and
demonstration (RD&D) activities could/should be focused to enhance the
market opportunities for UK manufacturers, developers and consultants,
including any small- and medium-sized enterprises (SMEs) who are active
but perhaps lack the resources needed to succeed.

The review aimed to address various types of HRSG technologies - from


large-scale units suitable for use in combined cycle gas turbine (CCGT) and
integrated gasification combined cycle (IGCC) power plant to medium / small
size units generating steam for small-scale power / combined heat and power
and other industrial process applications. However for the purposes of this
review, heat exchangers for producing hot water were not included.
The report is broken in to the following sections:

HRSG technologies: brief introduction to the development of HRSGs and


their applications; outline of the main designs and their features.
Current status of HRSG technologies: review of current commercial
applications and operating regimes; operational experience of current
HRSG designs.
New and developing technologies: discussion of emerging HRSG designs
and applications
World-wide activities: review of companies active in the HRSG market
world-wide and their capabilities.
Market potential: assessment of world-wide trends in the HRSG market;
measures required to stimulate the market and enhance opportunities for
UK companies.
UK activities: review of the capabilities of UK companies active in the
market

(1)

1.2

The Technology Status Review Partners

This review has been led by Power Technology (part of Powergen UK plc) in
partnership with Mitsui Babcock Energy Limited (MBEL) and ME
Engineering Limited. Given the respective activities of each company in
HRSG technology and markets, it was believed that a co-operative effort
would produce a better value report than could be offered by any of the
individual participants. The rationale in using this partnership was that MBEL
and ME would be able to offer their experience as suppliers of utility scale and
industrial scale HRSGs respectively, whilst Powergen would be able to offer
its experience as a user.
Power Technology is the focus of engineering and scientific consultancy
within Powergen, and employs approximately 230 specialist scientists and
engineers. It provides technical support to a large number of CCGT projects
and smaller CHP projects including the HRSGs, both within Powergen and for
external customers. Powergen alone currently has 4 CCGT sites (two of which
are joint ventures) and 14 CHP sites featuring HRSGs (with one fluidised bed
combustion plant under construction). However, Power Technology also
supports many external sites, and has experience of many leading
manufacturers.
Mitsui Babcock is a major energy engineering company incorporated in the
UK, and since 1995, a wholly owned subsidiary of Mitsui Engineering &
Shipbuilding of Japan. The company is a technology leader in large fossil fuel
steam generating plant, and specialises in the design, engineering,
manufacture, construction and commissioning and after sales servicing of high
efficiency, high availability coal, oil and gas fired boilers for the power
stations of electricity generating companies world-wide. The company is also
a major manufacturer and supplier of heat recovery steam generating plant,
industrial, fluidised bed and other clean burn coal fired boilers, coal milling
plant, flue gas desulphurisation plant and low NOx burners.
ME Engineerings UK business in steam generation has been established for
in excess of 50 years. As a part of Thermax group, ME have extended
experience in the design manufacture and supply of wide range of industrial
steam generating plants based on coal, liquid fuel, biomass and waste heat.
Principal product and project activities comprise industrial boilers, heaters, cogeneration, water and waste management, and absorption cooling.

(2)

2
2.1

HRSG TECHNOLOGIES
Introduction

Many industrial processes and power generation systems produce a high


temperature exhaust gas. Gas turbine exhaust temperatures are typically in the
range 425 - 600C, while the exhaust from a sponge iron plant for example
may be at 1000 - 1200C. If this hot exhaust is released straight to atmosphere
it clearly represents a large loss of energy. For a typical gas turbine the
exhaust heat loss might be greater than 60% of the fuel lower heating value
(LHV). In other industrial processes, the process requirements themselves may
dictate that a gas stream may need cooling. If some of this heat loss can be
recovered and converted to useful energy, the process efficiency will be
increased with both economic and environmental benefits.
The design of HRSGs in Europe has evolved from conventional boiler
designs. The earliest boilers were of fire-tube (also termed smoke tube)
design. In these designs, the hot flue gas is passed through a set of parallel
small diameter tubes. The tubes are enclosed in a water filled shell hence the
alternative name of shell boiler. The heat transfer across the tubes from the
hot gas to the water boils the water to raise steam, which is piped off from the
top of the shell. As higher steam pressures and flows were demanded, the
shells had to become increasingly thick and a practical limit was reached.
Boiler explosions occurred with increasing regularity in the 19th Century
emphasising the need for safer alternatives. The alternative is the water tube
boiler in which the water / steam is contained in the small diameter tubes with
the hot gas flowing around them. This allows the use of much higher pressures
with greater safety. The first patent for a water tube design was taken out by
William Blakey in 1766, but James Rumsey came up with the forerunners of
modern designs with water and steam spaces linked by tubes running through
the firebox [1].
2.2

Utility vs. Industrial HRSGs

Nowadays HRSGs are employed in a number of applications. The largest units


are used in large combined cycle power plants recovering heat from gas
turbines (GTs). These are referred to as utility scale HRSGs. HRSGs are also
used behind other engines and in various industrial processes and these are
referred to as industrial HRSGs. HRSGs used behind small GTs in combined
heat and power applications are also often termed industrial. In this case the
distinction between utility and industrial scale is somewhat artificial units
serving GTs with output above around 50MWe are usually considered as
utility scale.
2.3

Background to the Development and Use of HRSG Technology

Whilst the evolution of the gas turbine owes much to the development of the
jet engine, its use for power generation in fact precedes its use in aircraft

(3)

propulsion, with the first commercial gas turbine, a 2 MWe Brown-Boveri


machine originally installed in Switzerland in 1939 [2].
The use of gas turbine HRSGs evolved from the requirement to provide a
significant improvement in the overall efficiency of a gas turbine generating
plant by utilising the heat available in the exhaust flow of the gas turbine.
With the thermal efficiency of the gas turbine inherently low due to the high
exit gas temperatures (425 - 600C) and high excess air levels (220 300%) in
the combustion products, the thermal energy remaining in the exhaust gas was
targeted for recovery via a heat exchanger system which circulated water and
generated steam, thus combining additional electricity from a steam turbine
generator. A schematic of a simple combined cycle system consisting of a
single gas turbine generator, a HRSG, a single steam turbine generator, a
condenser and the associated auxiliary systems is shown in Figure 1.
Considering the HRSG alone, a variety of different physical configurations are
available alongside differences in the method of circulation employed, the
number of steam side pressure levels achieved, and the specific mode of firing
selected. A description and discussion of the impact of these factors is
presented in the following sections.

(4)

Figure 1: Schematic of a CCGT plant (Courtesy of Mitsui Babcock Energy Ltd).

(5)

2.4
2.4.1

Utility Scale HRSG Technology


Justification for Using Gas Turbine HRSGs

Today, the employment of a HRSG results in an electrical output from the


combined gas turbine and steam turbine in the general region of ~ 30 to 50%
greater than the output available from the gas turbine alone [3]. Significantly,
this extra electrical output is obtained with no necessity for any additional fuel
input.
Combined cycle systems make use of a Brayton cycle gas turbine firing
natural gas or distillate oil and a Rankine cycle steam system to achieve
efficient, reliable power generation. The Brayton cycle has high source
temperature and rejects heat at a temperature that can be conveniently used as
the energy source for the Rankine cycle. Table 1 shows the energy utilisation
for a typical combined cycle plant.
COMBINED CYCLE PERFORMANCE
% OF FUEL INPUT
Fuel Input LHV

100

Gas Turbine Power


Gas Turbine Losses
Gas Turbine Exhaust heat

36
1
63

Stack Loss
Input to Steam

22
41

Steam Turbine Power


Steam Turbine Losses
Heat to Condenser

19
1
21

Gross Electric Power

55

Auxiliaries Power
Total Net Power and
Efficiency

2
53

Table 1: Typical modern day combined cycle performance.


The gas turbine may typically convert 36% of the fuel energy into power
leaving 63% as heat passing to the HRSG from the exhaust of the gas turbine
(typical mechanical electrical and heat losses in the GT accounting for 1%).
The HRSG captures approximately two thirds of the gas turbine exhaust heat
with the remaining third being lost in the exit stack. Finally 19% of the fuel
input is converted into power via the steam turbine with 1% lost in the turbine
and 21% of the fuel energy lost in the spent steam which is sent to the
(6)

condenser. The combined gross power of gas and steam turbines equates to
55% (LHV) of the fuel energy. Plant auxiliaries account for ~2% of the fuel
input finally leaving 53% as net output combined cycle efficiency. Therefore
the main justification for utilising HRSGs within utility power plants lies in
the clear benefit from superposition of the gas turbine Brayton cycle over the
steam turbine Rankine cycle (Figure 2) which results in an enhanced overall
thermal efficiency.

Figure 2: Gas turbine combined cycle (Courtesy of Innogy plc).


In terms of emissions to atmosphere, CCGT plant is significantly better than
conventional coal-fired plant. Table 2 shows average emissions data for both
plant types from the Powergen UK fleet during 2000, including part-load
operation and starts [4]. It should be noted that emissions vary considerably
between units depending on fuel composition, plant design, emission
abatement technology and, to a lesser extent, running regime.

(7)

Emission

Typical operating
CCGT
0.43
0.00

Typical operating
coal-fired plant
0.92
7.17

CO2, kg/kWh
SO2, g/kWh
(highly fuel dependent)
NOx, g/kWh
0.28
2.31
Particulates, g/kWh
0.00
0.18
Table 2: Comparison of emissions from operating CCGT and coal-fired plant.
Further general benefits associated with the gas turbine combined cycle
include increased plant flexibility and a relatively low capital outlay. Gas
turbines can be used independently to provide a rapid start-up, peaking service
with the HRSG boiler system usually brought from a cold start to full load
steam generation in approximately 60 minutes. In terms of capital, gas turbine
HRSG systems are relatively low due to the standardised components,
modular construction, rapid erection and minimum support system costs.
2.4.2

Technical Considerations for Utility HRSG Design and their Economic


Implications

The gas turbine HRSG is essentially a counterflow heat exchanger consisting


of a series of superheater, boiler (or evaporator), and economiser sections
arranged from the gas inlet to the gas outlet in order to maximise heat recovery
and supply the rated steam flow at the required temperature and pressure to a
steam turbine.
The critical temperature differences that influence the amount of heat transfer
surface are the pinch point and both the economiser and superheater approach
temperatures. The pinch point and approach temperatures are illustrated in
Figure 3 for a single pressure HRSG and are defined as:

Pinch point: The difference between the gas temperature leaving an


evaporating section and the temperature at which the boiling is occurring
(i.e. the saturated water temperature).
Economiser approach point: The difference between the saturated water
temperature in an evaporating section and the incoming feedwater
temperature.
Superheater approach temperature: The difference between the inlet
exhaust gas temperature from the gas turbine and the exiting superheated
steam temperature.

(8)

Figure 3: Temperature profile of a single pressure HRSG (Courtesy of Mitsui


Babcock Energy Ltd).
Considerable efforts are made by HRSG designers to obtain maximum levels
of heat recovery from the turbine exhaust gas. The principles applied for
designing HRSG equipment are in many ways similar to those used for
conventional utility boiler design.
In general the technical design of any gas turbine HRSG is centred on the
following five features [5] and their respective economic implications:

Allowable back-pressure: The HRSG cross sectional area significantly


influences the gas turbine back-pressure. Smaller, more compact HRSGs
require higher gas turbine back-pressures to drive through the flue gas,
however, whilst the size reduction may reduce HRSG cost, the
requirement to provide a higher pressure at the turbine exit has a
detrimental effect on gas turbine efficiency. (Typical values of gas turbine
back-pressures are 2.5 to 3.7 kPa in most units).
Steam pressure and temperature: The steam pressure and temperature
are selected to provide an economical design. Higher steam pressures lead
to increased system efficiency but can limit total heat recovery from the
flue gas in single pressure HRSGs due to the higher saturation
temperature. Multiple pressure HRSGs as discussed in Section 2.4.5 are
used to overcome this constraint.
Pinch point and superheater approach temperatures: Small pinch
point and superheater approach temperatures correspond to a lower
temperature difference between flue gas and the steam within the
exchanger pipework. As a result of these smaller temperature differences
the surface area required is much greater in order to produce the same heat
transfer. The direct consequence is that more material is used and hence
capital cost is seen to rise (typical values of pinch point and superheater
approach point are between 11 to 28 K and 22 to 33 K respectively).
(9)

Economiser approach temperatures: The economiser approach


temperature is typically set to avoid the economiser steaming at the design
point. (Typical values of economiser approach point are 6 to 17 K).
Furthermore the economiser inlet water temperature must be fixed at a
level above that of the acid dew point of the combustion gases so that
corrosion from sulphuric acid condensation is avoided (a typical value of
economiser feedwater temperature is ~120C when sulphur is present in
the fuel; see Section 3.3.1.3).
Stack outlet temperatures: As for the feedwater temperature in the
economiser, the minimum flue gas exit temperature (stack temperature)
has to be controlled to avoid the financial penalties associated with
designing against acid corrosion.

2.4.3

Requirement for Finned Tube

Consideration is also required on the actual tubing utilised to form the gas to
water/steam heat exchanger. The heat transfer rate between the tube and the
high density water on the inside of the tube is far greater than the transfer rate
between the tube and the low density flue gas passing on the outside. The
outside heat transfer rate is said to be controlling and therefore responsible
for the overall heat transfer rate. In the case of a HRSG this overall rate of
heat transfer is lower in comparison with a fired utility boiler, due to the lower
flue gas temperatures and the reduced effect of radiation. Therefore, in order
to increase the rate of heat exchange in the HRSG tubes, the surface area on
the outside of the tubes is extended by finning.
There are many variations of fin design available. A commonly employed
finning process is where the fin is fabricated from an L shaped strip of metal.
The longer leg of the strip is slit and the strip is wound and welded in a spiral
around the tube. This results in the slits of the protruding long leg spreading
out as the L is wrapped around the parent tube as illustrated in Figures 4 and 5.
Alternatively a plain strip can be high frequency welded on to the tube to give
I finned tube as shown in Figure 6. The differences between I and L
section finning are discussed in Section 3.2.6. With either method, the
resulting fin area can be several times the area of the bare tube.

(10)

Figure 4: Finned tube manufacture (Courtesy of Mitsui Babcock Energy Ltd).

Figure 5: Finned tube bundle (Courtesy of Mitsui Babcock Energy Ltd).

Figure 6: High frequency welded segmented fin (Courtesy of Mitsui Babcock


Energy Ltd).

(11)

The actual gas side temperature limit for finned tubes will be influenced by fin
material considerations. In general, if the inlet gas temperature to the HRSG is
below about 750 to 800C finned tubes may be used exclusively. However if
the HRSG is to be fired in a supplementary mode with the use of duct burners
(as described in Section 2.5) and gas temperatures in excess of approximately
750 to 800C are encountered, the rows of tubes in the hotter zones may well
require to be bare or have a reduced level of finning (see Section 3.4.4 for
discussion of operational experience).
2.4.4

HRSG Circulation and Configuration

A HRSG may have a gas pass which is either horizontal or vertical in


orientation. In the first case (Figure 7), the gas turbine exhaust is ducted
horizontally through the casing of the HRSG and then passes over topsupported tubes before being turned vertically to a stack. For this horizontal
gas flow case, the evaporator tubes are vertical thus allowing water circulation
in the evaporator by natural convection without the need for a circulation
pump. Whilst the evaporator tubes are vertical, the superheater and
economiser tubes for the horizontal gas flow can be either vertical or
horizontal and are usually chosen on the basis of providing the best drainage.
When the gas flow is vertical, the evaporator tubes are horizontal (Figure 8)
and in order to ensure a more consistent flow of water, circulation is generally
achieved by a pumped or forced circulation means. There are however some
exceptions and HRSGs have been built with vertical gas flows and horizontal
heating surfaces which by utilising elevated drums ensure adequate circulation
via a natural circulation mechanism.

(12)

Figure 7: Typical modern horizontal gas-flow, natural circulation HRSG


(Courtesy of Mitsui Babcock Energy Ltd).

Figure 8: Typical modern vertical gas-flow, assisted circulation HRSG


(Courtesy of Mitsui Babcock Energy Ltd).

(13)

Both HRSG orientations have essentially the same components included in the
scope of supply. These typically consist of:

An expansion joint at the gas turbine exhaust interface


An exhaust by-pass damper
A by-pass stack and silencer
An inlet transition duct with flow correctors
Duct burner (required for supplementary and auxiliary firing units)
Heat recovery steam generator modules
Steam drums
Access ladders and platforms
Exhaust stack

Each circulation method has its own advantages and disadvantages as outlined
in Table 3 (assuming that vertical HRSGs are pumped circulation).

(14)

Natural circulation (horizontal gas pass) HRSGs


Advantages
Disadvantages
Flow disruption is not flow
Reduced power requirements due to
compensated as in forced circulation
absence of boiler circulating pumps
system
No maintenance of circulating pumps,
May have slightly increased plot area
motors, motor and pump controls etc
under certain circumstances
required
Greater care to maintain cleanliness
Vertical or inclined tube boilers can be
required, particularly with dirty
more effectively drained
gases
Overall height of the boiler plant can be
Can be difficult to withdraw single
more easily restricted.
tube elements
Tube spacers instead of tube supports
can be employed. This minimises
problems with fretting.
Easy to provide a water-cooled
combustion zone or furnace
More forgiving of flow and temperature
maldistribution, especially for fired units.
Easier to incorporate supplementary firing
and a split superheater design.
Pumped Circulation (vertical gas pass) HRSGs
Advantages
Disadvantages
May allow slightly reduced plot area
Increased power consumption due to
under certain circumstances
circulating pumps
On very restricted sites boiler can be
Additional maintenance items:mounted directly over GT with vertical
Circulating pumps, motors, motor and
exhaust straight into boiler
pump controls etc.
Stack being supported off the already
Effectiveness of the draining is not as
elevated boiler structure, height and costs
efficient
are reduced
Generally lower water content than natural
Restrictions on boiler height difficult to
circulation unit of equivalent output
achieve
Water-cooled tube supports required
Horizontal tubes with vertical fins
with gas temperatures above 760C
provides a self cleaning surface
Easy to arrange fully drainable
Not as easy to provide a water-cooled
superheaters irrespective of the size of the
combustion zone.
bank
Ease of individual tube element
Difficult to add SCR and CO catalyst
withdrawal without disturbance of other
support systems.
tubes
Designing for high earthquake
considerations is considerably more
difficult.
Difficult to incorporate a split
superheater with supplementary firing.

Table 3: Characteristics of natural and forced circulation HRSGs.

(15)

Both types of technology can be instrumented and controlled to the same level
of automation, both incorporate with equal ease an emission control catalyst
for NOx and SOx, both hold similar records for plant life and reliability and
both take up a similar plot area.
In order to maintain a position within the marketplace, equipment which
essentially performs the same function must be similarly priced. Thus no
significant price difference between the natural and pumped circulation
technologies is apparent, although the market currently demands a 90-95%
preference for horizontal gas-flow designs.
An area where the technologies do differ is supplementary firing (see Section
2.5) as this is easier to incorporate into a natural circulation, horizontal gaspass design. A major superheated steam temperature control tool, the split
superheater design, is difficult to incorporate into a pumped circulation design.
2.4.5

Significance of HRSG Pressure Levels and Reheat in Increasing Cycle


Efficiency

As mentioned previously in Section 2.4.2, a HRSG may be designed for


operation with multiple, separate pressure water/steam circuits in order to
maximise heat recovery. Over the years, the pressure levels available have
increased from the early single level installations, through the dual pressure
systems of plants built in the early 1990s to the more recent triple pressure
HRSGs. The number of pressure levels incorporated within an HRSG and the
use of reheat has a direct effect on steam cycle efficiency. Therefore the use of
a triple pressure with reheat circuit can, for example, contribute directly to
the combined cycle efficiency of the plant.
Consider, as the turbine exhaust gas passes through the HRSG, heat is
transferred to the circulating water and steam via the heating surface. Having
passed through the HRSG, if the temperature of the flue gas leaving the stack
is lower then the greater the amount of transferred heat to the steam and
therefore the greater the level of heat recovered.
In the case of a single pressure cycle, water from the condenser enters the
HRSG at the cold end and is then heated until near saturation by the exhaust
gases. Following this the water then enters the evaporator circuit via the steam
drum where the flow is circulated and heat is transferred at constant
temperature. Finally the steam is superheated by the hottest flue gas before
passing to the steam turbine. However, whilst the high steam pressure is
required at the steam turbine in order to achieve a high steam cycle efficiency,
the choice of a high pressure level simultaneously limits the amount of heat
transferred to the steam. Therefore in order to generate high pressure useful
steam for the steam turbine and maximise the amount of heat transferred from
the flue gas, other pressure levels are required.
With a dual pressure cycle the high pressure circuit ensures high steam
pressure delivery whilst the low pressure circuit ensures that maximum heat is
extracted from the gas turbine exhaust gas. Thus the problem of balancing

(16)

maximum enthalpy capture with an efficient steam cycle is addressed. Due to


the nature of latent heat, the evaporation of steam by a gas will always be such
that a large temperature difference will develop between the streams. This
resulting loss in performance can however be minimised by the adoption of
further steam evaporation levels such as the triple pressure cycle.
In addition to introducing various pressure levels to the steam circuit the
possibility of re-heating steam which has initially passed through the high
pressure section of the steam turbine is also considered by HRSG designers.
Reheat aims to optimise the lower pressure end of the steam turbine
performance. When re-heat is utilised, the steam turbine performance is
superior to non-re-heat cycles, due to the increased temperatures specifically,
of the lower pressure steam supply. There are however consequences in
enhancing this temperature. For example the additional heat required in the
HRSG re-heater section results in lower HP steam production. Furthermore,
whilst the steam turbine performance is improved, the gas turbine output
degrades slightly due to the need to overcome the pressure drop associated
with the additional HRSG re-heat surface. In general re-heat adds control
complexity and potentially higher capital costs due to higher costs of piping,
controls and a suitable steam turbine. Some consideration is therefore required
prior to selection to ensure the most effective configuration is achieved.
In general, the overall cycle efficiency can vary in the range of around three
percentage points depending on whether a single pressure or triple pressure
with re-heat cycle exists [6] . The obvious disadvantage, however, in increasing
the number of HRSG pressure levels is the associated increase in capital costs
with each new pressure level.
The following trends have been previously highlighted [6]:

The single pressure non-reheat cycle has a low installed plant cost and is
envisaged as a sound investment when fuel is inexpensive, ash bearing and
with a high sulphur content (e.g. oil firing).
The dual pressure non-reheat cycle has a higher installed plant cost than
the single pressure non-reheat cycle and has proven in the past to be the
most economical choice when fuel is more expensive and clean burning
with little sulphur content (e.g. natural gas).
The upper range of pressure levels i.e. the dual pressure level with reheat
and triple pressure level with reheat are usually matched to gas turbines
with high exhaust temperatures. In this case, there is sufficiently high
temperature energy to the HRSG to make the reheat steam cycle practical.
Therefore, the non-reheat dual pressure cycle is common for older and
smaller gas turbines, with the modern generation of larger gas turbines
such as the GE 9FA lending themselves to steam cycles with an additional
pressure level and reheat capability.

Typically, the largest step achievable in enhancing cycle efficiency by means


of incorporating various pressure levels is 1.7% points (single pressure to a
dual pressure non-reheat system). A triple pressure with reheat cycle provides
a further 1.3% points improvement over the dual pressure non-reheat case and,
(17)

of this 1.3% points increase, 0.5% points is directly attributable to the reheat
line itself.
An obvious approach to enhancing cycle efficiency is to continue adding
further pressure levels to the HRSG system thus further increasing heat
recovery of the exhaust gas and lowering stack temperature. However, in
practice there is a limit to the incorporation of pressure levels due to the fact
that the low pressure steam produced has an insignificant contribution to the
steam turbines power output. Furthermore the stack temperatures themselves
must be maintained above acid dew point thresholds in order to prevent the
expense of designing against possible corrosion.
2.5

Alternative Modes of Operating HRSGs

Due to the fact that gas turbines operate with a high air throughput (~ three
and a half times stochiometric) a gas turbine exhaust contains sufficient
oxygen to support further combustion (approximately 15 % w/w is present in
the exit stream of the gas turbine compared to some 23 % w/w in air). As
described previously, HRSGs can be unfired in which case they only utilise
the sensible heat of the gas as supplied. However, they may be fitted with
additional firing equipment (grid burners) positioned in the exhaust gas stream
across the inlet transition duct. These burners are commonly fired with gas,
although oil burners can be utilised but tend to be avoided due to
complications with the atomisation of the fuel.
When duct burners are present, two additional HRSG modes of operation are
possible, these are known as:
Supplementary firing mode
Auxiliary firing mode
2.5.1

Supplementary Firing Mode

Supplementary-fired HRSGs involve further combustion of additional fuel in


the gas turbine exhaust gas by utilising duct burners. The result of this
additional firing being that the flue gas temperature is substantially increased
which in turn improves steam production and raises superheated steam
temperature.
Normally large gas turbines provide an exhaust gas at a maximum temperature
of ~600C. However, by incorporating supplementary firing into a standard
HRSG this temperature can be raised to ~815C. HRSG inlet temperatures
higher than 815C are achievable when firing with duct burners, but the walls
of the HRSG will then need to be lined with refractory for protection of the
steel casing. Furthermore, at temperatures above around 1100C water cooled
walls may be necessary. This increase in exhaust gas temperature to ~815C is
associated with an almost doubling of steam production and thus provides a
mechanism of altering steam production by means that are independent of the
gas turbine operation.

(18)

2.5.2

Auxiliary Firing Mode

Auxiliary fired HRSGs allow steam to be generated in the HRSG when the gas
turbine itself is not in operation. The main advantage of this mode is the
flexibility it provides operators, allowing maintenance to be undertaken on the
gas turbine whilst still generating electricity with the steam turbine.
Disadvantages of this mode are that firing is therefore undertaken with air at a
lower initial temperature than that supplied previously by the gas turbine
exhaust gas. Therefore the fuel input required to obtain full steam output is
greater than for the case where the gas turbine is in operation. For auxiliary
firing mode, a system is therefore required which allows sufficient airflow into
the HRSG when the gas turbine is not in operation. This usually consists of an
arrangement whereby the duct from the gas turbine exit is isolated when the
turbine is off-line and the combustion air is introduced into the HRSG either
via a separate upstream forced draft fan or an induced draft fan positioned
downstream of the HRSG.
2.6

Industrial Scale HRSG Technology

A wider variety of operating conditions and applications gives rise to a variety


of designs and specialised equipment additions for industrial scale HRSGs as
compared to utility units.
As at the utility scale, the driving force for the development of the industrial
scale HRSG is the desire to improve the overall efficiency of fuel use. For
example industrial scale HRSGs are commonly used in combined heat and
power (CHP) schemes. In 2001 the average electrical efficiency of all
operating UK CHP schemes was 20% (GCV) and the average heat efficiency
was 54% giving an overall efficiency of 74% [7]. For comparison in 1998 the
mean efficiency of fossil fuelled electricity generation before transmission
loses was 40% and the efficiency of typical UK boiler stock was around 75%
(GCV). Based on the above efficiencies, compared to separate generation of
electricity and heat, the CHP scheme would use only 81% of the energy input
[8]
.
In addition, in some industrial applications there is in any case a process need
to cool a fluid flow, and the HRSG allows some use to be made of the rejected
heat.
2.6.1

Water Tube Designs

Many industrial HRSGs are broadly similar in design to utility scale units and
the same design criteria as discussed above apply. They are usually simpler,
often having just a single working pressure and no reheat. Some units
supplying steam to a steam turbine may have a second pressure level to allow
improved heat recovery with the lower pressure steam being admitted to the
turbine part way down the casing at the appropriate pressure level. As with the
larger units, they may have a horizontal or a vertical gas path, with hot or cold
casing and natural or forced circulation. In some applications, water wall
designs are used. This avoids the need for a refractory lining and is
(19)

particularly advantageous in applications where the exhaust gas is corrosive


and would attack a refractory lining. It is also sometimes used in units with a
high degree of supplementary firing.
2.6.2

Smoke Tube Designs

In some applications in which pressures and flows are lower, it is possible to


use the smoke tube (also known as fire tube or shell boiler) design in
which the hot gas is circulated through tubes within a water /steam filled shell.
The smoke-tube design has the advantage of simplicity, ease of construction
and lower capital cost. It removes the need for a separate steam drum and the
need to consider circulation and the provision of boiler circulation pumps. It is
therefore favoured in many smaller scale applications, especially in small
scale CHP schemes. Smoke tube HRSGs are available as factory built package
units, but these packages are usually limited to steam pressures of around 18
barg. The smoke tube design is also favoured in process applications where
there is a high gas side pressure.
However smoke tube designs are limited to the production of saturated steam
as the water and steam always exist within the same compartment (although a
separate superheater could be fitted). The stress in the shell due to the internal
pressure increases with diameter. Smoke tube designs are therefore limited in
their steam flow and pressure capabilities compared to the water tube design.
The multitude of small diameter gas passes means that they are not easily
cleaned and are prone to dust fouling. They are therefore not well suited to
applications where there is a high dust load in the gas. High gas velocities may
be experienced, especially if some tubes start to block increasing the flow
through the remaining tubes, which can lead to erosion problems.
Care needs to be taken in designs for higher temperature use. Steam
blanketing may occur in the upper regions of the shell reducing the cooling
flow of the water / steam. This is especially true in a vertical design where a
steam blanket can form over the bottom surface of the top tube plate. This can
allow high temperatures to occur in the top tube plate. This is dangerous if
such high temperatures have not been taken into account in the design of and
material selection for the top tube plate.
2.6.3

Once Through Steam Generators

Once through steam generators (OTSGs) are available at the industrial scale as
well as the utility scale and in some markets have achieved reasonable
penetration. The main supplier has been Innovative Steam Technologies (IST),
who have installed 65 units, mainly in the US [9]. However with four notable
exceptions, these have generally been on GTs of 50MW or less, and on the
>50MW plants, steam pressures have been relatively low (< 40 bar). The IST
design is not widely used in Europe yet. The IST design makes extensive use
of high temperature rated alloys that allow it to stand exposure to full GT
exhaust temperature when dry. The advantages of ISTs OTSG technology are
perceived as being [9]: (20)

Cheaper erection costs and quicker erection (for units less than ~50MW)
because all pressure welds can be done in the factory and there is no
welding of drums etc. to carry out on site. IST suggest that an OTSG can
be erected in only 25% of the erection time of a conventional HRSG.
Typically HRSGs will be equipped with a bypass damper/stack to allow
the GT to continue running if the HRSG trips. This is not needed with the
OTSG as the unit can run dry the tubes are able to withstand the full GT
exhaust temperature without cooling. The absence of a bypass stack and
damper gives a saving in capital cost. It will also give a slight efficiency
improvement as there is no GT exhaust gas loss through damper (normally
0.3 1.0% of exhaust gas gets lost through the damper and out of the
bypass stack).
There is no blowdown in the OTSG design, so there are no blowdown
losses, resulting in a higher efficiency.
Fewer parts, requiring less control and instrumentation typically only
50% of the valves that a conventional HRSG would need. This can result
in improved reliability and lower maintenance requirements.
The OTSG can be designed for fast maintenance with a single door for
access and fabricated entirely with single pass welds which are close
together for ease of access.
Remote operation as the unit is more robust - it will not be damaged if it
runs dry.
The nickel alloy tubes employed in the design are resistant to corrosion.
Carbon steel finning may suffer corrosion or the build up of deposits.
However these can be removed by running dry to bake them off. This
obviates the need for periodic acid cleaning. Alternatively corrosion
resistant alloys can be used for the finning as well.
The OTSG has a low water demand. It only uses around 16% of the water
that a conventional HRSG uses. This suits it to applications where good
quality water is not readily available e.g. areas of the Middle East. This
also allows a faster start up / shut down as there is a smaller reservoir of
water to heat up.
The OTSG has only ~40% of the weight of a conventional HRSG
(implications for transport costs and structural steelwork).
Smaller diameter and therefore thinner section components are used.
Thermal stresses are lower on start up, shut down and load varying. Higher
heating / cooling rates are acceptable. IST suggests that one of their dual
pressure OTSGs behind a 40 MWe GT could start from cold in under one
hour. The design is well suited to cyclic operation.
The material can handle a feedwater temperature as low as 15C
(compared to ~60C for carbon steel). This means that a plant can run with
a lower de-aerator temperature and feedwater temperature, which gives
higher boiler efficiency. However, it should be noted that any type of
HRSG can have material upgrades in the preheater if necessary or other
provisions for protection against cold end corrosion.
As the de-aerator can run at a lower temperature, vacuum de-aeration can
be employed, which uses less steam and at a lower pressure so giving
greater overall plant efficiency.

(21)

There are also a number of perceived disadvantages:

The advanced finning and construction techniques involved in the factory


fabrication of OTSG designs mean that few facilities are capable of
making them. This in turn means that the cost of transport to site may be
higher as manufacture cannot be carried out locally.
The extensive use of high temperature alloys increases capital cost.
One of the key advantages of the OTSG design is the fact that a bypass
duct is not needed, but the absence of one means that the GT must be shut
down in order to carry out HRSG maintenance.
The tube diameter is usually smaller to allow adequate strength at the
elevated temperatures that an OTSG may experience. This results in a
higher water side pressure drop and therefore a requirement for larger
feedwater pumps, which impose a larger parasitic load.
As impurities may not be blown down from a steam drum, the control of
water quality becomes more critical. A polisher is required in the circuit
adding to capital and operational costs.
The capital cost of OTSG designs may be slightly higher than
conventional drum designs.

Although once through designs have existed for many years now they are still
perceived as being novel with a higher risk than the conventional drum design.
This is especially true in Europe where there is little reference plant. The
design has been more widely used in the US, although they still form a very
small proportion of the total market. Many plants are now built under turnkey
contracts. Many of the advantages of the OTSG design are operational
whereas the capital cost may be slightly higher. This means that turnkey
contractors whose principal concern is normally capital cost rather than
operational advantage tend not to favour the design.
2.6.4

Design Aspects

Beside the thermodynamic design factors detailed above for utility HRSGs,
the design of the industrial scale HRSG is controlled by factors such as:

The dust load of the hot gas stream and its chemistry: in some process
applications the hot gas may have a high dust load and the dust chemistry
will influence the degree to which deposits build up on tube surfaces. In
such cases, means of online removal of deposits will be required, such as
rapping devices, sonic horns or soot blowers. Allowances for the reduction
of heat transfer rate due to the build up of deposits must be factored in to
heat transfer calculations by the inclusion of fouling factors.
The composition of the hot gas stream: compatible construction materials
may be needed (e.g. materials resistant to a corrosive gas stream).
Pressure of the gas stream: whilst for utility scale combined cycle
applications the hot exhaust gas is always at a low pressure to minimise
the back pressure on the gas turbine, in a process integrated application the
hot gas stream may be at high pressure. In such circumstances, and if a
high steam pressure is not required, it may be preferable or necessary to
use a smoke tube design.

(22)

Exhaust gas temperature: current GT exhaust temperatures tend to be in


the range 450-600C. Process exhaust gas temperatures may be far higher
necessitating the use of refractory linings or membrane panel water walls.
Steam requirement: the steam conditions produced from the HRSG may be
dictated not by the most efficient operating point of the HRSG, but by the
requirements of the process utilising the steam. Achieving the required
steam flow may also require supplementary firing.
Guaranteed availability of steam production: if a very high availability of
steam production is required to keep a downstream process running, the
design will be affected e.g. auxiliary firing may be provided to allow full
steam production to be maintained in the event of a gas turbine trip.

(23)

3
3.1

CURRENT STATUS OF HRSG TECHNOLOGIES


Introduction

This section covers the current design of HRSGs. It also discusses problem
areas, operational experience and typical costs. It is based on the experience of
users in specifying and operating both CHP and utility CCGT plant, and on
discussions with major equipment suppliers. Other applications of industrial
scale HRSGs are also discussed.
3.2

Specification and Design of HRSG Plant

The recent trend has been for the CCGT plant to be built under a turnkey
contract, to a functional specification with one overall contractor being
responsible for supplying the whole plant. Whilst this does have advantages to
the end user in terms of accountability, it does tend to mean that the choice of
HRSG supplier is often outside the users control.
An example of the role of the turnkey contractor is given by Siemens [10].
Siemens would typically carry out the overall thermal calculations and supply
the steam boundary conditions, customer requirements in terms of starts and
cycles and the number of pressure levels to the HRSG supplier, who then
carries out the detailed HRSG design. Siemens would also specify whether the
HRSG should be vertical / horizontal, fired / unfired and, if fired, whether this
should be supplementary / auxiliary. Siemens would then check the design,
ask the HRSG supplier to explain the rationale behind it and check the
calculations to make sure that it can perform theoretically. Siemens would also
thoroughly check the design of any non-standard plant with which the HRSG
supplier is unfamiliar.
Different turnkey contractors have different philosophies regarding erection.
For example, some use a single erection contractor for all plant items. The
Siemens approach [10] requires the contractor to design, manufacture and erect
the HRSG as they have knowledge, familiarity and experience of their own
plant and can install it correctly. This also means that responsibility is clearly
established. The boiler steel structure is also regarded as part of the boiler and
is included in the boiler suppliers scope.
The user can regain some control over the HRSG design through the
functional specification, although if this contradicts the suppliers standard
design philosophy, additional costs may result. The aim, therefore, is to allow
the supplier to provide his standard design as far as possible, whilst ensuring
that a minimum number of essential design features are incorporated into the
design. This will often involve meeting with the supplier at an early stage in
the tendering and/or design process to discuss the HRSG design and how the
required features can be incorporated.

(24)

Since the market for HRSGs is very competitive, most suppliers will provide
only the basic design unless otherwise required by the specification. In
addition, the suppliers warranty period will normally be limited to 1-3 years,
so the HRSG supplier will not always be focussed on the long-term
performance of the plant. In this respect, it is often the turnkey contractor
(who may supply a warranty of up to 5 years) or end user that dictate design
improvements, usually on the basis of previous experience.
Some of the main issues raised at the design/construction phase are: 3.2.1

Design Code

The Powergen requirement is normally that the HRSG be designed to an


agreed internationally recognised code. The preference would normally be for
the use of British Standards, EN Standards or TRD, principally due to their
recognition of fatigue as a damage mechanism, linkage of design stress to
design life and generally lower component wall thickness. However the
majority of HRSG suppliers provide their HRSGs to ASME (American
Society of Mechanical Engineers), and it often proves less problematic to stay
with the manufacturers standard practice and to concentrate on attaining
specific mechanical design features within that code. As implied above,
ASME Section 1 excludes any reference to low cycle fatigue as a damage
mechanism [11], so if this code is to be accepted, fatigue needs to be separately
assessed at the design phase, particularly if cyclic operation is envisaged.
Siemens experience also reflects this [10]. They regard ASME as a somewhat
over-conservative code that results in thicker components, and flexible plants
are therefore best designed to TRD (German Technical Rules for Boilers), BS
(British Standards) or EN (Euronorm). They also noted that inspectors are
involved in design and manufacture with the European codes, giving an extra
quality check, whereas ASME only requires sample designs to be checked.
Having said this, Siemens preference is for ASME as it is familiar to all boiler
manufacturers (European codes in particular are not familiar to Asian
manufacturers). As Siemens are under pressure to reduce delivery times
(Killingholme B was completed in 36 months but would now be done in
around 22), the use of unfamiliar codes can slow projects down. As a result,
no overseas Siemens projects have been designed to TRD. Occasionally the
choice of code is specified by the end-user. Another issue is that ASME only
allows the use of ASME materials and negotiations have to be made with the
inspector when other materials are wanted.
Similarly to Powergen, Innogy [12] is not prescriptive in its specification of
design codes (the description given is simply BS or equivalent). They also
recognise the fact that ASME has an inertia behind it and dominates, even in
Europe. The European codes have been written to exclude the legislative
aspects, as they are included in the PED (Pressure Equipment Directive).
ASME on the other hand is all encompassing. In retrospect, Innogy felt they
may have preferred to put more investment into HRSG projects up front (i.e.
improve the specification) to reduce operational problems experienced further
down the line. However, they did feel that it is sometimes more cost effective

(25)

to rectify issues early in a plants life than to change the specification in the
project development stage. As discussed above, this is because standardisation
from HRSG to HRSG allows turnkey contractors to save money and
additional features requested by the client deviate away from the turnkey
contractors norm.
As far as suppliers are concerned, opinion is divided. Nooter/Eriksen [13] carry
out all design to ASME unless specifically requested by the customer. NEM
[14]
fix the design code to the customers requirements, although once again,
they state that ASME has been the tendency over recent years. Standard Fasel
Lentjes have no preference, except for two-shift operation, where TRD is
preferred.
3.2.2

Quality of Supply

Quality of supply has become one of the major issues to the user. As the
HRSG market has become more competitive, fabrication has shifted to areas
where labour costs are low. The actual boiler fabricator is therefore often
several steps removed from the turnkey contractor, and in the worst case may
not have experience in HRSG fabrication. This means that the management of
quality by the turnkey contractor or main HRSG supplier is increasingly
difficult, and that changes to design agreed with the turnkey contractor may
take some time to filter down to the shop floor. The same problems can occur
on site, particularly if the HRSG erector did not supply the HRSG.
Siemens experience reflects this [10]. They do not tend to encounter problems
with the design of standard drum-type HRSG plant anymore - it is fabrication /
erection quality that causes greater concern and shop/site supervision is
therefore seen as crucial. For example, incorrectly installed baffle plates have
resulted in low thermal performance in the past due to gas bypassing. Site
welds are considered a particularly high-risk area. They also prefer to use
established designers although a boiler design may work thermodynamically
from a theoretical point of view, practical details can be critical and therefore
great emphasis is placed on experience.
Powergen experience is similar, and importance is placed on Powergen
personnel carrying out independent QA inspections during HRSG fabrication
and installation. This is believed to result in significant overall benefits in
terms of long term plant reliability.
3.2.3

Pressure Part Materials

HRSGs generally (but not always) operate at lower conditions than


conventional direct-fired boiler plant. The pressure parts are therefore
normally fabricated from well-proven materials. Modified 9%Cr (P91) and
12%Cr are now commonplace in HRSG design. P91 is the newer material,
having been in use in power plant for some 13 or so years and in HRSGs for
slightly less than this. P91 has excellent high temperature creep properties and
this is particularly beneficial in reducing the wall thickness of high
temperature components. This is often the driver for the use of this material,
particularly for headers, where design for cyclic operation is required.
(26)

Siemens [10] view was that materials are not really an issue, as higher
temperatures have been successfully handled on coal plant. It is also possible
to make a boiler more or less flexible with the same material by changing the
number of stubs into each header, for example. They believe the main issue
with materials is now quality of supply.
Nooter/Eriksen [13] use ASME materials as standard, subject to availability P91 is used as standard on high temperature pressure parts and they otherwise
tend to use commonly available materials. Standard Fasel Lentjes [15] and
NEM [14] give similar comments.
From an operators perspective, there are still some issues with P91, mostly
relating to the long-term life of the welds and inspection strategies to manage
this. Significant research / investigation is ongoing in this area.
3.2.4

Seamless Pressure Part Components

Powergens standard is for all tubes, pipes and headers to be fabricated from
seamless material, due to the perceived inherent risk in seam welded
components, in which the weld is subjected to the full hoop stress due to
internal pressure. This is particularly applicable in the creep range, where the
long term properties of the seam weld come into play.
Innogy [12] have likewise always used seamless headers. They have, however,
used ERW (electric resistance welded) tubing and have seen ERW tube
failures on some international projects. However, they believe that the
specification for ERW has since been tightened and that it is probably not
much cheaper than seamless (apparently only three companies world-wide
supply ERW). Innogys (now sold) Killingholme A plant contains ERW and
they acknowledged that the reliability of ERW is very sensitive to quality.
Siemenss approach [10] agrees with this. They prefer seamless headers and if a
manufacturer wants to use seam welded pipework, special requirements are
imposed. Siemens have not accepted a seam-welded header to date.
As far as suppliers are concerned, the situation is as follows. Nooter/Eriksens
standard [13] is to use seamless tubes for chrome alloy tubes and seamless or
welded ERW tubing for the carbon steel sections, with all headers being
seamless. However, in the UK, seamless tends to be used throughout due to
negligible difference in cost compared to seam-welded components. Standard
Fasel Lentjes [15] use seamless throughout. NEM use seamless in the creep
range and seamless or seam-welded below these temperatures.
3.2.5

Stub to Header Weld Detail

If incorrectly specified and executed, stub to header welds can be a major


source of long term unreliability. This is principally due to their great number,
difficulty of access (especially in horizontal HRSGs) and that fact that they are
usually required to accept bending stress as well as internal pressure stress. As
such, this is regarded as a key pressure part design feature.
(27)

Powergen normally require full penetration welds, particularly for cycling


plant (Figure 9a). Partial penetration welds (Figure 9b) are more susceptible to
corrosion fatigue cracking initiating at the unfused internal land. Whilst there
are plants operating flexibly with partial penetration welds, they are generally
best avoided other than for base-load operation, where starts are limited and
chemistry is well controlled. The use of full penetration welds can cause
problems for some manufacturers if it is not their standard, however in real
terms, there is little cost penalty in including this feature.

(a)

(b)

Figure 9: Full (a) and partial (b) penetration welds (Courtesy of Power
Technology).
Siemens [10] also prefer full penetration header welds, particularly if heavy
cycling is required. They believe that partial penetration welds are a risk as
acid can get trapped in the gaps following acid clean and cause corrosion. A
different cleaning procedure is therefore required. The unfused area can also
act as a site for corrosion fatigue cracking under cyclic operation (as
mentioned above). Siemens prefer fully penetrated stub welds that are actually
drilled afterwards to remove any protruding weld material. However, this is
more expensive. At one project, special welding equipment was developed to
ensure a consistent weld around its full circumference, despite the stub not
being normal to the header. Siemens stated that if full penetration welding is
not used then special requirements are imposed.
Innogy [12] has experienced corrosion fatigue failures where partial penetration
welds have been used and the headers have not been heat-treated. This has
resulted in header replacement. They are also aware of another project (not
one of their own) where this resulted in 5% failure on the site hydro test when
it had passed the same test in works. They have also seen inadequate provision
for differential expansion, which resulted in stub to header weld failures. In
this case the headers and tube-sheet in question were relatively close together
and the header pulled apart the tube-sheet on expansion.
Nooter/Eriksen [13] standard for all header connections is for set-on branch
connections with full penetration welds. This was upgraded in recent times
due to defects found between 1992 and 1996. The revised weld procedures
have proved successful. They have also moved to a straight through tube
design (i.e. no bends in the stubs) where flexible operation is required as this is
claimed to reduce bending moment to zero.
(28)

NEM [14] also utilise full penetration stub to header welds only. Standard Fasel
Lentjes [15] may choose partial penetration welds depending on the application
and design code however boilers for cyclic operation or designed to TRD
always have full penetration welds, as do the superheater modules on all types
of units.
3.2.6

Tube Finning Procedure

Powergen specifies the HF (high frequency I shaped fin) welding process


rather than the SF (standard frequency L shaped fin) welding process due
to problems experienced on an early CCGT project (Figure 10). Siemens [10]
also specify HF finned tubes (serrated or solid) for similar reasons. The defect
results from excessive heat input during manufacture causing a weakness in
the tube. An excessive heat input in the I type manufacturing process will
result in the fin failing rather than the tube.

Figure 10: Finning defect in an L type finned tube (courtesy of Mitsui


Babcock Energy Ltd).
Nooter/Eriksen [13] only uses serrated, HF finning. Standard Fasel Lentjes [15]
and NEM [14] do not express a preference on this issue, so may not have had
the poor experience with SF finned tubing of Powergen/Siemens.
3.2.7

Header End Cap Design

Forged header end-cap designs (Figure 11a) are preferred to the flat end-cap
type (Figure 11b). In particular, the latter is not favoured for use at elevated
temperatures (in the creep range), especially when constructed out of P91
material, due to the long-term effect of placing a significant bending moment
on the end cap attachment weld. A failure has occurred on such an end cap
design at a UK power plant (coal fired). It is believed that the failure was due
to a combination of material deficiency and the type of end cap design used
(flat plate), which gives a high bending stress at the weld.

(29)

(a)

(b)

Figure 11: Forged (a) and flat plate (b) type end caps (Courtesy of Power
Technology).
Siemens agree with the above [10] and go a step further. They believe that fully
formed end caps with relief grooves are the best design.
The Nooter/Eriksen [13] standard is for flat plate, set-in end caps with full
penetration welds to meet ASME (and BS1113), unless the client requests
otherwise. NEM tend to utilise a flat plate set on end cap. Standard Fasel
Lentjes [15] use either a flat plate end cap or hemispherical, depending on
application. This is therefore an area where discussion is often focussed during
the design phase.
3.2.8

HRSG Access

It often appears that little consideration is given to access for repair and
maintenance in HRSG design, particularly in horizontal gas-flow units [16].
One example of the difficulty this can create was experienced on a Powergen
CHP plant, where it was necessary to remove and replace 15 tubes to carry out
the repair of a single stub to header weld failure.
An approved manufacturers method for cutting headers to repair stub-toheader welds on vertical tube banks was rejected by Innogy [12]. However they
did have a positive comment with respect to the Alstom HRSG design which
has the advantage of harp construction (i.e. headers top and bottom with only a
single row of tubes) with headers connected by U-bends. In this case, when
repair is necessary, it is possible to take off the return bends on a single header
and lift out the entire harp.
This difficulty of access applies not only to repair work, but also to routine
inspections. The Written Schemes of Examination under the UK Pressure
Systems Safety Regulations require certain pressure part inspections and
assessments to be made during the life of the HRSG. This often requires the
taking of diametrical measurements for creep life assessment purposes or the
non-destructive examination of header butt welds, both of which can be
virtually impossible, particularly on a horizontal gas flow HRSG. Header
(30)

internal inspections may also be required, but dedicated inspection branches


are generally not fitted unless specifically requested, rendering this activity
more difficult and time consuming as the header end cap needs to be removed.
In many cases, these issues could be addressed at the design phase with little
additional cost, and experience is that there are considerable benefits in
discussing these with the supplier at an early stage.
3.3
3.3.1

Operational Issues with HRSG Pressure Parts


Preheaters and Economisers

3.3.1.1 Steaming
Though plant can be designed to minimise it, steam locking is a risk on startup in horizontal gas-flow designs, as many designs have high points in the
tubing that cannot be adequately vented [17]. This can result in differential
expansion, poor drum level control and chemical deposition with an associated
corrosion risk. In extreme cases, particularly where gas temperatures at normal
operating conditions are significantly higher than saturation temperature
within the economiser, the effects can be more severe. Economiser steam/air
locking was responsible for 50% economiser under performance at one
particular CHP site, and knock-on effects included high stack temperature, low
economiser flow, high attemperator spray flow, high primary superheater
outlet temperature and distorted economiser and superheater tubes due to
overheating [18]. This issue was eventually resolved by making significant
modifications to the economiser circuit, but demonstrates the need to assess
the risk of steaming, especially where the plant deviates significantly from the
manufacturers reference.
3.3.1.2 Thermal Fatigue
A survey covering around 50 HRSGs in the United States described low
temperature economisers as prone to both thermal and corrosion fatigue [19].
Thermal and corrosion fatigue are both usually start related and are
exacerbated under a flexible operation regime.
If the economiser inlet flow is not introduced gradually, a risk does exist on
cold starts, as feedwater is not required for topping up the drum until well after
the GT has been started. It takes some time for the drum swell to subside after
steam formation has commenced and by this time the economiser header and
tubes may be 100-150C above the temperature of the feedwater [11].
Depending on the header geometry, these mechanisms may be substantial
enough to cause low cycle fatigue under a flexible operation regime.
A similar situation can exist on a warm start-up. Whilst the use of a stack
damper has a net beneficial effect during overnight shutdowns (as described in
Section 3.4.7), it can also have some negative effects. When a unit is boxedup, heat tends to radiate from the hotter components and is absorbed by the
cooler components, with temperatures typically equalising at about the
(31)

temperature of the main (usually HP) evaporative circuit. The economiser and
preheater tube banks therefore warm up as heat is distributed from the
superheater (and reheater if installed). On start-up, this can result in a thermal
down-shock on the inner walls of the economiser and preheater inlet headers,
as cold feed water is pulled into them. These transients tend to be more severe
on HP economisers than LP economisers, and although they may not present a
short-term problem, the long-term effect on header integrity may be
substantial. Clearly, these effects are negligible following longer shutdowns,
as heat is gradually dissipated from the HRSG and the temperature difference
between the economiser/preheater headers and the feed water is much
reduced.
Another effect that can develop when off-load is thermal stratification within
the economiser headers, so that a temperature differential is actually
established in the vertical plane. The upper part of the header attempts to
expand relative to the lower part, causing header hogging. This can be
exacerbated on start-up when an initially gentle feed flow enters the header.
As this is relatively cold, it can flow along the header bottom increasing the
magnitude of the top-to-bottom differential. Depending on the degree of
flexibility built into the economiser antler tube design and support
arrangement, this can lead to failures of the stub-to-header welds under a
flexible operation regime.
The above effects can be minimised by the use of an economiser recirculation
system, which helps reduce stratification and dilutes the cold feed water flow.
These are not usually installed as original equipment, due to the additional
cost involved, but can prove a useful retrofit if a flexible operating regime is
envisaged.
Some HRSG designs, particularly those originating in the USA, may feature
economisers with tubes of different pressures (i.e. HP and IP or HP and LP)
connected to shared headers at the inlet and outlet, with internal pressure
plates being used to separate the two pressure sections. Whilst the temperature
difference between the circuits this may not be significant during on-load
conditions (typically 30C [20]), the two header sections may experience
differential expansion on start-up, especially if steam is produced in one drum
before it is produced in the other. Temperature differences of 150C have been
measured between stagnant tubes at gas temperature and tubes with relatively
cool feed flow passing through them [11]. This results in bending of the
pressure plate, high thermal stress at the welds adjacent to it and high hoop
stresses in the header itself as the HP section attempts to expand relative to the
IP or LP section. In fact stub-to-header connection failures have been known
to occur after 100-200 starts on a number of economisers featuring this
pressure plate design [21]. Leakage across an internal pressure plate has also
been observed [20]. Economisers of this design should be prohibited in new
HRSG plant specifications, particularly if frequent starts are anticipated.

(32)

3.3.1.3 External Dew Point Corrosion


External dew-point corrosion on start-up and during low load operation is a
risk on preheaters and, to a lesser extent, on economisers. The risk may be
increased when firing GTs with distillate fuel oil (which is often used as a
back-up supply) if the preheater and/or economiser inlet temperatures are
below the acid dew point temperature of the fuel. Typically (but not always) a
pumped condensate preheater recirculation system will be used to maintain
inlet temperature above dew point (typically 120C for distillate, 65C for
natural gas) by re-circulating a proportion of the preheater outlet water back to
the inlet. However for thermal performance reasons, these recirculation
systems are normally designed on the limit in terms of temperature, and it is
not uncommon to notice signs of corrosion of the last few (i.e. coldest) rows of
tubes.
3.3.2

Evaporative Circuits

3.3.2.1 Drum Sizing & Design


Drum sizing is an important issue, particularly for plant that will be required to
do large numbers of starts. Figure 12 shows the drum working volume plotted
against heated evaporator volume for HP, LP and, where appropriate, IP
pressure circuits at a number of stations. The drum working volume is
measured between the start-up/low-low level and the high-high level and the
heated evaporator volume is the volume of the inlet and outlet headers and the
evaporator tubes between them. Although there are small errors associated
with the volume of drum internals and the drum working volume can be
affected by the type of separation system employed and other design-specific
details, this gives a useful guide as to the suitability of a proposed drum size.

Figure 12: Drum working volume vs. heated evaporator volume for HP, LP
and IP circuits on several utility-scale HRSGs (Courtesy of Power
Technology).

(33)

Most drums evaluated on this basis have a drum working volume to heated
evaporator volume within the range 0.8 to 1.1. Circuits for which this ratio is
towards or beyond the upper end of this range may be oversized, and, as a
consequence, the drum itself may be unnecessarily thick. For circuits where
this ratio is lower than the range indicated, there may be difficulties
accommodating drum swell within allowable level limits on start-up or during
load changes. This is certainly true for Station A, which has comparatively
small drums on all three pressure circuits. As a consequence, a large quantity
of treated water has to be blown down during starts.
Innogy [12] stated that drum capacity is normally on the tight side, but that
drum level control is more of an instrumentation issue than a problem with
physical drum size.
If drum level is not controlled adequately, high blowdown flow, water
carryover to the superheater resulting in thermal fatigue damage or unit trips
can occur. Both Innogy and Powergen set different drum start levels for
different start types to make drum level control more manageable [12, 22].
3.3.2.2 Case Study: Drum Sizing and Design
Conventionally, vertical gas-flow (horizontal tube) HRSGs utilise assisted
circulation in the evaporative circuits. However, at a 3 x 330MW CCGT
station featuring triple pressure HRSGs with reheat sections, all three pressure
levels rely on natural circulation only. This removes the capital cost, operating
cost and loss of availability associated with the use of circulation pumps.
During commissioning, a comprehensive programme of cold, warm and hot
starts was conducted to ensure that the lack of circulation pumps would not
result in:

reverse evaporator flow at start-up


excessive fluctuations in drum level
thermal fatigue cracking of the evaporator header due to high thermal
transients on start-up
internal, on-load corrosion of evaporator tubing due to flow stratification
and/or tube dry out
overheating of HP evaporator tubing due to flow stratification and/or tube
dry out

The main issue was that of drum level control, which initially caused some
problems, leading to unit trips on start up. It was found that adjustment of the
start procedure to set different drum levels for each start type and allowing the
early venting of drums to bypass on warm and hot starts proved beneficial.
After the initial rise and subsequent drop in drum level following GT ignition,
the drum level became stable again as soon as the steam bypass valve was
opened (Figure 13 [23]). In addition, the blowdown system was modified such
that as well as looking at drum level, it also looked at rate of change of level.
This allowed the blowdown to be automatically opened if rapid changes in
drum level were detected. The units now start up reliably.

(34)

Figure 13: Fluctuation in drum level during start-up - shaded area represents
the drum level (courtesy of Power Technology).
3.3.2.3 Drums and Flexible Operation
Aside from the potential difficulties controlling drum level on start-up or
during load changes mentioned above, flexible operation could have other
potentially damaging effects of on drum integrity. This stems from the fact
that in a conventional (i.e. not once-through) HRSG design, the HP drum is
the thickest pressure part.
A HP Drum can be particularly susceptible to thermal fatigue damage during
cold starts. Prior to the start, both internal and external surfaces of the drum
essentially start from ambient temperature. As the evaporative section begins
to absorb heat from the GT exhaust gas flow, the internal surface temperature
of the drum remains close to that of the incoming water/steam mixture, but the
response of the external surface temperature is significantly delayed. As a
result, severe thermal stresses can be produced as the expansion of the internal
surface is constrained by the external surface. The magnitude of the stresses
established is a function of the temperature differential between the internal
and external surfaces, and this is dependent upon drum thickness. As drums
made of moderate or high alloy material can be thinner than drums of a lower
grade specification for the same pressure, temperature and geometry, they are
better suited to flexible operation.
The opposite effect occurs during unit shutdown. As the steam pressure falls,
the internal surface of the drum cools and contracts more quickly than the
external surface, putting the internal surface in tension. If the thermal stress
exceeds the yield strength of the material, it can result in crack
formation/propagation and thus consume fatigue life. For this reason, it is vital
that operators adhere to the start and shutdown procedures recommended by
the OEM in order to maximise component life, availability and reliability [24].
The through-wall differential of the HP drum and effects on the HP
superheater are the limiting factors on HRSG start rate.

(35)

3.3.2.4 Evaporator Headers & Tubing


The tube/riser connections to the drum and the design of the evaporator tubes
and headers on horizontal gas-flow HRSGs need to be sufficiently flexible if
the HRSG in question is expected to two-shift frequently. This is because the
front row evaporator tubes reach saturation temperature and establish a
circulation more quickly than rows of tubes towards the back of the evaporator
on hot/warm starts. This can result in large thermal stresses at the tube/riser
connections to the drum [11]. The horizontal, serpentine tube banks often
favoured on evaporators of vertical gas-flow HRSGs are inherently more
suited to flexible operation.
Flow accelerated corrosion (also known as flow assisted corrosion, FAC or
erosion corrosion and described in Section 3.5.3) is the largest cause of HRSG
tube failure in LP evaporators and preheaters as identified by NERC [19].
Powergen has experienced LP evaporator tube leaks due to FAC [25] and HP
evaporator inlet header to stub weld leaks caused by fatigue [26].
3.3.2.5 Once-Through Evaporators
At Cottam Development Centre (CDC), both HP and IP circuits are of a oncethrough design (i.e. there is no re-circulating steam drum), with a conventional
steam drum being used only on the LP circuit. Steam/water separators are used
on the HP and IP evaporative circuits during start-up (Figure 14). These are
smaller, thinner components than conventional drums which means that the
HRSG can start-up and shutdown more rapidly. This is aided by the fact that
once-through designs contain a lower water volume and therefore have a
lower thermal inertia than conventional drum-type designs. The HP separator
at CDC is only 45mm thick, compared to a 100mm thick steam drum for
equivalent conditions (580C and 160bar) with a conventional drum design.
The components that limit the start-time at CDC are therefore the HP
superheater and reheater final outlet headers.

Figure 14: The HP circuit of the once-through HRSG at Cottam Development


Centre (Courtesy of Siemens KWU).

(36)

3.3.3

HP Superheater and Reheater

3.3.3.1 Condensate Formation and Quenching


When a GT purge sequence is initiated, air from the compressor is forced
through the HRSG to clear any gas that could otherwise result in accidental
explosion. This purge process rapidly cools the HP superheater and, if
applicable, the reheater, particularly following a GT trip. This results in
condensate formation, which can collect in and thermally shock the lower
headers/tubes of horizontal gas-flow HRSGs. The damage is exacerbated if the
drains are not sized to remove the condensate at the rate at which it forms
and/or they are opened at the incorrect time or for insufficient duration. The
sizing of the drains, instrumentation and control for adequate removal of
condensate under these conditions is often an issue on earlier HRSGs designed
for base-load operation. Indeed, on virtually all horizontal gas-flow HRSGs,
there are no thermocouples installed on the lower headers that can be used
detect the presence of condensate. In these instances, additional
instrumentation may have to be installed and integrated into an automatic
drainage system. The blowdown tank should also be of sufficient size to
accommodate drainage of condensate during the purge sequence.
As steam flow is established during the subsequent start, any condensate that
has not been drained may be preferentially cleared from some tubes before it
clears from others. The tubes that have been cleared are cooled by the ensuing
steam flow whilst the other tubes stagnate until the condensate within them
also clears. This can result in tube distortion and differential expansion
between different parts of the header, which, depending upon the flexibility of
the stub to header attachments and/or header support system may result in
substantial loads on the stub to header welds. Tube sagging can occur on
vertical gas-flow HRSGs.
A potentially more damaging effect occurs as the condensate is cleared and
pushed forward into the outlet header(s) (and possibly the outlet manifold and
downstream pipework), causing a rapid decrease in inner surface temperature
known as quenching. This is particularly damaging on hot starts when the
header(s) will be at or close to normal operating temperature. As the inner
surface of the header tries to contract, the internal surface is put into tension
and can yield if the resulting stress is above the yield strength of the material.
3.3.3.2 The Superheater and Reheater on Cold Starts
The location of the final stage superheater as the first heat exchanger in the gas
path means that it inevitably sees the most severe temperature cycling. On
start-up, the tubes are exposed immediately to the highest gas temperature as it
enters the HRSG. A reheater, if installed, is also subjected to high ramp rates.
On utility-scale units, the GT exhaust temperature can increase from 80C to
450C in 5 minutes as it accelerates to synchronous speed [11]. This results in
high stress concentrations at the crotch corner of stub positions and uses up the
fatigue life of the component.

(37)

During a cold start, the gas flow causes the external surface temperature of the
superheater tubes to increase rapidly and the situation is exacerbated by the
fact that there is little cooling steam flow through the tubes until the drum
pressure has built up sufficiently. The tubes may reach temperatures well
above those experienced during normal operation and so may be at risk of
creep (as well as fatigue) damage during this period. The ramp up in
temperature is followed by a temperature drop when the drum pressure has
risen enough to initiate a substantial cooling steam flow. It is essential that
start-up vents (and drains) are sized and operated correctly to limit the ramp
rates in superheater components. These promote a cooling steam flow in the
superheater during the period leading up to admission of steam to the steam
turbine.
On horizontal gas-flow HRSGs, superheater headers with multiple tube rows
can be susceptible to the effects of differential tube expansion as the leading
row(s) warm up faster than those towards the rear of the bank. For this reason,
it is advantageous to specify smaller diameter headers with reduced numbers
of stubs/rows for plant that is expected to cycle regularly. Alternatively, the
leading row of tubes can be made unfinned so that the amount of heat pick-up
is similar to those on the finned row(s) behind. Vertical gas-flow HRSGs with
a serpentine tube arrangement between the inlet and outlet header can
accommodate thermal expansion of this type more easily and are generally
less of a concern.
Reheaters are also susceptible to thermal fatigue damage on cold starts. As no
cold reheat steam is available until the steam turbine has been rolled, reheater
tubes are often cooled with bypassed steam from the HP superheater during
the start, with flow being promoted by a reheater start-up vent. However, they
are still at risk due to the delay involved in establishing a superheater steam
flow as described above. Reheaters can also be susceptible to quenching with
condensate originating from the HP superheater when the turbine is being
bypassed.
3.3.3.3 Case Study: Excessive Ramp Rates and Through-Wall Differentials
on Superheater and Reheater Headers
Thermocouples were attached to the HP superheater outlet header and header
stubs of a dual pressure, vertical gas-flow HRSG to quantify the temperature
ramp rates on start-up. The temperature difference between the header external
surface and the stub/tube is a good approximation to that of the header
through-wall temperature, as the stubs are relatively thin. A severe thermal
upshock of around 200C in 13 minutes (15.4C/min) within an overall
change of 250C in 40 minutes was measured on a warm start, resulting in
substantial temperature differences between the header shell and the tube stubs
(Figure 15). As a rule of thumb, a tube ramp rate of 10C/minute is normally
considered to be at the upper limit of acceptability. A cold start showed an
overall change of about 380C in 45 minutes with no particularly high ramp
rates within that time. However, very high temperature differentials of up to
160C between the header shell and stubs were observed (Figure 16 [27]).

(38)

Figure 15: Temperature of the HP Superheater outlet header shell and tube
stubs on a warm start (Courtesy of Power Technology).

Figure 16: Temperature differential between the HP Superheater outlet header


shell and tube stubs during a cold start (Courtesy of Power Technology).
Following these preliminary observations, the effect of these transients and
through-wall differentials on the superheater outlet header was studied in more
detail. That study concluded that crack initiation could occur at the internal
surface of the stub hole ligaments after as little as 550 starts [28] and therefore
represented a serious threat to the long-term integrity of the headers.
Appropriate changes to the operational procedures (including gradual
admission of the exhaust gas flow to the HRSG using the bypass damper) and
the NDT strategy employed were recommended to the station concerned.
It should be noted that temperature differentials of comparable magnitude
were observed on the HP superheater outlet header of a similar HRSG, but
fatigue life calculations based on the start data observed did not provide cause
for concern [29]. This demonstrates that the ability of a header to withstand
thermal transients is dependent upon design (i.e. geometry, material and
(39)

construction detail) as well as the actual cyclic conditions that it is subjected


to.
Reheater outlet headers are also at risk of thermal fatigue on start-up,
especially when starting from ambient conditions on cold starts. Through-wall
differentials of 140C have been sustained for up to half an hour on the
reheater outlet header of a horizontal gas-flow HRSG (Figure 17 [30]).

Figure 17: Reheater header through-wall temperature differentials during a


cold start (Courtesy of Power Technology).
The fact that final superheater and reheater outlet header through-wall
differentials can be of a similar magnitude on shutdown as on a hot start was
demonstrated on the same plant (Figure 18). Differentials of approximately
60C were observed on shutdown and hot starts four to six hours later.

(40)

Figure 18: Comparison of start-up and shut-down temperature differentials


(Courtesy of Power Technology).
3.3.3.4 Reducing the Risks of Reheater and Superheater Fatigue
Many of the header design features described in Section 3.2 make high
temperature pressure parts less susceptible to thermal fatigue. These include
the use of P91 resulting in much thinner walls than when using P22 for a
comparable creep life and hence lower through-wall differentials. Full
penetration welds, flexible stub to header connections and the use of multiple,
smaller headers, each with a reduced number of stubs are also beneficial on
cycling plant.
For existing plant, the HP pressure should be kept as high as possible by
boxing up the HRSG for all hot/warm starts. This is beneficial for the HP
drum as well as the HP superheater outlet. Purging should be kept to a
minimum and any condensate formed should be removed prior to start-up,
with modifications made to drains if necessary. If possible, the GT load should
be kept low until a superheater steam flow has been established. Where
installed, bypass dampers may be able to be used to control the admission of
exhaust gas to the HRSG.
The beneficial effects of boxing up on shutdown, elimination of the shutdown
purge and condensate removal before start-up on a horizontal gas-flow, triple
pressure HRSG were estimated in [20]. Superheater outlet header cyclic life
was estimated to increase by a factor of six for cold starts, over five for warm
starts and over four for cold/warm starts following a trip.
3.3.4

Attemperators

Attemperator sprays are used to control the temperature of the main steam
leaving the final superheater/reheater outlet header within target limits.
However, poor attemperator control can result in quenching of

(41)

superheater/reheater headers, tube distortion (Figure 19) and damage to


downstream pipework. Innogy have experienced the latter where single-port
type attemperators have been installed [12]. Blocked or passing valves when
off-load can also result in off-design pipe movements.

Figure 19: Tube distortion due to quenching caused by poor attemperator


control (Courtesy of Power Technology).
Attemperators can be particularly problematic on start-up or when an HRSG is
being operated at part-load. The amount of attemperator spray required to
maintain the superheater temperature within target limits can exceed that
needed to cool the steam to saturation temperature. This can result in an
oscillating superheated steam temperature and result in thermal fatigue
damage to the superheater itself. This can be exacerbated by the fact that the
spray valve(s) may be too coarse for sufficient control at part-load operation.
Water droplets in the attemperated flow can also collect in tubes with the least
flow resistance, resulting in temperature differentials between superheater
tubes [11].
The design of attemperators has improved significantly over recent years, and,
especially on larger plant, it is now not uncommon to see multi-nozzle, piston
controlled type sprays or similarly sophisticated systems.
3.3.5

Tube Banks

Problems with tube fretting have been experienced on a number of CCGT


and CHP plants, where abrasion of tubes against the tube sheets through which
they pass has caused wall thinning and, in some cases, tube failure. Tube
(42)

sheets may be used for tube support, dead-space baffling, tube alignment and
to reduce vibration. Fretting can occur as tubes get snagged on these tube
sheets during contraction on shutdown, particularly where finned tubing is
used and/or tube/tube plate distortion has occurred (Figure 20). This may also
result in increased stress on the stub to header welds.

Figure 20: Fretting of a tube where it passes through a tube sheet (Courtesy of
Power Technology).
Tube failures have also resulted from gas bypassing up the sides of the HRSG
casing. Bypassing leads to a reduction in HRSG thermal performance, casing
distortion and excessive heating of the wing tubes in the tube bank being
bypassed and those downstream. The latter problem can also have an adverse
effect on the water/steam-side chemistry of evaporator tubes, increasing the
risk of flow accelerated corrosion and/or phosphate hide-out (Sections 3.5.3
and 3.5.5). Gas bypassing also puts the wing tubes at increased risk of
buckling/fretting as they are operating at substantially hotter temperatures than
the rest of the bank.
The likelihood of gas bypassing is reduced by the use of a cold-casing design
(as casing distortion is much reduced) and by the use of intermediate tube
sheets tied to the casing side-walls at intervals across the gas path. In addition,
gas baffling is generally installed by the OEM (original equipment
manufacturer) at the sides of each tube bank and at the header dead-spaces, but
in several cases this has been either poorly designed or executed. However it is
often possible to make improvement to these baffles later in the plant life, if
required.

(43)

3.3.6

Pipework and Pipework Support Systems

There can be interface problems between the pipework and the HRSG itself as
pipework systems are often neglected during design and construction. Often
responsibility is split for different sections of pipework between the steam
turbine supplier, boiler supplier and turnkey contractor, and as such
responsibility for the overall system is not clear.
Experience is that a high level of QA is necessary during the construction
process and any subsequent pipe support issues arising during commissioning
or early operation need to be addressed quickly. This is particularly true for
pipework operating in the creep range, where there may be a significant effect
on the future life of the piping. It is not uncommon to see significant pipework
support system defects relating to poor design and/or erection very early in life
eg supports stuck at the top/bottom of their travel, still gagged in the erection
position or simply not moving through their design hot to cold movement. In
these cases, it often falls to the plant user to highlight the defects and ensure
rectification work is carried out.
Although pipework systems generally see less damaging transients than the
final superheater outlet header and the geometry is comparatively
straightforward, annual surveys of hot and cold pipework movements and the
associated stresses are typically carried out, for comparison with design
values. It is particularly important to ensure that pipework movements are
acceptable if operating under a flexible operation regime.
Where more than one HRSG is interconnected, there is an increased likelihood
of temperature differentials in the common pipework, especially if the units
are two-shifting. The sealing efficiency of pipework isolation valves is also
important [22].
3.3.6.1 Case Study: Problems with Pipework and Pipework Support
Inspection of the major steam pipework support systems on a 4 x 350MW
CCGT shortly after commissioning showed that a significant number of
supports were incorrectly adjusted and that support name plate details were
incorrect [31]. It was confirmed that this was due to supports having been
swapped around during construction following late changes to the pipework
design philosophy. Analysis of the as built pipe support system identified
significant overloading at the boiler and turbine terminal points.
As a result, the HP steam, hot and cold reheat pipework were redesigned and
re-hung on all units. After 12 months operation with the new support system, a
check of cold to hot movement/support condition on all four units was carried
out. Whilst there had been a marked improvement in pipework behaviour
following the re-support, cold to hot movements appeared to have decreased
over the last year and were in some cases substantially less than design (Figure
21). It is now proposed to rerun the analysis against these lower movements
to check system stresses. This incident has confirmed the benefit of carrying

(44)

out an independent review of design and installation of major pipe systems


during the warranty period.

Figure 21: Pipe support movements - predicted and measured (Courtesy of


Power Technology).
3.3.7

Valves & Fittings

Valves and miscellaneous fittings (gauge glasses etc) supplied on new plant
are often of low quality, as this tends to be an area where OEMs can save
money. This can results in poor reliability during commissioning and the early
months/years of operation, generally in the form of poor isolation or leakage
to atmosphere. It is not unknown for wholesale replacement of all minor
valves (drain valves and so on), or for the fitting of Hydrastep (or similar) to
replace gauge glasses, to be necessary very early in station life.
The arrangement of valves is also often deficient - Innogy cited the lack of
master/martyr drain valve arrangements as a particular problem. Valve
problems are not always limited to the smaller valves, with Innogy also giving
an example where back heating from a passing main reheat isolation valve
caused pressurisation of a sister unit [16] in this case high quality isolation is
essential.
Innogys preference is that for many applications parallel slide valves should
be used for isolation duty, with globe valves being used for throttling duty [22].
Valves sometimes pass due to erosion or partial blockage with debris. In one
particular case, debris has clearly been responsible for valve seizures on a
HRSG containing large amounts of mill scale deposit arising from the fact that
no acid clean was carried out.

(45)

3.4

Operational Issues with HRSG Non-Pressure Parts

3.4.1

Ducting & Casing

Ducting is used to guide the exhaust gases from the gas turbine to the HRSG
where the tube banks are enclosed in an insulated casing to ensure maximum
heat recovery. There are essentially three ducting/casing design philosophies
used on HRSGs and these are described as hot, warm and cold. An
HRSG may also have different types of casing design in different areas
depending on the local gas temperature and preference of the manufacturer.
In all cases, a typical maintenance regime would comprise thorough visual
inspections at least annually to ensure that any minor defects are identified and
rectified. This is particularly important on cold and warm casing designs
i.e. those with insulation inside the casing, where experience has shown that
problems can develop beneath insulation liner plates which result in sudden,
catastrophic failure of the internal insulation system if not addressed. Thermal
imaging surveys are also a useful inspection tool as these enable operators to
identify casing hot spots where insulation may be dislodged or internal
insulation may be damaged. Prompt repair may prevent irreparable damage to
the casing welds.
3.4.1.1 Hot Casing Design
In a hot casing design, insulation is fitted to the exterior of the HRSG
ducting/casing and the interior is exposed to the full temperature (and
temperature fluctuations) of the gas flow. The key advantage of this design is
that it is simple to construct and the casing interior can be inspected and
repaired easily. However, there are a number of disadvantages:

If a hot casing design is used in the presence of high GT exhaust


temperatures (~ 570C or above), high specification (i.e. austenitic
stainless steel) materials must be used, which are relatively expensive and
have poor thermal properties compared to ferritic alloys.
Expansion joint steelwork is exposed to the full exhaust gas temperature.
This means that the through-wall differential across the joint flange
steelwork is increased and the joint fabric has to accommodate increased
thermal expansion of the duct as compared to that on a warm or cold
casing design. This can lead to distortion of expansion joint flanges and
cracking of joint components, particularly at the corners on joints with
square/rectangular cross-sections.
The casing can be vulnerable to damage during start-up and shutdown due
to the large temperature differentials set up across the external casing
stiffeners.
On larger units, the use of a hot casing design normally requires the use
internal bracing in addition to the external stiffeners fitted as standard. The
designs of these systems can be very complicated and may result in
components fighting against each other during transient conditions when
the internal and external surfaces of the ducting are at different
temperatures. Widespread, recurring defects can result.
(46)

3.4.1.2 Case Study: Hot Casing Design


A HSRG ducting design featuring a large number of welded internal and
external stiffeners is shown in Figure 22 [32]. During transient conditions (e.g.
on start-up or at shutdown), the gas inlet temperature can peak at up to 650C,
and temperature differentials occur between the ducting, internal stiffeners,
external stiffeners and pressure parts. Even on an extremely gentle start,
differentials of up to 350C have been measured between the casing and the
extremity of an external stiffener, causing extremely high stresses. Visual
surveys of the duct have highlighted a large number of resultant defects
including: failed internal bracing pipes
cracked floor beam to tube plate welds
cracked internal stiffener attachment welds
torn ducting at corners local to external stiffeners
splits in the duct radius-corner welds

Figure 22: Internal bracing and external stiffeners on a hot casing design
(Courtesy of Powergen UK plc).
The ducting is made of grade 316 S51 stainless steel and the relatively high
thermal expansion coefficient and low heat transfer coefficient of this material
increases the temperature differentials between internal and external stiffeners
during start up. Whilst the units operate at base-load, the defects are
manageable by routine inspection and repair, though it is a major cause of
additional expenditure during boiler outages. However, if a more flexible

(47)

operating regime was introduced, the ducting could become a significant risk
to boiler integrity. In anticipation of this, a pin-connection design has been
successfully retrofitted to some internal stiffener locations to increase duct
flexibility and reduce the extent of the repeat defects.
3.4.1.3 Cold Casing Design
In a cold casing design, the insulation system is fitted to the casing interior
using a dedicated support system. Internal insulation is generally used as
standard on horizontal gas-flow boilers [10, 14] and a cold casing design is the
type adopted by Nooter/Eriksen [13]. Vertical gas flow boilers may also use a
cold casing design, depending on the manufacturers preference. There are
many different designs of internal insulation in use and these have a number of
advantages:

The lower casing temperatures resulting from the use of internal duct
insulation enable a lower specification of casing material to be used.
As the thermal expansion of internally insulated casings is considerably
lower than that of externally insulated designs, less movement has to be
accommodated by duct expansion joints.
External duct stiffeners can be welded directly to the casing and there is no
need (or at least a reduced need) for sliding connections.
For higher temperature applications (e.g. where supplementary/auxiliary
duct firing is installed typically CHP), ceramic block systems can be
used as duct liners as long as gas velocities are low (less than
approximately 36m/s). This provides a useful bridge between stainless
steel liners, which a limited to around 850C and the more costly watercooled furnace option.

The disadvantages of using an internal insulation system are:

Internal insulation systems can be relatively expensive and time


consuming to install.
The design of the insulation system and in particular the quality of the
installation must be excellent to prevent failure, which may be sudden and
dramatic. For example, Siemens place great emphasis on quality during
erection and specify a number of design details including pin size, liner
plate size and the use of pre compressed insulation that does not compress
when wet [10].
Attention to detail is crucial. For example, welds between insulation
system top nuts and washers are essential, otherwise washers are free to
vibrate and eventually wear through. Once washers have been worn away,
liner plates are then free to vibrate which allows insulation fibres to escape
into the gas flow. Casing hot spots and eventual distortion/detachment of
the liner plates themselves can result.
Liner systems can be unsuitable for very high temperature applications
(above 850C) which is typical of ducts with supplementary firing. Very
high temperatures (typically over 1100C) require water-cooled membrane
walls.

(48)

There have also been incidences of stress corrosion cracking of the steel pins
welded to carbon steel duct [11].
3.4.1.4 Case Study: Cold Casing Design
At a CHP site, the GT exhausts into a right-angled duct before entering the
HRSG. This increases the exhaust gas velocity in some areas and reduces it in
others. The original liner system failed because the plates and bolts were too
thin (at 10mm diameter and 2mm thick respectively) and the insulation
material was incorrectly selected (mineral wool). A ceramic block system also
proved unsuitable, as the conditions in the duct were too turbulent. The
replacement design [33], featured small liner plates each anchored to a central
location. The thickness of the bolts and liner plates was increased (to 12mm
diameter and 3mm thick respectively) and the insulation material (now
Superwool 607) was compressed beneath the liner plates by up to 25%. This
design has operated satisfactorily for three years [34].
3.4.1.5 Warm Casing Design
As the name suggests, warm casing designs consist of thinner layers of
internal and external insulation, which, combined, result in a similar level of
heat retention. Such systems tend to have some of the advantages and
disadvantages of both hot and cold casing designs, but may be beneficial in
certain circumstances. For example, the NEM standard design for vertical
HRSGs is a warm casing design up to the HP Evaporator level with external
insulation only fitted above that [14].
3.4.1.6 Case Study: Warm Casing Design
An internally and externally insulated GT diffuser duct between the GT
exhaust (V94.3a gas turbine) and HRSG inlet has a circular inlet, rectangular
outlet and is made of P265HG (carbon steel) with hot external stiffeners
throughout and additional cold stiffeners on the rectangular section only. The
liner plates of the internal insulation system overlap in the direction of gas flow,
contain 150mm ceramic fibre and are exposed to 631kg/s of GT exhaust flow at
595C and 33mbar. The duct has 150mm thickness of man-made mineral fibre
for external insulation.
After about 18 months operation, a major failure of the internal insulation
system was discovered with approximately 80% (~200m2) of the internal
insulation and stainless steel liner sheets detached from the duct walls, leaving
the carbon steel ducting exposed to full gas temperature (Figure 23). Only
very minor defects had been noted previously. The cause of failure was
determined to be constrained and uncontrolled thermal expansion of the liner
plates [35], resulting in thermal fatigue and insufficient radial clearance for
thermal expansion of the externally insulated (hot) GT exhaust duct within the
internally insulated (warm) GT diffuser duct. The design has since been
modified to be more flexible, with a strengthened liner sheet attachment
system and increased clearance between the duct and GT exhaust.

(49)

Figure 23: Destruction of the internal insulation system in a GT diffuser duct


(Courtesy of Power Technology).
3.4.2

Expansion Joints

Gas duct fabric expansion joints are typically located at the HRSG inlet,
HRSG outlet to the main stack and, if applicable, at the bypass stack inlet. The
GT to diffuser duct connection is more usually a metallic bellows type, due to
the higher gas conditions, although specially designed fabric types have been
used in this location at times.
Gas temperature, velocity and turbulence local to the main stack is usually
relatively low, so there are rarely problems associated with fabric expansion
joints in these areas. However, high temperature gas duct fabric expansion
joints are often of relatively complex design and experience onerous operating
conditions with respect to gas temperature and thermal expansion, particularly
on the larger GTs. In particular this can result in very large temperature
differences being established between the inner and outer flanges of a joint,
particularly during a unit start, leading to steelwork deterioration. It is
typically this mechanism which ultimately leads to fabric damage.
Expansion joints can be round or square depending on the shape of the duct
and consist of a gas tight fabric membrane with internal insulation, which
attaches to steelwork flanges either side of the gap that it has to bridge.
Although defects have been experienced on round joints, more severe defects
have been experienced on square joints square joints inherently tend to
concentrate thermal expansion loads and stresses at certain points within the
frame, whereas in a round joint, these tend to be evenly spread. Typical
(50)

defects include flange distortion and cracking of the joint steel frame,
especially at the corners. In a limited number of cases it has been possible to
reduce the incidence of corner cracking by retrofitting radiused corners into
the original square joint design.
If gas leaks as a result of a defect in the joint steelwork, the life of the joint
fabric itself is greatly reduced due to operation at above-design temperatures.
The outer fabric is typically of a PTFE type material, with an operational limit
of around 250C. Furthermore, guarantees for joint fabric and insulation are
often rendered invalid if failures are caused by defects associated with the
joint steelwork. Operational experience therefore shows that integrity of the
steelwork around the joint is the key to ensuring reliable expansion joint
operation. Expansion joints often represent the most pressing obstacle to more
flexible operation, as low cycle fatigue exacerbates existing defects and can
lead to the creation of new ones. Many in-service HRSGs in the UK feature
expansion joint designs from the mid-1990s that were not designed to
withstand large numbers of unit starts. However, this is now widely
recognised by the joint manufacturers and more advanced design solutions are
now available.
Siemens KWU has experienced problems with expansion joints on more than
one of its plants and in their opinion, many of these stem from the erection
process. In recent years, Siemens have ensured that the joint fabric and pillows
are installed as late as possible to minimise accidental damage and this has
resulted in a marked decrease in incidents of gas leakage. Siemens also
stipulate that the steel flanges of any joint are protected from the residues of
GT on-line water washing by the installation of a dam and small (typically 3
bore) drain at the bottom of the gas duct [10].
A typical inspection and maintenance programme would include annual visual
and thermal image surveys to identify expansion joint defects in their early
stages. Prompt action in addressing joint defects may considerably extend the
life of the fabric if not the joint itself. Anticipated life of the fabric is typically
3-6 years, with the replacement interval often being planned to coincide with a
major gas turbine outage.
3.4.2.1 Case Study: Expansion Joints
Expansion joints situated downstream of a GE 9FA gas turbine are exposed to
modest gas pressure inside the duct (around 42mb), but the gas temperature is
up to 650C peak (during start up) and the flow is highly turbulent. Thermal
surveys revealed a number of hot spots with surface fabric temperatures in
some cases exceeding 400C. This is 250C above the maximum continuous
operating temperature for the joint material and at least 300C above that
typical for a joint of this type. By comparison, the predicted joint fabric
temperature from the original design calculations was only 60C.
Cracks in the ducting had been repeatedly ground out and re-welded during
outages, only to re-appear. This was particularly the case at the duct corners.
This led to significant gas leakage, which was both a safety risk and a major

(51)

contributor to the joint overheating noted above. There was also significant
flange distortion and consequent joint distortion and this created problems in
replacing the joints because the dimensions of any replacement must match
the flanges (Figure 24).

Figure 24: Expansion joint flange distortion (Courtesy of Power Technology).


A replacement joint was installed using the existing flange steelwork, which
had to undergo considerable repair prior to installation. However, the
steelwork deteriorated again quite quickly, with fabric joint temperatures
rapidly increasing to above design levels. For example, thermal survey one
year after new joint installation showed joint temperatures up to 327C, flange
temperatures up to 430C and obvious gas leaks local to flanges, with peak
casing temperatures of up to 566C (Figure 25). Differential temperatures
between hot and cold parts of the steelwork of up to 400C were recorded
during start-up.

(52)

Figure 25: Thermal image of expansion joint (Courtesy of Powergen UK plc).


Ultimately, the enhanced cost of steelwork repairs and joint replacement, the
increased risk of unexpected joint failure and the potential release of exhaust
gases into the main plant building at ground level were sufficient justification
for a change in design of the joint steelwork. It is also worth noting that due to
the mechanism of failure (i.e. thermal fatigue), the extent and severity of
damage is predominantly related to units starts and will be exacerbated by
flexible operation. This has since been installed on one of four units and been
problem-free to date.
3.4.3

Bypass Duct and Damper

Bypass ducts and dampers enable gas turbines to continue running in open
cycle (rather than combined cycle) mode whilst the HRSG is shutdown and
also allow auxiliary duct burners to be fired when the GT is off-load. If a
HRSG defect requiring off-load maintenance occurs on a plant without bypass
capability, it means that the GT also has to be shut down before any
maintenance can take place and vice versa. This can result in heavy financial
penalties for some CHP operators and for this reason they are often installed
on CHP plant. They are sometimes installed on utility-scale CCGT plant, but
since pure utility CCGTs are generally uneconomical to operate in open cycle
mode and their HRSGs are generally unfired, there is less of a convincing case
for the additional expenditure on a bypass. It should also be noted that in the
US, a GT cannot usually meet the stringent regulatory emission limits for NOx
in open cycle mode, as the Selective Catalytic Reduction (SCR) device which
reduces NOx emissions is situated within the HRSG itself. Bypass stacks are
therefore less useful under these circumstances [36].

(53)

Bypass dampers are part of Nooter/Eriksens scope about 80% of the time in
the UK. They prefer hydraulic rather than motor-driven dampers as these give
a greater degree of control, albeit at increased cost. They also prefer flap
dampers to the multi-louvre type, unless the client specifically requests the
latter. The advantages of multi-louvre dampers lie in better control of flow at
partial opening, so these tend to be found more often on CHP plants where
they are sometime used to help control GT exhaust flow to the boiler and
thereby control process steam flow.
Flap type bypass dampers are usually reliable in operation, however problems
can be experienced with sealing efficiency, usually due to distorted damper
seals. On one Powergen CHP plant, incomplete sealing of the diverter damper
when in the combined cycle position resulted in 13-20% of the GT exhaust gas
being lost to the bypass stack. On another Powergen plant, 2-3% loss of
exhaust gas to the bypass stack during normal operation was estimated to
result in a 0.5MW loss on the steam turbine [37]. For these reasons, damper
blades should be well insulated to minimise distortion and the seals regularly
inspected and maintained.
It is generally also possible to use diverter dampers to control HRSG warming
rates on start-up, thus reducing thermal transients to the HRSG components
and the HP superheater in particular. However opinion is still divided as to
whether the benefits are substantial enough to justify the additional cost of a
diverter damper on a large utility scale HRSG.
3.4.4

Burners

The supplementary and auxiliary firing modes of duct burners are described in
Section 2.5. Both Innogy and Powergen have experienced operational
problems stemming from duct burner operation (predominantly superheater
tube distortion and damage to internal insulation).
It is vital when operating in supplementary firing mode that the duct burners
are not brought into service until a cooling steam flow has been established in
the superheater tubes. Failure to do so can result in operation at above-design
temperatures and potentially in tube failures.
It should be ensured at the design stage that the first bank of tubes is not too
close to the burners, otherwise flames may impinge directly onto the tubes and
it will be more difficult to evenly distribute the heat across the gas path. This
can result in tube overheating and distortion. On one Powergen plant,
excessive heat-pick up combined with poor boiler water quality led to tube
failures on the first row of evaporator tubes. The solution involved reducing
the heat pick-up of the leading row of evaporator tubes by replacing them with
unfinned tubes.
It should also be ensured that the duct insulation is suitable for the
temperatures resulting from the combination of GT exhaust gas and the
simultaneous operation of the duct burners. In this respect, experience has
been that even when mixed gas temperature is nominally acceptable to the

(54)

insulation system, flame impingement due to poor burner set up or an


inadequately sized firing duct has resulted in much higher temperatures at the
wall.
3.4.5

Silencers

Silencers are frequently used on HRSGs and may be situated at the HRSG
inlet, at the base of the HRSG stack and/or at the base of the bypass stack.
Their purpose is to attenuate the noise of the GT exhaust gas and, in the case
of silencers at the HRSG inlet, to improve the distribution of the exhaust flow
across the tube banks. Designs typically consist of a number of vertical
silencer pods arranged in the duct with the gas flow passing through gaps
between them or a matrix of silencer pods with gas passing through horizontal
and vertical inter-spaces. The pods themselves are packed with insulation
material appropriate to the operating temperature.
High quality installation is important in ensuring that there is no uncontrolled
movement of silencer pods, particularly at the HRSG inlet where turbulent
conditions, high temperatures and severe temperature ramp rates may be
experienced. At one site, for example, detachment of silencer pod guides due
to incorrect installation has resulted in damage to the guides and pods on one
HRSG, whereas the equivalent components on sister units have had no
problems.
Significant slumping of pod insulation and loss of insulation fibres to the gas
stream has also been experienced on a number of units. This degrades the
sound attenuation capability of the silencers and the fibres tend to be deposited
in finned tube banks. This can present a health risk to personnel working in the
boiler if refractory ceramic fibre (RCF) has been used as the insulating
material.
RCF has been classified as a Category 2 carcinogen under European Directive
97/69/EC and, where practicable, should be replaced with material which falls
outside the scope of the European Directive. When insulation has been lost to
the gas stream, it can be replaced with insulation, which is injected into the
pods in liquid form and then hardens. This is a relatively straightforward
process. Whilst this method has been used successfully on HRSG silencers [38],
it may not always be appropriate for replacing insulation lost from duct
insulation systems, as it can result in the jacking off of internal liner plates
[39]
.
3.4.6

Case Study: Silencers

A 42MW Frame 6 gas turbine exhausts gas into a duct which turns through
90 and passes to a silencer via an asymmetrically enlarging duct before
reaching the bypass stack. The original silencer construction featured six
modules, each with two pods measuring 5m high x 2.5m long situated back to
back. Each pod had a flow plate of semicircular cross-section, which also
acted as a stiffener, side panels made of perforated plate and guides on the
ductwork roof and floor.

(55)

After approximately five years operation, damage to the front and rear
semicircular flow plates was noted, with some being completely detached
(Figure 26 [40]). As a consequence, silencer packing material had been
dispersed throughout the boiler and to atmosphere. The damage suggested that
the silencer design was inadequate to cope with the flow induced vibration and
thermal expansion experienced. Operating temperatures are typically 530C,
with gas velocities up to 50m/s and high turbulence due to the duct geometry.

Figure 26: Silencer damage downstream of the GT exhaust in a CHP plant


(Courtesy of Power Technology).
The replacement design adopted was of the matrix type, which has inherently
lower gas velocities between the modules and was therefore considered to be
less prone to flow induced vibration damage. This design has been problem
free to date.
3.4.7

Stack Damper

Following GT shutdown, a natural draft draws cool air through the turbine and
HRSG which cools the HRSG components. However, this natural draft and
the associated cooling effect can be minimised by the inclusion of a stack
damper in the design specification. This is usually of the louvre or butterfly
type and is positioned close to the base of the main stack. The stack damper is
usually closed when the GT has run down to barring speed, putting the HRSG
into hot storage. This greatly reduces the pressure decay and cooling rates of
(56)

the HRSG components, so that they are much closer to their normal
operational temperatures on start-up. The likelihood of fatigue damage to the
high temperature components is therefore reduced and, as it is these that tend
to limit the speed with which a unit can return to service, the unit can be run
up more quickly.
For these reasons, stack dampers are widely regarded as a positive design
feature [11, 16]. Nooter/Eriksen stated that stack dampers are currently used
more in the US than in the UK but this is becoming more of a necessity in the
UK due to the increased need for cycling plant. NEM and Standard Fasel
typically include stack dampers in their designs [14, 15] . Unfortunately there are
difficulties involved in retrofitting a stack damper to an existing, vertical gasflow HRSG due to the imposition of significant loads onto an existing
structure. This is not necessarily the case for a horizontal gas-flow HRSG.
A stack damper also reduces the risk of tubing/ducting corrosion from rain.
This is particularly true on coastal sites where a damper can reduce or
eliminate the amount of salt deposits reaching the preheater when off-load [19].
Despite reducing the risk of thermal fatigue damage to the high temperature
components of an HRSG, the use of a stack damper can have some damaging
effects to components such as economisers at the cold end of the boiler. These
are discussed in more detail in Section 3.3.1.2.
3.5
3.5.1

Cycle Chemistry Issues in HRSG Plant


Introduction

The basic principles of chemical control applied to the water and steam side of
HRSGs is no different to that applied to conventional boilers. These are to
maintain chemical conditions in the feed and boiler water which minimise
both corrosion of the circuit materials and also the transport of impurities from
the feed system to the evaporator and steam circuits. The main principles are
to maintain a mildly alkaline feed water and evaporator water to limit
corrosion of both ferrous and copper based materials.
As with conventional fossil fuel fired plant the chemistry applied to HRSGs
has to be compatible with materials in other parts of the circuit. As the primary
purpose of the HRSG is to produce steam, this steam must be of suitable
purity for the purpose for which it is to be used. This may be powering a
condensing steam turbine or, alternatively, some or all of the steam may be
passed out to a steam customer, either directly or via a back pressure steam
turbine. The particular use of the steam by the customer may place specific
demands on steam purity, which in turn may restrict the applicable feed and
evaporator water chemistry. Where steam is exported for third party use the
loss of steam must be made up with fresh make-up water. This make-up may
be entirely fresh water or may be comprised, in part, of condensate returned
from the process where the steam was used. Such condensate may be
contaminated relative to the required make-up water purity.
(57)

HRSGs may be single pressure or multiple pressure designs. In general, but


not exclusively, single pressure designs are used for steam export CHP
systems, while multiple pressure designs are more commonly used with
conventional condensing turbines. Because HRSGs are constructed with thin
walled, finned tubing, to maximise heat transfer, the margins for corrosion are
much less than in conventional plant. Therefore, there is often a requirement to
use a more pure feedwater or a more demanding evaporator chemistry regime
than may have been used with a conventional boiler of similar pressure rating.
This is reflected in the latest guidance contained in British standard
BS2486:1997 [41], which now includes a specific section on non fired heat
recovery boilers. Other information sources commonly used for HRSG cycle
chemistry specification are the VGB guidance [42] and documentation
produced by EPRI [43]. In addition many individual boiler manufacturers and
major turnkey contractors have their own internal guidance for feed water and
evaporator water chemistry. However, some of this guidance extrapolates
from one pressure range to another without paying due regard to all of the
plant and materials interactions which can vary with pressure and temperature.
As noted above, the principles of chemical control are to maintain mildly
alkaline conditions in both the feed water and evaporator water. A second
requirement is to limit the amount of oxygen entering with the feed water. The
combination of oxygen and impurity salts within the feed water or evaporator
will cause serious corrosion and component failure. Removal of oxygen
dissolved in the feedwater may be achieved by mechanical deaeration using
steam scrubbing, by vacuum deaeration of the make-up water, by dosing
chemicals which react with the oxygen or by a combination of these processes.
In low pressure HRSGs there may be only chemical oxygen scavenging. As
the pressure increases a combination of mechanical deaeration and chemical
scavenging is more common. Non volatile chemical scavengers (such as
sodium sulphite) cannot be used where the dosed feed water will be used for
spraying into superheaters to de-superheat the steam, as the non-volatile salts
will deposit in the superheater. Volatile oxygen scavengers, such as hydrazine
or carbohydrazide, are generally used in higher pressure plant, and in all plant,
irrespective of pressure, where the feed water is used for de-superheating.
The alkalinity of the feed water, and the choice of alkali used, is dictated by
the materials of the feed system and those of the steam system and condenser
or condensate system. In most HRSGs the feed water alkalinity is adjusted by
dosing a volatile alkali. This may be ammonia or an organic amine. Ammonia
is used for its simplicity and low cost, but it is highly steam volatile. Organic
amines are less steam volatile and are often used as mixtures with differing
volatilities and basicities. However, they have a tendency to degrade at high
temperatures and pressures to release low molecular weight organic acids into
the steam and water circuit. Organic amines are preferred where there are long
condensate return lines or there are particular system or design requirements.
Where the steam is to be used for food processing or food packaging the
volatile components in the steam are both limited and strictly regulated [44].

(58)

The ideal pH range for feedwater to an all ferrous system is 9.2 - 9.6
(measured at 25oC), but if copper or copper alloy components are present this
must be reduced to pH 8.8 9.2, where ammonia is present, to avoid copper
alloy corrosion. For evaporator water, the ideal pH range varies between pH
9.0 10.0, with a higher pH being more prevalent in lower pressure systems,
but each system must be considered for its particular design, circuit materials
and steam purity requirement.
3.5.2

Steam Droplet Carryover

As noted earlier, steam purity limitations can be influenced by the steam


volatile components added to the feedwater but it is also influenced by the
concentration of non-volatile salts in the evaporator water. These may be
added as boiler water conditioning agents or they may have accumulated from
impurities in the make-up water. Such salts will be contained in droplets of
water which are entrained in the steam leaving the steam drum. Droplet
separators and scrubbers are normally fitted in to the roof of the drum to scrub
the steam free of water droplets and impurities. The efficiency of these
separators is a key factor in determining the final steam purity. Even in drums
with properly designed droplet separators the droplet carryover will increase
as the drum pressure rises.
The early operation of HRSGs produced a number of specific problems that
were not expected. Two of these were flow accelerated corrosion, which
affected the integrity of the HRSG, and the hideout of sodium phosphate
dosed to the evaporator water for pH control. These two features are discussed
below.
3.5.3

Flow Accelerated Corrosion

Flow accelerated corrosion (FAC) occurs where the protective oxide at a tube
or component surface is removed, by a combination of flow turbulence and
unsuitable chemistry, faster than it can reform at the tube wall / liquid
interface, resulting in ongoing corrosion of the underlying metal. This type of
attack is characterised by a polished surface, very thin or sometimes no
surface oxide and by accretions of horseshoe shaped pits. It is dependent on
component material, component geometry, operating temperature and the
chemistry of the water phase in contact with the component. Further
information on FAC in power plant is presented by Chexal et al [45] and more
specific information relative to HRSGs was presented by Harries and Willett
[46]
.
HRSGs are mainly constructed from carbon steel, to limit costs, and the major
FAC risk is generally with bends in economiser or evaporator pipework.
Carbon steel components, and pipework, with a protective magnetite layer, are
more susceptible than those with a small percentage of molybdenum or
particularly chrome, which form a more protective oxide. Designs with sharp
bends of 90o or greater can create significant turbulence, although high flow
rates in plain pipework can also increase susceptibility to FAC. FAC can occur
in a single water phase, or in two phase steam / water mixtures where the
water phase is in contact with the tube wall. It cannot occur in the pure steam
(59)

phase. The temperature range over which FAC occurs is 120 200oC, with a
peak around 150 170oC. Therefore, the lower temperature economiser and
low pressure evaporator circuits are most at risk from FAC. The chemistry of
the water phase also has a significant effect on the rate of FAC. Above a pH of
9.5 there is little risk of FAC occurring, but the rate of corrosion increases as
the pH is reduced below this value. Similarly, the presence of a small, but
controlled, concentration of dissolved oxygen, up to 10 g/l (ppb) O2, in the
water phase will inhibit FAC by creating a protective oxide containing the
more compact and less soluble ferric oxide, haematite. It has also been
established that strongly reducing conditions, e.g. overdosing of an oxygen
scavenger, can also promote FAC.
3.5.4

Design Features Promoting FAC.

3.5.4.1 Horizontal Tubed HRSGs


The early examples of FAC in HRSGs were found in the 180o return bends of
horizontally tubed, low pressure evaporators (Figure 27). The inner return
bends were particularly at risk (Figure 28), although less pronounced FAC
damage was found in the middle and outer bends. In the example shown, the
chemistry was being operated with an all volatile regime in both HP and LP
evaporator, with a feed water pH in the range 9.2 9.4 and dosed with
hydrazine as the main alkali, relying on its degradation to ammonia as the
source of alkalinity. Ammonia has a significant distribution into the steam
phase of the water steam mixture in an evaporator. In a low pressure
evaporator, typically 5 8 bar pressure, the distribution coefficient for
ammonia between steam and water is 10:1 8:1 respectively. The pH of the
feed water was already below the key value of 9.5, but in the LP evaporator
the pH of the water phase was reduced further to a value in the range pH 8.5
8.8. Therefore, FAC damage was almost inevitable, and the worst carbon steel
bends were fully corroded within three years operation. The rate of FAC was
exacerbated by hot gas bypassing up the side of the HRSG between the casing
and the tube banks. This raised the steam fraction in the outermost tubes,
which in turn increased the velocity of the steam / water mixture, promoting
greater turbulence. Figure 29 shows the pattern of wall thinning for the final
inner return bends across the tube bank. Note the outer tube shows less
thinning as it was believed that this bend was entirely in the steam phase, due
to the hot gases raising the temperature.

(60)

Figure 27: Example of flow accelerated corrosion damage to tube (Courtesy


of Power Technology).

Figure 28: Diagram illustrating point at which worst FAC damage occurs
(Courtesy of Power Technology).

(61)

Figure 29: Decrease in wall thickness at bends associated with flow


accelerated corrosion (Courtesy of Power Technology).
The solution to FAC in the plant described above has been two fold. Replace
all thinned bends with new bends manufactured from 2.25%Cr, 0.5% Mo
steel, for extra resistance to FAC, plus changing the chemistry to non-volatile
alkali dosing to the LP evaporator and eliminating the oxygen scavenger
dosing to the feed system. In the particular instance described above, sodium
hydroxide was dosed to the LP evaporator water to maintain the pH in the
range 9.5 9.8. This has successfully suppressed further FAC attack.
3.5.4.2 Vertical Tubed HRSGs
Vertical tube HRSGs generally incorporate fewer 180o return bends within
their design. However, they can incorporate a number of bends up to 90o
angle, which are equally at risk from FAC, where there is a steam / water
mixture in the key risk temperature range and a low pH chemistry. Again this
generally refers to the use of all volatile chemistry in the LP evaporator.
Modelling of the potential for FAC corrosion has shown several designs to
have a high risk of FAC bend wall thinning to an unacceptable degree within
the expected 20 year lifetime of the plant. This can affect both bends in the
evaporator and bends in pipework linking the evaporator top header to the
drum. Again the recommended solution is to use a non-volatile alkali, sodium
hydroxide or tri-sodium phosphate as the boiler water pH conditioning agent
in the LP evaporator. However, for some designs this is not practicable.
3.5.4.3 Combined LP Drum / Deaerator or LP Drum / HP Feedwater Tank
This design of plant combines the LP evaporator drum with the feed water
tank for the IP and HP evaporators. It is not possible to dose the LP
evaporator circuit with a non-volatile alkali at a pH suitable to protect against
FAC, i.e. pH 9.5 9.8, as this high pH water would be fed to the IP and HP
evaporator circuits where it would concentrate further, by up to two orders of
magnitude. This would produce both pH values and salt concentrations that
(62)

would be too high for safe operation. Therefore, the LP evaporator can only be
dosed on an all volatile treatment (AVT) basis. If the HRSG components are
all mild steel and the feedwater dosing is ammonia, then a very high
concentration of ammonia is required to achieve the minimum pH of 9.5 in the
water phase, due to the high partition of ammonia into the steam phase at low
pressures. Typically a feed water pH of 9.8 10, which equates to 5 10 mg/l
NH3, is required. Operating with such a high pH should provide water of
suitable alkalinity as feed to the IP and HP circuits without additional nonvolatile dosing. Use of such high ammonia concentrations may have an impact
if the steam is exported for third party use and will require elimination the use
of any copper alloy components in the circuit.
Two examples of the problems associated with this type of design are given
below.
A small CCGT with dual pressure vertical tube HRSG, incorporating a
combined LP drum/deaerator/ HP feed tank and all carbon steel construction
also included a condenser tubed in brass. If the cycle chemistry was set up
with low pH (9.2), ammonia dosed feedwater to minimise corrosion of the
condenser, modelling of the LP evaporator showed a high risk of FAC in the
bends feeding into the LP evaporator outlet header. If the chemistry was
optimised to limit FAC attack, by increasing the LP feedwater pH to around
9.8, the condenser would be at serious risk of corrosion from attack by
ammonia and oxygen. The proposed solution was to dose with a volatile
amine mixture, which would maintain the water phase pH in the LP
evaporator due to its lower partition into the steam phase.
A second example is of a similar two pressure CCGT, with horizontal tubed
HRSG and combined LP drum / deaerator. This plant also has steam export
both for GT NOx suppression and to a third party consumer. The HRSG has
1% Cr return bends, but this did not entirely eliminate the risk of FAC in the
LP evaporator. The original cycle chemistry had low pH (9.2) feedwater to
match the proposed brass tubed condenser. This was subsequently changed to
a stainless steel tubed condenser, at the design stage, which allowed a higher
feedwater pH and ammonia dosing level to combat FAC.
The above examples are concerned with FAC of bends in LP evaporators.
However, there are a number of reports of serious FAC attack on the steam /
water separators in LP drum boilers.
3.5.5

Phosphate Hideout in Drum Evaporator Circuits

The use of sodium phosphate for pH control in drum evaporator circuits has a
long history, but also produces some significant problems. Sodium phosphate
has been preferred over the simpler alkali, sodium hydroxide, because it has
the ability to form a buffered pH solution if it concentrates within the boiler.
Sodium hydroxide has the potential to concentrate to a degree at which
corrosion can occur, if the dosing is not properly controlled. This degree of
control for sodium hydroxide has been successfully applied in a number of
utilities.

(63)

Sodium phosphates have a complicated chemistry, with complex interactions


with both hardness salts (calcium and magnesium), as well as the iron oxides
which inevitably occur within a carbon steel system. In addition, as the
temperature rises the solubility of sodium phosphate falls dramatically, so that
there is little scope for concentration at the tube surface before the sodium
phosphate precipitates out of solution. All of these factors contribute to a
phenomenon known as phosphate hideout. Simply, at operating
temperatures and pressures phosphate dosed to the evaporator water
disappears from solution, i.e. hides out, in the evaporator, only to reappear
again when load is reduced (hideout return). If dosing to a pre-set phosphate
reserve in the evaporator water, excess phosphate will be dosed. On load
reduction the phosphate returns, raising the boiler water conductivity and
phosphate levels and also affecting the boiler water pH. The complex nature of
phosphate chemistry means that the pH can be either raised or lowered,
depending on the particular interactions involved.
Phosphate was originally dosed as tri-sodium phosphate (TSP), but
subsequently mixtures of di-sodium hydrogen phosphate and TSP became
more popular, with the aim of maintaining a fixed sodium:phosphate ratio of
2.6 2.8. To maintain this ratio under hideout conditions some operators
added mono-sodium di-hydrogen phosphate, but this resulted in an acid
phosphate attack on the evaporator tubes. Subsequently, further phosphate
based chemistries using TSP or TSP plus up to 1 mg/l (ppm) sodium
hydroxide have been developed [47].
Phosphate hideout was generally a phenomenon observed in conventional
fossil fuelled higher pressure evaporators, with higher heat fluxes than
normally found in HRSGs. Therefore, the initial view was that phosphate
hideout would not affect HRSGs and certainly not the lower pressure circuits.
This has proved to be a false conclusion. A significant number of HRSGs have
suffered from phosphate hideout in their high pressure circuits, but equally
some have shown significant phosphate hideout in the low pressure circuits,
particularly those with horizontal tube evaporators.
One design feature that has been identified as responsible for hideout is hot
gas by-passing of the evaporator tube banks. In theory all tubes see a similar
gas flow and heat input. In practice there can be variable gaps between the
boiler casing and the tubes nearest the wall. This allows hot gas to
preferentially bypass some of the stages so that the gas seen by the next lower
temperature stage is higher than design. This increases the heat input into the
tubes at the edge of the evaporator, consequently increasing the evaporation
rate and the concentration factors of any salts in the evaporator water. If a salt
such as sodium phosphate, with a limited solubility, is present it will
concentrate and deposit in this area of high heat input and hide out. In severe
cases even a minor load reduction, i.e. variation in heat input, will move the
evaporation zone and some of the phosphate will be taken back into solution.
This will increase the evaporator water conductivity and affect the pH.

(64)

This was the case at one particular plant with horizontal tubed HRSGs. There
was phosphate hideout in the HP evaporator, but very serious hideout in the
LP evaporator. The latter was traced to hot gas bypassing producing an
increase in temperature of up to 40oC at the walls relative to the main part of
the gas flow. Figure 30 shows the temperature profile across the gas path
immediately above, i.e. at the exit from, the LP evaporator.

Figure 30: Temperature profile immediately above LP evaporator (Courtesy


of Power Technology).
The preferred solution at this plant, which was cooled by estuarine water
would have been to move to an all volatile chemistry. However, the risk of
FAC in the LP evaporator was too high and the LP system has been
successfully dosed with sodium hydroxide, to maintain a pH in the range 9.5
9.7. The greater solubility of sodium hydroxide relative to sodium phosphate
ensures it does not hide out. The high pressure circuit has been operated on an
all volatile chemistry, ammonia dosed, except when there are condenser leaks
when it reverts to sodium hydroxide dosing, again maintaining the pH within
the 9.5 9.7.
3.5.6

Organic Amines for Feedwater pH Control

The use of volatile organic amines as feedwater conditioning agents has the
advantage of using alkalis with lower steam volatility so that the pH of the
water phase in evaporators will be maintained at a higher level than for the
same feedwater pH when dosed with ammonia. Volatile organic amines also
have applications where condensate is returned from an industrial process via
condensate lines which contain both liquid and vapour. If ammonia is used
under such circumstances a large proportion of the ammonia partitions into the
(65)

vapour phase, leaving the water phase at a lowered pH which can promote
corrosion of the condensate pipework. However, organic amines have two
limitations, their significantly higher cost than ammonia and their propensity
to degradation to low molecular weight organic acids in the evaporator
circuits. The higher the temperature and pressure the greater the risk of
degradation, and the degree of degradation will also depend on the type of
amine used. If the steam is to come into contact with food products or
packaging, only morpholine or cyclohexylamine are allowed, and these have a
relatively high rate of degradation. Examples of other organic amines used in
non-food applications, or for pure generation CCGTs, are methoxypropylamine, ethanolamine, diethylamine and amino methyl propanol.
3.5.7

Effects of Organic Matter in the Steam / Water Cycle

The low molecular weight acid breakdown products from organic amines, are
typically formates and acetates. The latter can also arise from the thermal
degradation of natural organic matter derived from the raw water. Some of this
organic mater may pass through the ion exchange based make-up water
treatment plant and into the feed water. The effects of such breakdown to
organic acids will be two fold. As volatile anions they partition between the
steam and water phases. The proportion that pass into the steam will increase
the steam conductivity, measured after cation exchange. The proportion that
remains in the water phase will lower the pH of evaporator water. This is a
particular problem where all volatile chemistries are in use, as both ammonia
and organic amines are weak bases with relatively little neutralising power.
The net effect is an increased corrosion risk. Evaporators with non-volatile
alkali dosing (NaOH or Na3PO4) are less susceptible to the effects of organic
acids and the dosing to the evaporator water can be adjusted to maintain the
pH within target. However, in plants where there is a high make-up demand
and significant organic matter in the make-up water, there can be noticeable
pH reduction effects in evaporator water dosed with non-volatile alkali.
Research has shown that organic matter, whether specifically identified as
acetate or formate or more generally identified as Total Organic Carbon
(TOC), preferentially concentrates in the low pressure evaporator circuit of
multi-pressure HRSGs. In the higher pressure circuits the concentration effect
is much less pronounced. This concentration effect occurs with both AVT and
non-volatile alkali treatment of the evaporator water, but is more pronounced
for non-volatile alkali dosing. Therefore, although the effects of volatile
organic anions may be most apparent in the high pressure steam, the most
effective method for their removal from the steam / water cycle is by
blowdown from the low pressure evaporator. Provided the organics mass
removed by blowdown is greater than the mass input from make-up, or other
sources, there will be an overall reduction in organic matter in the steam /
water system.
3.5.8

Cycle Chemistry Selection

A general principle is that the feed water purity is set at the targets appropriate
for the highest pressure circuit in the HRSG. Choice of alkali for dosing of the
feed water is dependent on the specific materials of construction, plant design
(66)

and any steam purity requirements, but generally the first choice is likely to be
ammonia.
Evaporator water chemistry is determined by a range of factors. For once
through designs, there is no evaporator dosing (see Section 3.5.10). For drum
type evaporators the choice of alkali, volatile or non-volatile is dependent on
plant design and on the risk of cycle contamination by cooling water. If the
CCGT has an air cooled condenser, there is virtually no risk of cycle
contamination, and all volatile chemistries may be considered. If the
condenser cooling water is highly saline there is a much higher risk of
contamination via condenser leaks and a non-volatile alkali dosing regime is
preferred as this gives additional protection against acid chloride attack.
In single pressure systems the feed and evaporator water chemistry is selected
on the basis of operating pressure and the required steam purity. As the
operating pressure increases, the purity requirement for the feed water and
evaporator water also increases, i.e. less salts and other impurities are
acceptable.
Where the HRSG has multiple pressure stages it is preferable, for operating
simplicity, to select the same chemistry for each circuit. This was the case
with early HRSGs, but experience has since shown that this is not always the
best option. As a general principle the low pressure circuit should be dosed
with a non-volatile alkali, TSP or sodium hydroxide, both to protect against
FAC and promote the concentration of organic acid anions in that circuit. The
only exception is for the combined LP drum/ IP and HP feed tank design,
which requires a high pH volatile alkali dosing. The operational and corrosion
consequences of this design should be fully reviewed before acceptance.
The intermediate pressure drum circuit generally operates at design
temperatures that are above the FAC risk area. However, part load operation,
on sliding pressure, may reduce the operating temperatures in the evaporator
to the upper end of the FAC risk envelope. Therefore, a non-volatile alkali
dosing regime is the preferred choice. However, a volatile alkali regime may
be an acceptable option, particularly with air cooled condenser plant, either via
high concentration ammonia dosing or organic amine dosing to the feed water.
If there is no evidence of hideout in high pressure drum circuits a non-volatile
alkali is acceptable, although some main plant contractors prefer to use a
volatile alkali regime so as to minimise the risks of steam contamination. This
is perfectly acceptable, but should be accompanied by a risk assessment of the
potential for impurity ingress into the circuit. As noted above, for plant with
air cooled condensers, the all volatile regime is to be preferred. For plant with
water cooled condensers, the integrity of the condenser and the salinity of the
cooling water have to be taken into account. High pressure drum evaporator
circuits have been successfully operated on a volatile alkali chemistry with
high salinity cooling water. However, sufficient control procedures must be in
place to ensure the appropriate actions are taken in the event of a condenser
leak. Generally, this will be an increase in blowdown and re-instatement of
solid alkali dosing.
(67)

3.5.9

Steam Purity from HRSGs

Steam injection into gas turbines has been practised for many years to cool the
combustion path and reduce the NOx emissions. Steam purity has not been an
issue for this application as the steam purity required for use in a steam turbine
is significantly better than that required for a gas turbine. However, in the next
generation of combined cycle plant, steam from the HRSG is used to cool the
blades of the GT combustion system by passing steam through internal
passages within the blades (see Section 4.3.1). In order to avoid deposition or
corrosion within these passages the steam has to be extremely pure. Whitehead
and Rowe [48] noted the extremely low particulate concentrations, < 0.12 _g/kg
(ppb), required for steam used in cooling GE gas turbines. Geary and Bellows
[49]
reported on the limits for soluble impurities in steam, when used for
cooling Siemens-Westinghouse GTs. Table 4 compares typical steam purities
specified for a steam turbine and for blade cooling in a gas turbine.
MAXIMUM IMPURITY
CONCENTRATION
Sodium (_g/kg Na)
Silica (_g/kg SiO2)
Iron (_g/kg Fe)
Copper (_g/kg Cu)
Chloride (_g/kg Cl)

STEAM TURBINE
(Typical)
5 - 10
10 - 20
20
2-3
5
(where specified)
0.2 - 0.3

Conductivity - after cation


exchange at 25oC (_S/cm)
Ammonia (_g/kg NH3)
No limit
Phosphate (_g/kg PO4
Not specified
Sulphate (_g/kg SO4)
Not specified
Degassed Conductivity
Not specified
after cation exchange at
25oC (_S/cm)
_g/kg = ppb, _S = micro Siemens.

GAS TURBINE
BLADE COOLING
1
30
5
0.01
1
0.11
2500
1
1
0.06

Table 4: Comparison of steam purity for steam turbines and gas turbine blade
cooling [49].
From the above table it is clear that for most impurities the maximum
allowable limit for gas turbine blade cooling is at least an order of magnitude
lower than for admission to steam turbines. To achieve the steam purities for
GT blade cooling on a routine basis, for all parameters, requires a modified
approach to HRSG evaporator chemistry. Assuming measures are taken to
limit volatile impurities, by effective treatment of make-up water, the major
source of impurities in steam is the physical carryover of droplets of water. To
achieve the very high steam purities demanded for GT blade cooling requires
strict limits on the impurities in the evaporator water. It also depends on the
drum pressure from which the cooling steam is taken. Mechanical carryover
increases as the pressure rises, being around 0.03 0.04% at 40 bar and
increasing to 0.1% at 140 bar pressure. Therefore, mechanical carryover is

(68)

significantly less from a 40 bar IP drum compared to a 130 bar HP drum. To


achieve steam of a specific impurity target, extraction of steam from an IP
drum allows greater impurity levels in the drum water compared with
extraction of steam from an HP drum. Therefore, use of IP steam places less
stringent demands on the drum water chemistry control. Two major sources of
impurity remain: condenser leaks and evaporator water conditioning
chemicals. Geary and Bellows [49] showed that if steam was extracted from an
IP drum, controlled dosing of low concentration of non-volatile alkalis could
still be used, although the alternative of using AVT chemistry gives a greater
degree of security. If AVT chemistry is considered a parallel study of FAC
risks must be undertaken and appropriate design measures taken to limit FAC.
Condenser leaks could be tolerated if small or if sufficient blowdown
capability is available. The alternative, to ensure the availability of high purity
feed water, at all times, is to install condensate polishing. This is an expensive
solution, both for capital and revenue expenditure.
3.5.10 Once Through HRSG Chemistry
The majority of HRSGs built to date are sub-critical drum type. A limited
number of sub-critical once through HRSGs have been built, although the low
pressure circuit is normally a drum type evaporator. Supercritical once through
HRSGs are still to be introduced. Certain designs of GT have used GT
compressor air coolers that are in effect small once through steam generators.
In principle the cycle chemistry requirements for once through HRSGs are
identical to their fossil fired counterparts. Because all impurities in the feed
water either pass through into the steam or, in the case of sub-critical
evaporators, may deposit in the evaporation zone, the allowable impurity
levels are very low and are generally equivalent to the impurity targets set for
the steam. In the past this high degree of feed water purity has been achieved
by installation of full flow condensate polishing. However, Alstom have
introduced a design without deaerator or condensate polishing [50]. This relies
on high purity feed water, plus an occasional operation where the load on the
boiler is reduced and the evaporation zone is flooded with feed water. This
dissolves deposited impurities and is then disposed of to drain. The system
relies on a high integrity condenser, but should any condenser leak occur the
HRSG must be taken off load. Therefore, the benefits of a low capital cost
CCGT must be balanced against the lack of operational flexibility afforded by
a condensate polisher.
Condensate polishers have also been provided for HRSGs with steam turbines
located on saline cooling waters or where a significant proportion of the makeup is return condensate from a third party process, where contamination is
possible. This is often a site specific decision.
3.5.11 Dew Point Corrosion on the Gas Side of HRSGs.
The risks of acid dew point corrosion in cooling gas streams are well known.
In general HRSGs are designed to ensure that the gas temperature at the exit of
the last stage of the HRSG does not fall below the acid dew point for the gas
or fuel oil supply to be burnt. In the UK most gas supplies have very low
sulphur contents, although distillate fuel oil may contain more sulphur.
(69)

Therefore, the risk of sulphuric acid dew point attack is minimal as the dew
point is well below the gas exit temperature. However, some HRSGs are fed
with exhaust gas from GTs burning a slightly sour gas, up to 2000 vppb
sulphur equivalent. While operation at base load provided a gas exit
temperature above the sulphuric acid dewpoint, operation at part load reduced
the gas temperature at the outlet of the condensate pre-heater to the point
where the dew point was breached and sulphuric acid deposition occurred
(Figure 31). This particular plant cycled between base load and part load on a
daily basis. The deposit layer was very highly concentrated sulphuric acid,
which also attracted very fine particles of siliceous material and particles of
gas duct internal lagging. The fact that the sulphuric acid remained highly
concentrated precluded corrosive attack on the tube or finning, while the plant
was in operation. Off load moisture ingress into the HRSG via the rain damper
mobilised some of the acid deposits down the casing walls. One option being
considered to prevent further deposition is to bypass the preheater at low loads
(a bypass line is fitted), although this would result in a further performance
penalty.

Figure 31: Condensate preheater deposits (Courtesy of Power Technology).

3.6

Control and Instrumentation Issues on HRSG Plant

Modern CCGTs are operated with a large degree of automation to minimise


the risk of plant trips and other damaging incidents. Automated sequences are
used for common plant procedures such as start-up and shutdown, minimising
the participation of the operator and hence the risk of error and variability
between the actions of different personnel. However, these sequences can stop
partway due to failures on field devices such as limit switches and
thermocouples. Moreover, control and instrumentation problems on some
plants mean that a high level of spurious and consequential alarms may be

(70)

initiated and presented to the operator. In this situation, there is a danger that
important alarms could be overlooked, leading to a higher risk of plant trips.
Some of these problems are related to the presence of poorly specified
instrumentation, particularly actuators and valves, and this means that
upgrading of such equipment is an ongoing process on new plant. It is known
that the incidence of plant trips that are control and instrumentation related
drops considerably in the first months and years of operation from levels that
can be as high as 50% during commissioning. It is vital to eliminate spurious
trips due to faulty instrumentation early in a plants lifetime, as these are very
damaging. For example, calculations performed on a P91 superheater header
with full penetration welds under an optimised hot-start/shutdown procedure
demonstrated that a hot restart following a unit trip is 41 times more damaging
in terms of thermal fatigue damage than a hot start following a normal
shutdown [51].
The level of detail provided on some sequence displays is often insufficient to
identify the plant condition preventing a sequence from progressing. The
sequence logic is often so complex that even minor faults can be difficult to
pinpoint and rectify, with sequences having to be overridden manually and
stepped through to try and identify the sequence hold. It is usually essential to
have a member of staff proficient in control and instrumentation issues
available to deal with any automation related problems that may arise.
The start-up of a unit can be potentially influenced by a wide range of active
operational constraints generated from within the GT itself and also from the
HRSG and the ST. These constraints adversely affect the run-up process by
inhibiting firing on the GT until the current active constraint has been relieved.
This gives rise to the risk of variable run-up times for each start-up on each
unit and clearly becomes even more relevant in the event of moving the
operational regime away from base-load. Faults with field devices can
exacerbate this, making it difficult to predict overall run-up and loading times
with absolute accuracy. The above factors have become highly significant
under the New Electricity Trading Arrangements (NETA), where significant
financial penalties can exist for not getting up to load on time.
3.7

Flexible Operation of HRSG Plant

To take economic advantage of fluctuations in the wholesale price of


electricity, it has become advantageous within certain markets (particularly the
UK) for generating plant to operate flexibly. This may involve regular twoshifting where plant is taken off load for several hours overnight, shut down
at weekends and/or fluctuations between full load and part load or minimum
stable generation. In the UK, the volatility in gas price and increased
dominance of gas generation also means that companies with a portfolio
comprising generation reliant on more than one fuel source can make up their
contracted output as they see fit. These factors have resulted in many CCGTs
(and hence HRSGs) designed for largely base-load operation being subjected
to a flexible operating regime, with many plants conducting daily starts for at
least part of the year. Many of the effects of flexible operation on HRSG

(71)

components in terms of thermal fatigue damage, cycle chemistry and


control/instrumentation are described in Sections 3.3, 3.4, 3.5 and 3.6 above.
As far as the HRSG is concerned, the effects of load fluctuation are small
compared to those arising from plant shutdown/start-up where the temperature
differentials and ramp rates involved are much more damaging in terms of
fatigue life.
Power Technologys approach to assessing the level of risk to HRSG
components under a flexible operating regime is to carry out a flexible
operation study, which is conducted in three phases:
Phase 1 involves establishing the existing level of instrumentation on the
HRSG and highlighting those that would be required for a series of monitored
flexible operation trials. The critical HRSG components at risk of early
failure/increased degradation or that could create operational problems under a
flexible operation regime are also identified based upon station-specific HRSG
component materials and geometry, inspection results, previous flexible
operation experience and known problem areas on the plant being studied. The
number and location of additional instruments (typically thermocouples on
boiler headers and stubs) required for flexible operation trials are then
specified, with their installation justified against the potential risks identified.
Thermocouples can also be fitted to structural components such as expansion
joints, duct supports, casings and so on to quantify the temperature
differentials across them.
A desk study of the effects of flexible operation on water/steam chemistry
would also be completed and would typically review phosphate hideout and its
control, flow accelerated corrosion risk in low temperature / pressure circuits,
water treatment plant capacity, steam quality at start-up, de-aeration capability
on start-up and condenser integrity.
After the additional instrumentation required had been installed, a programme
of shutdown/start trials would be agreed with the station for Phase 2 and all
plant data received would be processed and analysed. The extent of any
damaging transients (specifically ramp rates, through-wall temperature
differentials and peak temperatures) would be quantified and a detailed
thermal fatigue stress analysis carried out on the worst affected components.
The damage would be quantified with the results expressed in terms of impact
on component life and the likelihood of any failure mechanisms. In addition,
one or two starts would be observed on site to fully understand any operational
problems being experienced at first-hand. Appropriate recommendations to
manage any issues identified would then be made. These might typically
include enhanced, targeted non-destructive testing/visual inspections,
proposed modifications to operating procedures to minimise impact on HRSG
components and/or changes to the cycle chemistry.
Phase 3 (if required) would explore in more detail with the station the
feasibility of any proposed modifications to operating procedures in order to
reduce component damage and / or reduce start time. Where the required
changes are significant, it may be necessary to widen the scope of the study to
(72)

include the gas and steam turbines and to investigate the implications for
control and instrumentation. Trials of the optimised two-shift procedure would
then be carried out and the extent of improvement assessed.
Many of the design features that determine how flexible a HRSG is are
described in Section 3.2. However, the quality of manufacture and
construction and the operational procedures adopted also play a key role in
ensuring that the effects of flexible operation (predominantly thermal fatigue
damage) are minimised. Manufacturers are aware of the threats posed by
cyclic operation, and have already started to address the issues raised by the
flexible operation of power plant in response to customer demand.
Both vertical and horizontal gas flow designs can be equally suited to flexible
operation providing sufficient measures are taken during the design stage.
Horizontal gas flow designs tend to be more susceptible to flexible operation
damage due to lack of flexibility between the header systems (particularly if
bottom-supported), though this is being addressed on more modern designs.
The use of serpentine tube banks on vertical gas flow HRSGs inherently
makes them more mechanically flexible, although there can be difficulties
with drainage of the horizontal tube banks leading to the risk of off-load
corrosion.
Relatively simple ways of improving the ability of an HRSG to withstand the
rigours of flexible operation include the correct sizing of drains and vents and
the use of bypass valves and recirculation systems. GT and control and
instrumentation reliability are important in avoiding trips, as hot re-starts are
particularly damaging to upstream HRSG components. More substantial
features such as the inclusion of a stack damper or the use of higher-grade
alloys to reduce component thickness should be made at the design stage, as
these are much more expensive to retrofit later in the plant life.
3.8
3.8.1

HRSG Costs, Reliability and Maintenance


Capital Cost

Capital costs for new build CCGT plant are difficult to predict accurately
without going out to tender, and even then a wide spread of costs can be
possible at any given time. However the approximate total project cost for a
new build CCGT in the UK [52] is estimated to be in the region of 425/kW.
This includes not only the EPC (engineer/procure/construct) contract, but also
other items such as project management, connection to gas/electricity
networks etc. The HRSG it likely to account for around 10-15% of this total.
This is still significantly lower than for other forms of fossil fuel generation
e.g. the equivalent capital cost for a new build advanced pulverised fuel plant
is around 800/kW [53].
CHP plant capital expenditure is generally more expensive at around 750/kW
[52]
with the HRSG likely to account for 10-15% of this total.

(73)

3.8.2

Operating Costs

Fixed operating costs, excluding fuel, are estimated to be around 12/kWyear


for an existing CCGT plant [54], although this could be higher for the most
advanced class of GT plant due to GT maintenance costs. Generally, though,
this represents the lowest value across the range of fossil fuel plant, the next
lowest value being that of 14/kW for an existing coal fired plant without
FGD. These CCGT fixed operating costs will again be dominated by the gas
turbine, perhaps even more so than with the capital costs.
Taking into account all costs (fuel cost at 23p/therm, fixed operating cost and
cost of capital), the estimated delivered energy cost for a CCGT in the UK is
around 2.2p/kWh [53].
3.8.3

Reliability

Reliability/availability will vary greatly depending upon the original build


quality and design of the HRSG, the operating regime and the maintenance
performed. EPRI [55] predicted theoretical total availability of a drum type
HRSG to be 98.52% e.g. 1.48% availability loss. Powergen data from 19971999 [56] suggests a figure of around 0.2% average HRSG availability loss,
which may be due to the relative youth of the plant and a fairly tight functional
specification. The losses [56] appear to be mainly due to one of three causes;
tube leaks, leaks from flange connections or trips due to incorrect (high or
low) drum level on start up.
A survey of the causes of tube leaks on Powergen CCGT and CHP plant
indicates that around 50% are due to wear out mechanisms such as flow
accelerated corrosion, fretting, long term overheating, on load corrosion, stub
weld cracking etc. The remaining 50% can roughly be categorised as arising
either from original manufacturing (usually weld) defects / previous site
repairs or of being of a miscellaneous nature [57].
3.8.4

Maintenance

As well as the scheduled routine maintenance (e.g. valve/pump maintenance,


safety valve maintenance and testing, instrumentation checks, etc), typical
preventative actions would include annual HRSG visual inspections of:

HRSG/duct supports, expansion and alignment.


HRSG/duct external framing & internal stiffeners.
HRSG/duct internal insulation (if fitted).
Bypass damper and stack.
Main HRSG stack.
Tube modules and headers.
Duct fabric expansion joints (including thermal imaging whilst on-load to
identify areas operating at above-design temperatures and areas of gas
leakage).
The condition and tightness of pipe penetration seals
The condition and movement of main feed and main steam pipe supports.

(74)

The following measures are also typically taken during major outages to
guarantee the continued integrity of the HRSG through its design life (which
may range between 15 and 30 years):

Sample header / drum internal inspections for corrosion/debris, thermal


fatigue, blockage and/or FAC of orifices.
Measurement of creep pips (header diametrical measurements).
Sample header end-cap and butt weld inspections.
Pipework butt-weld inspections.
Tube sampling/thickness checks (for flow accelerated corrosion and offload corrosion).
Valve casting inspections.

3.9

Industrial HRSG Applications

HRSG applications are more diverse at the industrial scale and can be broadly
classified as below.
3.9.1

Industrial Gas Turbine HRSGs

At the small scale, gas turbines may be used in smaller CHP schemes or to
provide shaft power e.g. for pumping stations on gas pipelines. In CHP
schemes the demand is usually for process steam (e.g. for enhanced oil
recovery) and the generation of electricity by using a GT to burn the fuel and
generate electricity rather than just burning it in a boiler is an economic bonus.
Because the provision of steam to the process is usually the paramount
concern, an auxiliary burner is usually fitted to allow continuation of boiler
operation even when the GT has tripped. In this case a fresh air inlet duct is
needed. Units may also be installed for marine use in gas turbine driven ships,
floating production storage and offloading vessels (FPSO) and offshore
platforms. GT based CHP schemes typically achieve an electrical efficiency of
around 23% (GCV) and a heat efficiency of around 49% (GCV) [7].
In recent years a new breed of microturbines has been introduced, based on
turbocharger rather than aero-derivative technology. These are usually in the
range up to 0.5MWe, at which scale the aero-derivative type becomes more
economic. At present a typical microturbine unit from Bowman Power
Systems has an output of 80kWe and a thermal output of 130 260 kWth in an
exhaust gas stream at a temperature of around 600C [58]. Most units installed
so far have recovered heat as hot water, but in some specialised applications
steam has been generated.
3.9.2

Reciprocating Engine Exhaust Gas Boilers

Internal combustion engines may be used on a small scale for electricity


generation and HRSGs may be added to run in combined heat and power
mode. Low grade heat is usually recovered as hot water from the engine
cooling circuit. Higher grade heat may be recovered from the exhaust as
steam. Normally smoke tube design package units are used to generate
saturated steam. Engines may be run on liquid or gaseous fuels. Increasingly,
alternative fuel sources are being used, such as landfill gas, coal mine methane
(75)

and bio-gas from anaerobic digestion of sewage sludges, agricultural wastes


and industrial wastes. Reciprocating engine based CHP schemes typically
achieve an electrical efficiency of around 27% (GCV) and a heat efficiency of
around 42% (GCV) [7].
3.9.3

Heat Recovery from Other Industrial Exhaust Gases

HRSGs are used to recover energy from the hot exhaust of other industrial
processes such as:

Glass and metallurgical furnaces


Kilns (e.g. sponge iron plants): Coal based high temperature reduction of
iron ore produces a flue gas with a temperature in the range of 10001200C and a dust load as high as 40gNm-3. The first stage of the heat
recovery system is a radiant section with water membrane panel walls
where the gas is cooled to around 750C to reduce the risk of slag
deposition on downstream heat transfer surfaces. This is followed by a
plain tube superheater fitted with soot blowers and evaporator and
economiser sections.
Roaster based plants: a typical application is in roasting of pyrite ores.
Pyrite ores are oxidised in a fluidised bed to produce the oxide required for
manufacture of the primary metal. The flue gases contain sulphur dioxide
and are at a temperature of around 900C with a high dust load. The
Thermax design [59] uses a water tube boiler to recover heat. A vertical
tube alignment and a wide tube pitch are used to minimise problems of
dust deposition. A hammering device dislodges dust from the tubes into
hoppers below from which it is continuously removed.
Smelters and converters: for example in copper and zinc smelting. The
high temperature waste gas has a very high dust load. The heat is
recovered in two stages. In the first stage the gas passes through a large
water membrane walled radiant section where some of the dust and slag is
allowed to settle. The partially cleaned and cooled gas then passes through
a conventional convection section with vertical bare tubes, again fitted
with hammering devices to dislodge dust [59].
Coke ovens
Solid / liquid / gas waste incinerators
VOC thermal oxidisers

3.9.4

Process Integrated HRSGs

Many process industries use HRSGs to recover heat from the necessary
cooling of process gases. Industries include:

Petrochemicals (e.g. in sulphur recovery units)


Sulphuric acid plants: The double conversion double absorption process
for the manufacture of sulphuric acid from elemental sulphur generates
considerable heat in the exothermic conversion of sulphur to SO2 and SO 3
gases. The optimum working temperature for the V2O5 catalyst is about
440C, so it is essential to have a process integrated HRSG in the system
to cool the gas to this temperature.

(76)

Nitric acid / Caprolactum plants: nitric acid may be manufactured by the


catalytic oxidation of ammonia at around 950C to give a gas rich in
nitrogen oxides. The next step of the process requires a gas temperature of
around 250C creating a requirement for a HRSG. Either water tube or fire
tube designs may be used in either case the design needs to take into
account a typical gas side pressure of 5 7 barg.
Ammonia plants
Hydrogen gas plants
Fluidised catalytic converter units

3.10 Conclusions

Current state of the art utility scale HRSGs operate at HP steam conditions
of up to 124 bar/565C and allow the associated CCGT to deliver
electrical power at a claimed net efficiency of up to 60%. They are
generally two pressure or three pressure with reheat, and may be of either
vertical or horizontal gas flow.

The capital cost of new-build CCGT plant is around 425/kW, with the
HRSG accounting for 10-15% of this total. The estimated delivered energy
cost for a CCGT in the UK is around 2.2p/kWh.

Current state of the art industrial HRSGs generally operate at lower steam
conditions than utility scale plant, and are usually of single pressure
design. They are integrated into a wide range of industrial plant and often
include provision for supplementary or auxiliary (stand-alone) firing.
Highly fired units may incorporate a water-cooled furnace. Lower pressure
industrial boilers are usually of shell rather than water tube design.
Designs tend to be bespoke for particular process applications.

The recent trend has been for CCGT plant to be built under turnkey
contract. Whilst this does have advantages to the user in terms of
accountability, it does tend to mean that the user has less influence on the
detailed HRSG design.

Operational experience with HRSGs indicates that inclusion of specific


design features and attention to detail during fabrication are just as
important as the overall HRSG design, and that non pressure parts can be
as problematic as pressure parts.

Key areas for improvement include build quality, access for in service
inspection & maintenance, control & instrumentation and capability for
flexible operation. Overall cycle chemistry philosophy also needs to be
more thoroughly considered at the design stage.

The current challenge for operational HRSGs, particularly in the UK, is the
need to cycle plant which has been designed for and/or previously operating at
base load. Many users are currently carrying out investigations/trials and plant
modifications.

(77)

NEW AND DEVELOPING TECHNOLOGIES

4.1

Introduction

This chapter describes and reviews HRSG technologies which are envisaged
as becoming available in the near and longer term. A number of HRSG
suppliers and users were consulted during the preparation of this report about
technological developments and the need for further research. The consensus
of opinion amongst those consulted was that the technology is mature and that
large or revolutionary advances in technology are not expected. However,
small incremental improvements are expected to continue. None of the
industrial scale companies consulted stated that they have research projects of
their own going on currently. At the industrial scale, most businesses are not
large enough to take on large R&D commitments. Developments at the utility
scale, such as the use of higher temperature materials, will cascade down to
the smaller industrial scale market eventually and give gradual advances. A
number of new applications or small areas of technological advance were
identified and the following specific categories have emerged:

Developments in the design of HRSGs themselves.


Developments in other parts of combined cycle plant or the overall cycle.
New applications.

4.2
4.2.1

Developments in HRSG Design


Utility Scale Once Through HRSG Designs

In terms of components, the once-through steam generator is the simplest


HRSG design for recovering heat from the exhaust of a gas turbine. Water
entering at the cold end of the gas-pass, moves through a serpentine tube
bundle where heat absorption occurs and a phase change takes place, and exits
as superheated steam. The circulation ratio is one and there is no requirement
for circulation pumps.
Conventional (i.e. not once through), sub-critical HRSGs utilise drums in
which steam and water from the evaporative part of the cycle are separated.
The water is then recirculated within the evaporator with additional feedwater
while the steam passes to the superheater for further heating. Supercritical
pressure boilers cannot utilise this type of design as there is no distinct
water/steam phase transition above the critical pressure. A once-through
design is therefore required. The OTSG design also has advantages for flexible
operation. The steam drum is the component in a conventional HRSG design
with the thickest wall section and is therefore the most prone to the occurrence
of stresses associated with differential thermal expansion. It is the limiting
component in setting maximum heat-up and cool-down rates and a design that
eliminates the drum is therefore better for flexible operation.

(78)

Once-through technology has in the past generally been limited to projects


based on aero-derivative or small industrial gas turbines. However, designs
that can handle the larger frame gas turbines are now evolving.
With regards to utility HRSG applications in the UK, once-through technology
is currently at a demonstration stage with a full-scale once-through
demonstration HRSG (supplied by Deutsche Babcock, now Babcock Borsig)
currently in operation at Cottam Development Centre. This is a sub-critical
Benson design with superheater steam conditions of 580C and 160 bar and is
described further in Section 7.3.1. This technology, which originated with
Siemens, is currently licensed to a number of other companies including
Nooter/Eriksen.
However in North America the commercial acceptance of once-through
technology is far more apparent. ABB have 7 once-through (sub-critical)
industrial sized HRSG units in operation and a further 23 under construction
and are moving towards larger scale applications with a significant new 270
MWe once-through HRSG built recently at Agawam in Massachusetts. In
addition, Innovative Steam Technologies (IST) of Canada won a contract for a
once-through technology plant at Calpines Broadriver Energy Centre,
although in this case the HRSG is only sized to provide steam for GT
injection.
Successful and extensive pilot trials have been undertaken by Cockerill
Mechanical Industries (CMI) of Belgium in their Seraing works [60] with a
view to achieving supercritical conditions in the once through HRSG.
Indications are that HRSGs will gradually adopt once through technology and
then move to supercritical pressures as gas turbines become larger and exhaust
gas temperatures continue to increase.
4.2.2

Industrial Scale Once Through HRSG Designs

The application of the OTSG design to CHP plants is sometimes limited by


the critical need for a continuous supply of steam for some users. In a
conventional drum HRSG design there is a significant reservoir of steam and
hot water in the drum. In the event of a GT trip, this will provide a buffer
supply of steam to maintain the supply to the users plant while the auxiliary
burner starts up and reaches the necessary output. The only water in the OTSG
is in the tubes, which does not provide such a large reservoir of steam. For
applications where maintenance of a continuous steam supply is critical,
provision of steam buffer capacity needs to be investigated. The higher
pressure drop on the water side of the OTSG has been identified as a minor
disadvantage of the design. The development of new balanced header designs
that distribute the flow evenly over all of the tubes will reduce this effect.
4.2.3

Reliability Improvements

Within the UK, as a direct result of the New Electricity Trading Arrangements
(NETA) in England and Wales, generators and suppliers now have to contract
directly with each other for the physical supply of power. The effect of this
and similar legislation throughout the world on the plants themselves, is that
(79)

many existing combined cycle gas turbine plants need to move towards more
flexible operation than was initially envisaged at their design stage.
This move from what was initially a system engineered for base loading to one
with substantial requirement for two-shift operation has a detrimental impact
on plant reliability, specifically with regards to the HRSG. For the plant, the
main risk associated with the use of the HRSG under flexible operation is the
impact on the achievable lifetime of pressure part and non-pressure part
components such as tubes, headers, casing components etc. In general, a
reduction in lifetime is expected resulting from causes such as Low Cycle
Fatigue (LCF), localised header stresses and flow accelerated corrosion (FAC)
and from a combination of LCF and Stress Corrosion Cracking (SCC).
The challenge therefore faced by engineers is to design solutions that ensure
the reliability of the next generation of HRSGs to be installed and upgrade the
existing plants (by correctly sizing drains for example) before major problems
occur.
In order to do this companies have had to invest heavily in effective transient
thermal modelling capabilities that allow them to analyse thoroughly every
component of the HRSG and make design changes to limit thermal stresses.
One such example where new features have specifically been designed to
provide flexibility for the plant during thermal transients is with Foster
Wheelers Fort Meyers repowering project [61]. This novel feature is a wet
bypass unit which is designed to absorb the instantaneous power loss of a
steam turbine trip and enable the gas turbine to continue operating at a full
simple cycle load. In the event of the steam turbine tripping, main steam is
attemperated and its pressure reduced, before it is bypassed to the condenser.
Reheated steam is either bypassed to the atmosphere, when the condenser is
not available or also attemperated and reduced in pressure before passing to
the condenser dump. Thermal fatigue of the steam headers is greatly reduced
by this innovative process.
Likewise Alstom have adopted features specifically to reduce thermal stress at
welded joints albeit at a potential increase in capital outlay. Most natural
circulation HRSGs use a multiplerow harp-shaped design, comprising of one
horizontal upper header and one horizontal lower header joined by two or
three rows of vertical tubes. Alstom maintains that, the temperature of the
exhaust gas drops sufficiently as it passes through the multiple tube rows to
cause the individual tube rows to operate at different temperatures, inducing
differential thermal stress at the weld joints. Their solution is to form a single
row of tubes between headers to remove these differential stresses. As a single
row allows for smaller header diameters, circumferential temperature
gradients in the headers are also minimised. Alstoms analysis concluded that
small-diameter headers reduce thermal stress by as much as 60% when
compared to headers used with multiple rows.

(80)

4.2.4

Modularity and Improved Maintenance Features

One noticeable innovation with regards to HRSGs is the drive towards


modular designs. Aalborg Industries Inc., Alstom, Mitsui Babcock and
Nooter/Eriksen are some manufacturers who have designed around the
concept of modularity in order to exploit it as a selling point for their HRSGs.
Aalborgs rapid delivery of standard units in less than 90 days alongside quick
field erection of their pre-assembled, pressure-tested systems has won them a
considerable HRSG business. This modularity extends to the point where
some of the auxiliary features of the HRSG are already in place (e.g.
feedwater systems, air and flue gas ducting, etc).
Nooter/Eriksen incorporate a modular design to increase shop fabrication and
minimise field man-hours. Their flexible construction method and attention to
detail allows setting of up to five modules per day.
Other manufacturers such as Foster Wheeler believe modularity to be
advantageous. Their approach has been to make modules as large as possible
leading to a requirement for fewer components to be assembled in the field.
Foster Wheeler maintain that the increased constructability of their HRSGs
helps reduce the risk of possible delays in the erection schedule and the
financial penalties that may then result.
Other companies have patented design features such as enhanced accessibility
to their HRSGs. The manufacturer Deltak, for example, claims this very
feature reduces repair time to half of the industry standard.
4.2.5

Control and Instrumentation

At the simple smoke tube design end of the industrial HRSG market, control
technology is being improved. Up to date touch screen control technology is
only now being introduced to these traditional designs in less demanding
applications.
4.2.6

Highly Fired HRSG Designs

An increasing number of industrial scale HRSGs are being used to supply onsite power and process steam requirements. In some instances the steam
demand substantially exceeds what can be supplied from the exhaust of a GT
meeting the on-site power demand, so a high degree of supplementary firing is
required. This results in a very hot, high moisture content gas flow and a need
to fire down to low oxygen levels while still meeting emission limits for CO
and NOx. Often the supplementary burners must be sized to maintain full
steam output even if the GT trips. This creates design challenges for HRSG
manufacturers. More exotic materials are required for highly fired HRSGs and
water membrane walls are becoming more common in designs for these
applications. These technologies are well established in fired boiler designs
and are now being transferred across to HRSG designs.

(81)

4.3
4.3.1

Improvements to Cycle and Other Plant Components


Steam Cooled Turbine Blades

For the gas turbine, the maximum allowable inlet temperature is governed by
both the materials available for the turbine blading and the cooling technique
employed.
Previous generations of gas turbines utilised air to cool turbine components in
an open loop system. The cooling air was supplied from a bleed in the gas
turbine compressor, and then ducted to the internal chambers of the turbine
blades and discharged through small holes in the blade walls. This air
provided a thin, cool insulating blanket along the external surface of the
turbine blade. As a result, there is a significant exhaust gas temperature drop
across the first stage nozzle and significant flow of air required to cool down
the relevant turbine stages. An integrated closed loop steam cooling system
significantly reduces this temperature drop in addition to eliminating the
requirement for air bleed for the turbine cooling. This technology is envisaged
as contributing around 2% points in thermal efficiency.
The thermodynamic advantage of utilising steam in cooling circuits was
recognised in the early 90s [6] (Figure 32). This has been developed
accordingly over the last decade to allow integration of the HRSG steam flow
with the gas turbine cooling loop to further enhance cycle performance. The
implications of this on steam purity are significant if corrosion and fouling of
the cooling passages is to be prevented. This is discussed further in Section
3.5.9.

(82)

Figure 32: Effect of gas turbine cooling methods on efficiency (Courtsey of


Innogy plc) [6].

(83)

The worlds first operational steam cooled gas turbine, built by MHI, was
commissioned in Japan in 1998 and installed in Unit 4 at the Higashi Niigata
Power Station. In the United States, it was not until early 2001 that SiemensWestinghouse achieved commercial operation with a 360 MWe power plant
located in Massachusetts followed immediately by a second 249 MWe plant in
Florida. The gas turbine installed in the American plants is considered
partially steam cooled, with just the first stage vanes being incorporated
within a closed loop. Whilst specific teething problems were found during
start up conditions of the W105G gas turbine, a cycle efficiency of ~58%
(LHV) was achieved with this technology.
More recently General Electric installed their first H-class unit at the Baglan
Energy Park in South Wales. This unit, which features steam-cooled rotor and
stator vanes, is currently under commissioning tests with engineers aiming to
break the 60% (LHV) cycle efficiency barrier.
4.3.2

Fuel Heating

In the late 90s methods of heating fuel prior to combustion in the gas turbine
were being introduced to the combined cycle in order to enhance efficiency.
Preheating of the fuel results in a reduction in the amount of fuel needed to
achieve a given firing temperature in the gas turbine. However, whilst the
efficiency of the cycle improves, the plant power output is found to reduce
slightly. This originates from the fact that when a gas turbine is fed warmer
fuel, it requires less mass flow of that fuel to obtain the previous cold-fuel
firing temperatures, thus the exhaust mass flow and water vapour content of
the combustion products is lower. Less power is therefore obtained from the
combustion gas expansion through the turbine. Furthermore the HRSG
generates a little less steam from the decline in gas turbine exhaust mass flow
and hence a drop in steam turbine power also occurs. However, the overall
improvement in the cycle efficiency results from the fact that the energy
diverted from producing steam power is of relatively low grade, and is better
employed as a heating medium for the fuel.
The fuel heating source may be either steam or water. For the case of steam
this can originate from the steam turbine bleed or directly from one of the
HRSG pressure level circuits. For the case of water heating, the hot water is
drawn from the HRSG economisers.
There is a threshold at which the benefits associated with the increase in
efficiency are found to be at the expense of the level of electrical power
produced [62]. For example fuel heating to around 200C from an intermediate
pressure economiser exit water source on a typical three pressure reheat
combined cycle, results in a net heat rate gain of about 0.6% with a
corresponding net power loss of about 0.3%. If the temperature of the fuel
was raised to around 300C by a high pressure economiser exit water source,
the loss of power becomes more evident at 0.75% whereas the noted increase
in the heat rate gain is less apparent as it only rises to 0.8%.

(84)

4.3.3

Gas Turbine Steam Injection for Power Augmentation

The injection of steam into a gas turbine for NOx control is well established.
Steam injection into the combustor reduces combustion temperatures by
diluting both the level of oxygen in the combustion air and the heat generated
from combustion of the fuel. However the use of steam injection for power
augmentation is considered to be somewhat under developed for larger heavy
duty gas turbines.
Injecting large amounts of steam for power augmentation creates a fairly
efficient power only plant that effectively makes the steam turbine, condenser
and cooling tower in a combined cycle redundant. Aero-derivative engines
such as LM 5000 and the Alison 501 were proven in the mid 80s to be suitable
for steam injection power augmentation [63]. The LM 5000 can absorb all the
steam generated from the heat recovered from its own exhaust, and in doing so
increase its power output by up to 15%.
The reason for the aero-derivative engines success at power augmentation was
because aero-engines were initially designed to pick up load faster than heavy
industrial gas turbines and consequently have a greater surge margin when
operating in an industrial application. Some of this surge margin can therefore
be exploited to accept steam injection.
However the limited electrical generation capacity of these gas turbines (<
50MWe) means that the benefits of this application have been limited to
mostly small scale industrial power supply uses. Although in some cases
several of these aero-derivative gas turbines have been successfully combined
to form a reasonably sized utility plant (i.e. 7 x 50MWe). Currently no gas
turbine greater than 50 MWe has been designed which allows for such heavy
steam injection that the need for a separate steam turbine is removed (as is the
case for the LM5000).
However, elaborate arrangements where steam/water is injected into large gas
turbines are in the development stage, albeit that relatively few have actually
been constructed. A current list and description of these proposals has been
generated by Foster-Pegg [64]. Those which have reached demonstration status
include:

Simple Steam Injected Gas Turbine SIGT Cycle: this system involves
moderate steam injection into the combustor of a standard large gas
turbine and has been used to augment power under hot ambient conditions.
Humid Air Turbine or HAT Cycle: this system involves evaporating
moisture into the air flowing into a gas turbine and requires a special gas
turbine in order to operate effectively. It has been highlighted as
particularly appropriate for gasification combined cycles.

The retrofit of gas turbine steam injection for power augmentation may be
particularly attractive in areas where there is existing industrial scale open
cycle GT plant and a demand for greater generating capacity. A variety of
HRSG designs could be used in this application, but it is a developing

(85)

application for which OTSGs may be particularly suitable. The OTSG can be
designed to run dry at full exhaust temperature, removing the need for a
bypass stack and providing a more simple and robust system.
4.3.4

Gas Turbine Inlet Air Chilling

Air intake cooling technology is used to enhance combined cycle power


output and is particularly relevant for utility plant situated in warm climates
where the air is dry and hot. The cooling of the air at the inlet results in a
noticeable increase in mass flow through the gas turbine. This in turn results in
a higher gas turbine power output as well as a slight increase in the steam
production in the downstream HRSG.
The cooling of the inlet air is achieved by means of a refrigeration system
similar to the type employed in large building air conditioning units. The heat
exchanger used to cool the incoming gas turbine air is formed from a coil of
finned tubes located in the inlet housing of the gas turbine. Cooled water is
circulated through the tubes. The prior cooling of the water is achieved by
chillers which are mass produced pieces of equipment and essentially fall into
two categories:

Centrifugal chillers
Absorption chillers

In general, the absorption chiller delivers a lower gross plant power output
than the centrifugal chiller due to the use of steam bleed from the HRSG to
drive it. However in terms of plant net power outputs the two systems are
approximately the same. This is because the centrifugal chiller requires an
electric pump for circulation and therefore consumes a far greater amount of
auxiliary power as opposed to the steam driven circulation for the absorption
method. Further differences between these two systems are described in
greater detail by Elmasri [62].
Chilling the inlet of a large combined cycle allows extra power to be obtained
from a plant (~ 5% increase in the net kWe output). The cost of that additional
power output is the additional expense of the capital and operational costs of
the upstream chilling unit and other necessary modifications to the gas turbine
itself. These have been estimated by Elmasri [62] at ~$250 per kWe of capacity
gained above the initial hot design condition.
The small-scale OTSG design may also find application in power
augmentation for existing open cycle GT plant by GT air inlet chilling, as
described above. A small OTSG unit can be added to an open cycle GT to
provide steam for an absorption chiller. Again the advantage of the OTSG
design is that protection in the event of a boiler trip is not required as the
OTSG may be designed to run dry at full exhaust temperature.
4.3.5

Increases in Gas Turbine Exhaust Temperature

The pursuit of higher efficiency CCGT plant has driven the rapid increases in
GT exhaust temperature and mass flow rate imposed on HRSGs. The exhaust
(86)

conditions of some established and more recently developed gas turbines are
shown in Table 5 below:
Gas Turbine

Siemens/Westinghouse
501G (a 60Hz machine)
GE 9351FA
Alstom GT 26
GE 9001H

Turbine inlet
temperature
(C)
1427

Turbine exhaust
temperature
(C)
600

Mass flow
(kg/s)

1327
1260
1427

606
641
621

636
549
685

535

Table 5: Turbine inlet/exhaust gas temperatures and mass flow rates on some
modern gas turbines [65].
The Alstom GT 26 (and GT 24) gas turbine uses sequential combustion in two
annular combustion chambers to achieve improved efficiency. This is different
to the conventional approach to gas turbine combustion, which is carried out
in a single stage and requires increasingly high firing temperatures and
complicated cooling technologies [66]. The exhaust temperature is the highest
of any commercially available gas turbine.
The GE 9001H is designed to give a combined cycle thermal efficiency of
60% and the first of its kind is being installed at Baglan Energy Park in South
Wales. The efficiency improvement is due to the high firing temperature,
which is made possible by the use of large single crystal airfoils, superior
component and coating materials and a closed-loop steam cooling system (see
Section 4.3.1) [67]. The implications of this type of cooling system for HRSG
water/steam chemistry are not trivial and are discussed in Section 3.5.9.
In general, the latest generation of gas turbines, with their increased gas
turbine outlet conditions are not anticipated to be a major concern as far as the
HRSG pressure parts are concerned, as materials issues at these temperatures
have been successfully tackled on conventional coal-fired plant.
For these increased temperatures, GT diffuser ducts, HRSG inlet ducts and
casings are likely to be internally insulated in the higher temperature regions.
Currently used internal insulation liner materials can cope with these relatively
small increases in temperature (internal insulation has been proven on CHP
plant with supplementary firing up to around 850C).
The most likely area to be impacted is that of non-cooled support components,
particularly on vertical gas-flow HRSGs e.g. HP superheater/reheater tube
sheets and support links. This may require more extensive use of higher grade
alloys such as modified 9% chrome (P91) or stainless steel for support
components. There may even be a need to employ water/steam cooled support
tubes for high temperature tube banks on vertical gas-flow HRSGs (akin to the
rear pass of a conventional coal fired boiler). This method of support has
largely been limited to supplementary fired vertical gas flow HRSGs in the
past.
(87)

4.3.6

Supercritical Technology

Currently state-of-the-art supercritical pulverised fuel fired steam power


generation plants exist and operate at up to nominally 300 bar and 600C
steam output with net efficiencies of ~45% LHV [68]. Due to advances in
materials technologies steam temperatures and cycle efficiencies have
gradually improved and are set to continue to do so. Targets of final steam
conditions of 650-700C have been set for 2020 and associated cycle
efficiencies of around 50-55% are expected.
The recognised advantages of adopting a supercritical steam cycle in addition
to the obvious improvements to cycle efficiency are [68]:

CO2 emissions are reduced by about 15% per unit of electricity generated
when compared with typical existing sub-critical plant.
Exceptionally good part load efficiencies are achievable, typical half the
decrease in efficiency exhibited by sub-critical plant.
Plant costs are considered comparable with sub-critical technology.

Much of the technology surrounding supercritical technologies is not new and


a great deal of development work was done in the 1950s and 1960s. At this
time countries such as the UK kept a predominantly sub-critical power base
due to the unreliability, expense and poor operational flexibility of these early
designs. However, elsewhere in Europe and in Japan, development and
refinement continued to the extent that supercritical steam is now considered
one of the leading clean coal technologies. Currently 10% of orders for new
coal fired power generation plant are for supercritical steam cycles and whilst
future orders are difficult to predict, estimates suggest a steady rise in the
adoption of this technology [68].
Supercritical steam cycles are not limited to coal fired plants exclusively. In
theory, supercritical steam cycles can be used for any technology
incorporating a steam cycle to generate electricity. Therefore the benefits are
considered applicable to HRSGs within combined cycle gas turbine systems.
With advances in gas turbine technology, combined cycle units are now larger
and HRSGs are operating at higher temperatures. Previously both these factors
were lacking and thus directly affected the commercial and technical viability
of the supercritical HRSG.
4.4
4.4.1

New Applications for HRSGs


The Role of HRSGs in IGCC Plant

4.4.1.1 IGCC Plant Description


Whilst gas turbine technology has been applied previously in natural gas and
oil fired combined cycle plants, the development of the Integrated Gasification
Combined Cycle (IGCC) allows both solid and liquid fuels to be the main
(88)

source of fuel for the plant. After several years of technology development and
demonstration operation, this power plant technology is approaching the status
of commercial operation.
The concept of an IGCC power plant incorporates an oxygen or air-blown
gasifier operating at high pressure and producing a raw gas, which is cleaned
of most pollutants and burned in the combustion chamber of a gas turbine
generator set for power generation. The sensible heat of the raw gas
production process along with the hot exhaust gas from its combustion in the
gas turbine are used to produce steam. The steam, in turn, is then utilised to
generate additional electrical power through a series of steam turbines. The
main subsystems of an IGCC power plant are therefore:

Gasification plant including feedstock preparation system


Raw gas heat recovery system
Gas purification system with sulphur recovery
Air separation unit (ASU); required only for oxygen-blown gasification
Gas turbine with electrical generator
Heat recovery steam generator (HRSG)
Steam turbine system with electrical generator

The actual coal gasification process for IGCC power generation is classified
into three categories namely, stationary bed (Lurgi and British Gas Lurgi
systems), fluidised bed (High Temperature Winkler, U-gas and Kellog-RustWestinghouse systems) and entrained-flow bed.
The entrained flow bed gasification process uses an oxygen blown gasifier and
is the most proven technology for single unit with large capacity applications.
The entrained flow bed gasifier has essentially five distinctive types according
to manufacturer (Texaco, Destec, Prenflo, Shell and GazCobinat Schwarze
Pumpe). Of the five, two distinct categories are apparent:

Wet feed processes such as Texaco and Destec utilise a coal slurry feed.
Dry feed process such as Prenflo, Shell and GazCobinat Schwarze Pumpe
(GSP) utilise a dry powder feed.

Generally the temperatures within the gasifiers are lower in the case of the wet
feed process than the dry feed process. Water-cooled walls, rather than
firebrick are therefore necessary. Manufacturers claim that a dry powder feed
gasifier has a slight advantage in terms of cycle efficiency over the coal slurry
gasifier. Dry processes systems are however, considered to be more
complicated than their counterpart. These complications are generally
associated with reliability penalties. On the whole the unit investment for the
dry feed gasifier is considered greater than the wet feed unit.
A schematic showing a typical IGCC utility plant layout is shown in Figure
33. This figure illustrates the plant arrangement based on a Shell dry feed
entrained-flow gasification process. The cycle is described in detail below:-

(89)

Initially, the coal is pulverised in a roll mill and then conveyed to a dryer.
Within the dryer the coal is flash dried in a stream of hot nitrogen which has
been supplied from an ASU and heated by low-pressure steam. The dried coal
is separated in cyclones and the nitrogen cooled and the water removed. This
process reduces the moisture content of the coal considerably. The coal is then
pressurised with additional nitrogen from the ASU and fed into the dry-feed,
entrained flow, slagging gasifier through lock hoppers. In addition to the
nitrogen, the ASU also provides a steady supply of oxygen into the
gasification chamber.
The gasification pressure vessel is protected from the hot gasification products
by a tube wall construction in which intermediate pressure (IP) steam is raised.
The gasifier operates at a pressure of 25bara and a temperature of 1400C and
produces a raw fuel gas, mainly composed of carbon monoxide (CO) and
hydrogen (H2). Most of the coal ash forms a molten slag, which falls into a
water bath at the bottom of the gasification chamber. Sensible heat is
recovered from the raw gas in a waste heat boiler situated at the top of the
gasifier. This boiler evaporates water bled from the high pressure (HP) circuit.
In addition, further heat is also recovered from a syngas cooler after exiting
the gasifier. To ensure that the fly ash is solid prior to entering the syngas
cooler, cooled raw gas is recycled from the outlet of the syngas cooler.
The syngas cooler heat exchanger consists of economiser and evaporator
surfaces and generates IP steam supplied directly from the IP pump. The
remaining heat in the raw fuel gas is exchanged in a gas-to-gas heat
exchanger. This exchanger utilises the raw fuel gas to re-heat the cleaned fuel
gas after an acid gas removal process. The gas cleaning process itself is done
in stages. Initially, the fly ash is removed by cyclones and by water scrubbers,
which also absorb any hydrogen chloride (HCl) present. Heat is extracted for
boiler feedwater heating and the cooled raw gas then passed to the purification
stage where sulphur-bearing compounds, mainly hydrogen sulphide (H2S) and
cabonyl sulphide (COS), are removed in order to protect the gas turbine and
also to meet environmental legislation. These compounds are absorbed by
counter-current washing with Purisol solvent and are recovered from this
solvent in a series of flash columns. The solvent is recycled and the sulphurbearing compounds are sent to the sulphur recovery plant. This plant is based
on a Claus design. Unreacted gases are treated in a SCOT tails gas recovery
unit. Heat generated in the Claus plant is used for boiler feed water heating.
The clean fuel gas is saturated with hot water in a humidifier, which helps to
reduce the NOx formation in the gas turbine combustor. Prior to entry to the
combustor, the fuel gas is further heated by an exchanger using a bleed from
the high-pressure, high-temperature (HP/HT) economiser. At the combustor,
the clean fuel gas is mixed with air supplied directly from the GE 9FA turbine
compressor alongside compressed nitrogen, the original nitrogen source being
air from the gas turbine compressor which has been separated in the ASU.
From the gas turbine, the hot flue gas passes to the HRSG where steam is
raised at two pressure levels (NB the IP stream is only superheated within the

(90)

HRSG, the steam itself is originally raised within the gasifier). The cooled
gases are then exhausted via the stack to atmosphere.
The condensate from the low pressure LP steam turbine is passed through a
condensate preheater prior to entry to the deaerator. Here the incoming water
is heated by direct contact with steam. From the deaerator, there are three
supply lines to the HRSG, namely the HP, IP and LP lines.
The LP pump supplies feedwater directly to an LP evaporator within the
HRSG. Following evaporation, some saturated steam is extracted after the LP
evaporator and used in the dryer to remove moisture from the incoming
pulverised fuel (PF). The remaining saturated steam from the evaporator feeds
into the LP superheater. After being superheated, the LP steam is split into two
lines. One line supplies the superheated LP steam to the LP turbine another
recirculates the LP superheated steam back to the deareator.
The IP pump supplies feed water to the syngas cooler and gasifier membrane
wall where the heat from the gasification process is utilised to generate IP
steam as previously described. Some steam is also bled off and fed to the
gasifier itself as part of the gasification process. The IP steam from the gasifier
is then fed into the HRSG for superheating. Following superheating, the IP
steam is added to the exit line from the HP turbine. These lines combine and
are reheated in the HRSG before entering the IP turbine.
The HP pump supplies feedwater to the high-pressure, low-temperature
(HP/LT) economiser and then on to the HP/HT economiser. Upon leaving the
HP/HT economiser, feedwater is extracted to supply the waste heat boiler
within the gasifier (see above) as well as supplying a source of heat for the
clean fuel gas line. Saturated steam is therefore produced from both the HP
evaporator within the HRSG and the waste heat boiler within the gasifier. The
two steam lines then recombine to be superheated within the HRSG. The
superheated HP steam is then fed to the HP turbine for power generation.
The IP turbine steam supply consists of the HP turbine exit flow combined
with the HRSG IP line. The LP turbine steam supply consists of the IP turbine
exit flow combined with the HRSG LP line.

(91)

Figure 33: Diagram of a typical IGCC plant with dry feed gasifier (Courtesy of Mitsui Babcock Energy Ltd)

(92)

4.4.1.2 IGCC Plant Performance


As can be seen from Figure 33 and the description above, in order to enhance
the plant efficiency, the steam cycle of the HRSG in an IGCC plant is very
much integrated with other plant components such as the gasifier. For such
plant, in general, overall cycle efficiencies of around 43% can be achieved.
The efficiency of an IGCC plant is however still essentially lower than that of
a typical gas fired combined cycle plant. In addition to the loss of chemical
energy from the removal of sulphur and other combustible contaminants, the
hot gas leaving the gasifier must be cooled in order to allow effective chemical
and ash removal. Therefore the combination from both gasification and the gas
cooling are responsible for the lower overall cycle efficiency [69].
Since the 1950s there have been 24 IGCC plants constructed or planned for
construction throughout the world. These are based on several different
variations of the gasification process. Of these 24, some 3 are dismantled, 17
are in operation and the remaining are either at the planning, engineering or
construction stage.
In recent years the numbers of large Utility IGCC plants have been growing
and in both the USA and Europe IGCC plant have reached the
commercialisation stage. In Europe three large scale IGCC (>250MWe) plants
based on combining state-of-the-art gasifier technology and a high degree of
HRSG process steam integration have been constructed and successfully
operated during the last few years. Two of these, Buggenum (Netherlands) and
Puertollano (Spain) employ coal as the main fuel source whilst Priola Gargallo
(Italy) utilises refinery residues. In both coal based plants dry feed entrainedflow gasifiers were selected and in the case of the refinery residues-based
plant the wet feed entrained-flow gasifier process was chosen.
In the USA there are currently two IGCC units generating electricity
commercially - the United States Tampa Electric unit at Polk Power (250
MWe) station and the Cinergy owned (260 MWe) plant at Wabash River.
Characteristics of these key plants [69] are outlined in Table 6 alongside those
of other IGCC plant world-wide.

(94)

Plant
Steag
Steag
Coolwater
Plaquemine
Demkolec
Tampa,Polk
Eldorado
Wabash
SchwarzePumpe
Pernis

Country
F.R. Germany
F.R. Germany
USA, Cal.
USA, La
Netherlands
USA, Florida
USA, Kansas
USA, Indiana
F.R. Germany

Start
1952
1959
1984
1986
1993
1996
1996
1996
1996

Fuel
Coal
Coal
Coal
Coal
Coal
Coal
Coke
Coal
Coal/Oil

Process
Lurgi
Lurgi
Texaco
Destec
Shell
Texaco
Texaco
Destec
Shell

MWe
50
150
100
220
250
250
40
262
40

Status
Dismantled
Dismantled
Dismantled
Operating
Operating
Operating
Operating
Operating
Operating

Netherlands

1997

Shell

127

Operating

Pinon Pines
Puertallano
I.S.A.B
Saras
Star
A.P.I.
Fife Power
I.B.I.L
G.S.K
Fife Power
Zuv
S.P.C.C.
K.P.E

USA, Nevada
Spain
Italy
Italy
USA, Del.
Italy
UK, Scotland
India
Japan
UK, Scotland
Czec Rep.
China, Yantai
USA,
Kentucky
USA, Ohio

1998
1998
1998
1999
1999
1999
1999
2000
2000
2000
2000
2003/4
2003/4

Heavy
Oil
Coal
Coal
Asphalt
Tar
Coke
Tar
Coal
Lignite
Tar
Coal/Rdf
Coal
-

KRW
Prenflo
Texaco
Texaco
Texaco
Texaco
BGL
Tampella
Texaco
BGL
HTW
-

80
300
540
550
240
250
120
60
540
400
400
2x400
-

Operating
Operating
Operating
Operating
Operating
Operating
Operating
Operating
Operating
Operating
Planning
Planning
Planning

2003/4

580

Planning

Global
Energy

Table 6: IGCC plants world-wide.


Two main disadvantages which are usually associated with IGCC are the
reliability / availability of these combined cycles and the initial significant
capital outlay [70]. Whilst the reliability/availability factor is believed to be
improving as is illustrated from the efficient running of the Pernis plant in the
Netherlands [71], the cost of an IGCC plant still remains relatively higher than a
PF plant with a flue gas desulphurisation system installed.
However, the main benefit of an IGCC plant is its ability to allow coal to be
fired in a clean and efficient manner. The removal of contaminants during the
gas clean up results in a process which is potentially the cleanest type of coalfired power plant in operation. Whilst coal remains the largest unused source
of fossil fuel in the world, it makes environmental sense to develop
technologies that allow it to compete with its naturally cleaner counterparts
and therefore reduce the rate of consumption of premium liquid and gaseous
fuels.
4.4.2

Biomass Integrated Gasification Combined Cycle

Biomass IGCC applications tend to be sized at the industrial rather than the
utility scale due to the logistics of fuel supply. They are unlikely to reach
utility scale, even once the technology is mature, due to the low energy density
(95)

of biomass fuels and the land area and fuel supply infrastructure required to
support even a relatively small plant.
There are a variety of approaches to biomass gasification including fixed bed,
rotary kiln, pressurised circulating fluidised bed (CFB), atmospheric CFB and
the Batelle process [72]. The air blown CFB, fixed bed and rotary kiln
approaches produce a low calorific value (typically 5-6MJNm-3) syngas
while the Batelle process by excluding air and using steam as the gasifying
medium produces a medium calorific value gas (~15 MJNm-3). The syngas
from all of these processes can be burnt in a GT or a reciprocating engine with
exhaust heat recovery. If syngas is to be burnt in an engine, it must be cleaned
first to remove particulates and in some cases tars as well. Current gas clean
up technology requires that the gas be cooled from the gasification
temperature (typically 850 - 950C) before filtration. There are therefore two
HRSGs in the system, one to recover heat from the hot process gas prior to
filtering and the second to recover heat from the engine exhaust. The syngas
cooler HRSG will be exposed to carry over of bed material and ash. Biomass
fuels may be rich in alkali metals which increase the potential for tube fouling.
This needs to be taken into account in the design.
Very few BIG-CC projects have yet been successfully developed. If BIG-CC
technology can be made to operate reliably and economically it will open up a
new market for industrial scale HRSGs.
4.4.3

Microturbines

In general, microturbines are unlikely to be coupled with HRSGs. The


relatively low exhaust temperature (if recuperated as most are) and flow are
not normally sufficient for economic steam generation at useful steam
conditions. However, microturbine suppliers are working on scaling up their
units. Bowman Power Systems expect to release a 200kWe unit soon and
believe that microturbines up to 500kWe are feasible. At the larger sizes they
are more likely to be coupled to HRSGs to provide steam flows for smaller
consumers in some specialist applications. It is possible that they could be
used for air pre-heating or provide heat for an LP circuit as part of a larger
boiler system.
4.5

Conclusions

Future increases in HRSG operating conditions will largely be dictated by


increases in GT exhaust temperature.

One area of significant interest is once through design. The main benefit of
this technology at present is its suitability for flexible operation. In the
long-term future, it should pave the way for supercritical cycles with even
higher thermal efficiencies.

Another area of significant interest is the use of HRSG steam for GT blade
cooling in the latest class of GTs. This presents significant challenges for
HRSG design in attaining the high steam qualities required.

(96)

The use of HRSGs within IGCC plant is now approaching the status of
commercial operation, although the costs still remain relatively high.

Other development areas include modular design to reduce build costs,


improving reliability, and improving access. The latter two items address
specific problems experienced by plant operators. Improvements in these
areas are perceived to provide product differentiation in an extremely
competitive market place.

Industrial scale HRSG technology is relatively mature. Most development


comes from the integration of HRSGs within new processes, and the
trickling down of technology from utility scale HRSGs. One exception is
the use of once through technology which is already standard practice for
one HRSG supplier.

(97)

5
5.1

WORLD-WIDE ACTIVITIES
Introduction

This chapter reviews the current trends in the Global HRSG market. The main
sources of HRSG supply and the countries responsible for purchasing HRSGs
are highlighted, alongside capabilities of the key players in the Global HRSG
market.
Two approaches were taken to finding information for this section of the
report. An internet search was carried out to try to identify as many companies
as possible that are active in the field. Appendix A shows a list of the
companies identified and a summary of their capabilities. A brief
questionnaire was sent to each of these, but the response to this survey was
disappointing, with only thirteen complete responses received. The second
approach was to examine published data. McCoy Power Reports [73] provides a
suitable source, but is focussed on the utility sector of the market.
5.2

Survey Responses

The capabilities of the companies that responded are summarised in Table 7


below.

(98)

Turnover
Range (M $)
Application

Scale

Capabilities

Technologies

>100
Utility Combined Cycle
BIG-CC
Petrochemicals
Other chemical / process
Iron, steel & coke
Furnaces / Kilns
Waste incineration
Industrial GT exhaust
Diesel engine
Gas engine
1-5
5-10
10-20
20-50
50-100
>100
Consult
Design
Manufacture
Commission
Operate
Smoke tube
Water tube
OTSG

1-5

10-50

50-100

X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X

X
X
X

>100

50-100

1-5

X
X

X
X
X
X
X
X

X
X
X
X
X
X
X

X
X
X
X
X
X
X
X

X
X
X
X
X
X

X
X
X
X
X
X
X
X
X

X
X

X
X

X
X
X

>100

X
X
X
X

X
X
X
X
X

X
X
X

X
X
X

X
X
X

X
X
X

X
X

X
X

TEI Greens

Vogt-NEM
1-5

X
X
X
X
X
X
X
X

50-100

1-10

X
X
X
X
X
X

X
X
X
X
X
X
X

X
X
X

X
X
X
X
X
X

X
X
X

X
X
X

Table 7: Summary of capabilities of companies responding to survey

(99)

Wellmann Robey

Thermax

TBW

>100

X
X
X
X

X
X
X

SFL

Nooter/Eriksen

10-50

X
X

X
X
X
X

NEM

Mitsui Babcock

0.5-1

X
X

X
X
X
X
X
X
X

M E Engineering

Innovative Steam
Technologies

Erie

Alstom

Aprovis

Company

X
X

X
X

5.3

Utility Scale Market Published Information

5.3.1

Source of Market Information

McCoys data was gathered through a written survey of US and non-US


suppliers of electric power generating equipment and engineering services. As
data from each power plant are cross-checked from various sources, it is
believed that McCoy provides the most complete and accurate information
from an independent organisation. The calculations are based on the steam
turbine MWe output in combined cycle applications and half the gas turbine
output in cogeneration, (non combined cycle) projects. The final figure
generated for total installed electrical capacity is therefore somewhat
conservative by this method, however, the data provides an excellent
indication of trends within the market (Section 5.3.2) and major market
players (Section 5.3.3).
5.3.2

The HRSG Buyers

In terms of geographical distribution, it is apparent that over the past ten years
(1992-2001), the biggest buyers by far of HRSGs have been in the US. A
massive 48% of all world-wide purchases have been made by operators in the
US. Next to the US the United Kingdom and Japan are the joint closest in
terms of purchases over this period, however, at just 4% of the total ten year
sales the sheer size and domination of the US market is obviously apparent.
In terms of customer type three categories emerge. These are classed as:

Non-Utility Generators (NUGs)


Electricity Utility Power Generators (EUPGs)
Industrial Power Generators (IPGs).

The NUGs account for some 81% of orders over the past ten years in the US
market and some 49% of orders for the rest of the world. Over the same ten
year period, EUPGs account for some 17% of the orders in the US alone and
some 44% of the non-US market.
5.3.3

The HRSG Manufacturers

With respect to the outlined ten year period, the key manufacturers on an
individual company basis were identified as Alstom Power (14.2%),
Nooter/Eriksen (12.6%), Deltak (9.5%), NEM (7.7%) and Aalborg Industries
(7.5%). Companies outside this top five, in the 1-7% share of the market
included Foster Wheeler Energy (6.4%), Mitsubushi Heavy Industries (4.3%),
Doosan Heavy industries (4.0%) and Mitsui Babcock Energy Limited (1%).
Below the 1% threshold, some 50 companies, partnerships or joint ventures
compete for the remaining market share.
It is worth noting that many companies are involved in various types of
commercial agreements with related companies throughout the world,
therefore if all related companies, joint projects and licensee relationships are
(100)

considered the top ranking is altered with NEM taking first place followed by
Nooter/Eriksen and then Alstom Power.
5.4

Conclusions

Over the past ten years (1992-2001), the biggest buyers by far of utility
scale HRSGs have been in the USA, with 48% of all world-wide
purchases. Next are the United Kingdom and Japan with 4% of the total
sales.

Key manufacturers in the above period were Alstom Power (14.2%),


Nooter/Eriksen (12.6%), Deltak (9.5%), NEM (7.7%) and Aalborg
Industries (7.5%). Companies outside this top five in the 1-7% share of the
market included Foster Wheeler Energy (6.4%), Mitsubishi Heavy
Industries (4.3%), Doosan Heavy industries (4.0%) and Mitsui Babcock
Energy Limited (1%).

For industrial scale HRSGs, there were around 33% of sales in each of the
USA and Europe, with the other leading market being Asia and Australasia
(excluding China) with 19%.

(101)

6
6.1

MARKET POTENTIAL
Introduction

This chapter discusses the potential of various world-wide markets. The views
expressed are derived from internal consultations amongst the project partners
and consultation with other organisations [74]. The external consultation
included the completion of a questionnaire by a number of companies, which
included information about their geographical market breakdown by value.
The chapter focuses on the US and China as two main areas with significant
potential for development within the global market for utility HRSGs, and on
the home market in the UK. Non- technical barriers to future success within
these markets are identified and discussed.
6.2

Market Survey

The results of the questionnaire survey have been used to give an idea of
which markets are most active. The small response to the survey means that
confidence in the results is low.
The geographical market share averaged over the thirteen questionnaire
responses received are shown in Figure 34 below:-

(102)

Figure 34: Average percentage of business by geographical market.


Looking at the industrial scale sector alone, a breakdown of 68 enquiries
received by M E Engineering over the last 18 months is shown in Figure 35
below:-

Figure 35: Percentage of enquiries coming from geographical market


(Courtesy of ME Engineering Ltd).

(103)

10 respondents provided a breakdown of the proportion of the HRSGs they


supply (by value) in each of 6 size categories. The averaged information is
shown below:-

Figure 36: Average percentage of business by HRSG size.


This simply shows that there is activity at all scales. It is based on a simple
average of the percentage of units (by value) that each company supplies.
Since the companies that supply the larger units also have far larger turnover,
in value terms the market is dominated by the large units.
6.3

Market Perception amongst Consultees

Consultees in the US suggest that the market for utility scale HRSGs is very
poor currently. Prior to the last year or so, the market in the US was buoyant.
HRSG companies had expanded to meet a demand for new utility scale power
plant. However the market is now largely saturated and there is generation
over-capacity. Transmission and distribution networks are close to full
capacity and finance for merchant plant cannot be obtained in the current
economic climate. Siemens [10] expect the US HRSG market to slump
considerably over the next few years. Consolidation amongst HRSG
companies in the US is expected. At the industrial scale there is more activity.
There are opportunities for the development of CHP schemes on industrial
sites, largely driven by security of price and supply issues in the volatile
deregulated electricity market. Concern over climate change is not yet
perceived as a significant influence on policy or the market in the US.
However, the Clean Air Act is having an influence at the industrial scale.
Consents are specifying lower NOx emission levels and rather than retrofitting
low NO x boilers and selective catalytic reduction (SCR) some companies are
opting to switch to a completely new CHP scheme.
One consultee identified Russia as a good current market due to the need to
replace ageing and inefficient plant. The same is true of Central and Eastern
(104)

European (CEE) countries. The market in CEE countries is likely to be


enhanced as the EU enlarges as trade with them will become easier and they
will be striving to meet EU environmental standards. Turkey was identified as
a good market due to low gas prices and a large demand for electricity. The
Middle East was identified as a market with good potential due to the
abundance of open cycle gas turbine plant that could be retrofitted with
HRSGs for improved efficiency by operating in combined cycle mode or for
power augmentation by turbine inlet chilling. One consultee identified Italy
and Spain as good markets for oil replacement plant and the Middle East for
desalination plant.
The market in China is expanding rapidly, but is viewed as a difficult place to
do business. This is due to the bureaucracy of complying with the local codes
and standards and the fierce competition from local manufacturers. The
market in Indonesia is also apparently growing, but competition from Chinese
manufacturers is stiff here too.
A number of other markets were identified by consultees as still being active,
despite the general depression of the CHP market throughout Europe
associated with falling electricity prices and rising gas prices. Within Europe
the markets in France, Spain, Italy, Germany and Scandinavia are perceived as
being most active for the development of CHP projects. German and
Scandinavian markets are seen as being more highly regulated still (less
competitive pressure on wholesale electricity prices) and there is price support
for CHP schemes in Germany for a limited period. Some new CHP schemes in
Germany benefit from a guaranteed feed in price. France, Spain and Italy also
have support mechanisms in place for CHP.
6.4

UK Market

The UK electricity generator market has essentially two sub-sectors, utility


CCGT / CHP and industrial CHP as outlined in Table 8.
DISTRIBUTION BY MWe
Utility
CCGT/CHP
Industrial
CHP

+1 MWe

+10 MWe

+40 MWe
60

200

80

20

+500 MWe
20

Table 8: UK power generation market sectors.


Utility CCGT/CHP are required to compete with conventional plant under the
New Electricity Trading Arrangements (NETA). Industrial CHP plants are
struggling under NETA trading conditions and most are currently operating
their steam and power supply contracts at a loss. There is therefore a degree of
turmoil within the market with attrition expected amongst some of the players.
However despite competition problems both sectors cannot ignore the need to
consider widespread integrity and performance improvements to meet NETA
market and client contractual demands.
(105)

6.4.1

Non-Technical Barriers in the UK Utility HRSG Market

In terms of sustaining a profitable business based on the internal UK utility


HRSG market certain barriers currently exist:
6.4.1.1 Current Surplus of Generating Capacity in the UK
The electricity market in the UK is currently oversupplied, with the Seven
Year Statement published by the National Grid Company in 2002 [75] stating
that the capacity available for the 2002/3 winter is 67564MW. This capacity is
made up as shown in Figure 37 below, with CCGTs contributing
approximately 32.5% to the overall generation mix. Electricity demand in
extreme winter weather conditions is expected to reach 55306MW, giving a
surplus plant margin of 22.2%. However, in normal weather conditions, the
peak demand is projected to be 52500MW (this was the peak demand during
the winter of 2001/2), resulting in an even higher plant margin. This is
attributed to the fact that governmental responsibilities for sufficient and
reliable power generation in the past have led to capacity above actual need.

Figure 37: UK generation capacity available for the 2002/3 winter.


Although a reasonable margin on plant capacity is obviously a necessity, the
UK remains oversupplied and, as a consequence, energy prices are very low
and are likely to remain so for the foreseeable future. This means that although
CCGTs are relatively cheap to build and operate (see Sections 3.8.1 and
3.8.2), they remain economically unviable under current market conditions.
This, coupled with the high price of natural gas (see below) not only makes
the building of new, utility-scale HRSGs unlikely, but has also resulted in the
recent mothballing of some UK CCGT plant.

(106)

6.4.1.2 Fluctuations in the Price of Natural Gas


Recently investors have been dissuaded from investing in CCGT plant by the
volatility in the price of natural gas. Price volatility results in difficulties in
reliably forecasting the overall costs occurred over the entire operating life of
the combined cycle plant.
6.4.1.3 Current Unpredictability of the UK Retrofit Market
With little in the way of large new-build utility plants in the UK, the market
for upgrading existing facilities within the country must be the primary focus.
The effects of flexible operation are becoming more and more apparent to
operators, so the potential for upgrade opportunities is large. However with
flexibility upgrades on drainage, pressure parts and casings expected to be
between 100k and 400k per HRSG, the depth of the market remains
uncertain. Factors within the operators market as a direct result of deregulation
suggest that the availability of finances to fund these upgrades is questionable.
6.4.2

UK Industrial CHP Market

The annual Directory of UK Energy Statistics (DUKES) for 2001 gives


various data for CHP schemes in the UK. The data is gathered through the
Governments CHP quality assurance scheme. The majority of schemes are
fuelled with natural gas (61%), with fuel oil accounting for 7%, 2% from
renewables and the balance from various process exhausts or by-product fuels.
The majority of the 1573 schemes are small, but generating capacity is
dominated by the minority of larger schemes.
Electrical capacity size
range
< 100 kWe
100 kWe 999 kWe
1 MWe 9.9 MWe
> 10 MWe

% of total number of
schemes
43.2
40.1
12.1
4.6

% of total generating
capacity
0.9
3.2
16.2
79.7

Table 9: CHP Schemes by size, 2001.


Figure 38 below shows the installed CHP capacity in the UK for the last 5
years.

(107)

Figure 38: Installed CHP capacity in the UK 1997-2001.


The chart shows that through the 1990s the pace of addition of CHP capacity
was accelerating with the rate peaking in 2000 with 844MW of plant being
commissioned, a 22% increase on the previous year. However in 2001 only 38
MW of capacity was added. As part of this study the Combined Heat and
Power Association and a number of companies involved in the supply and
installation of CHP schemes were consulted. All of these reinforced the view
of the market given by the DUKES statistics. The market for CHP schemes
was buoyant towards the end of the 1990s but then collapsed. The reasons for
the current depressed state of the market are discussed below. The trend is
mirrored in the figures for electrical export from CHP schemes. In 2000 8482
GWh were exported. In 2001 the figure had dropped to 5960 GWh. Similar
problems with the CHP market apply elsewhere in the EU, such as in the
Netherlands and Germany [76].
6.4.2.1 Stricter Consents Policy
In 1998 the Government introduced a stricter consents policy on gas fuelled
electricity generation plant. Under this policy, consent for the development of
any gas fired power plant of generating capacity greater than 10MW was
required from the Government. The measure was introduced because of
concern that the electricity market was being distorted and coal fired plant was
being unfairly penalised by this. The policy was kept in force while a review
of electricity arrangements was carried out. It ended in 2000 with the
introduction of the New Electricity Trading Arrangements (NETA). The only
plants granted consents during this period were CHP plants. The stricter

(108)

consents policy had the effect of concentrating attention on the development


of CHP projects.
6.4.2.2 Natural Gas Prices
Most CHP plant is fuelled with natural gas. The interconnection of the UK gas
grid with that of continental Europe via the Bacton-Zeebrugge pipeline was
accomplished in October 1998. This has exposed the UK to the higher prices
of the continental European market and UK natural gas prices have
consequently risen from a low of under 14p/therm in 1999 to 19p/therm in
2001 [77].
6.4.2.3 New Electricity Trading Arrangements
The New Electricity Trading Arrangements (NETA) were brought into force
in March 2001. NETA introduced new regulations governing the deregulated
electricity market in the UK. Previously the pool price at any one time was
paid to all generators supplying power at that time and the price was set by the
most expensive generator. Under NETA, most wholesale electricity sales
between generators and retailers are now concluded well in advance with the
timing, volume and price of supply fixed. A balancing mechanism provides a
means of meeting short term demand fluctuations. Fierce competition since
the liberalisation of the market has driven down the prices at which contracts
have been concluded. Since 1998, when NETA was first proposed, electricity
wholesale prices have fallen by 40%. Between April 2001 and February 2002
baseload prices have fallen by 20% and peak prices by 27% [78].
What is more, CHP schemes are unlikely to get the best prices. CHP schemes
are generally small compared to utility scale plant so have little market
influence. Penalty payments are levied if a generator fails to meet its
scheduled supply profile. Many grid connected CHP schemes tend to run to
meet the heat or steam demands of its customer, rather than an electricity
supply profile, and the electrical output is therefore not entirely predictable.
The CHP scheme is exposed to a potential mismatch between the heat and
electrical demand profiles that it is trying to meet. This results either in NETA
penalties or running at lower efficiency than the design point.
6.4.3

Other Barriers for UK Firms

The grid has evolved to distribute electricity from a small number of large
generators. A substantial amount of power is wasted in the form of
transmission losses. In 1999 it is estimated that transmission and distribution
losses accounted for 1336 TWh, equivalent to 11.6% of the Worlds final
electricity demand [79]. CHP plants are generally connected to local grid
networks and the separation between generator and consumer is smaller,
resulting in lower transmission and distribution losses. Decentralised or
embedded generation can also have benefits in strengthening the local grid and
improving power quality. These benefits of embedded generation are currently
not fully reflected in the sale price of electricity from CHP schemes. However
the connection of a large number of smaller generators may also cause
difficulties for the grid system relating to fault levels, islanding and power
(109)

quality. Where substantial modifications are required to the grid to allow


connection, the cost to the scheme developer may be prohibitive, especially
for smaller schemes. The cost and complexity of grid connection is currently a
barrier to the development of smaller CHP schemes.
Anecdotal evidence suggests that banks are not willing to invest in energy
projects at present due to the risk associated with price volatility and
uncertainties within energy markets. The market is in such a poor state that the
major CHP developers have disbanded their development teams. The only
projects that may go ahead currently are either ones where all of the power
will be consumed on site or that have another factor driving them, such as
avoided grid connection strengthening costs.
One UK consultee identified the diversity of European standards as being a
problem for companies trying to export to other EU countries. A common
European standard for shell boiler design is being introduced (EN12953), but
the consultee was concerned that oversees clients will continue to specify their
own national standards.
The current strength of sterling was also identified as a problem for UK HRSG
suppliers. There is strong competition in the market place and other European
suppliers can undercut UK companies, even in the UK despite their higher
transport costs.
6.4.4

Future Industrial CHP Market Potential in the UK and Mainland


Europe

The European Commission sponsored Future Cogen study assessed the


potential for the expansion of CHP within EU member states and Central and
Eastern European (CEE) states. It modelled the growth of CHP under four
scenarios ranging from the pessimistic deregulated liberalisation scenario to
the optimistic post Kyoto scenario. In the deregulated liberalisation
scenario EU CHP capacity grew by only 16 GW from a base level of 65GW to
81GW in 2020. In contrast, under the Post Kyoto scenario installed capacity
grew by 130GW to 195GW by 2020. Under the Post Kyoto scenario, CHP in
the UK would grow from a base level of 3453GW to 27215GW in 2020. Not
all of this growth would come from CHP schemes involving steam generation
but it would represent a substantial opportunity for growth in the HRSG
industry[80].
At present CHP opportunities in the UK are limited to those that have specific
driving factors other than just more efficient use of fuel. The UK government
has set a target of achieving an installed CHP capacity of 10GW by 2010,
compared to the current capacity of 4801MW (2001). Some policy measures
have been introduced to stimulate the CHP market. Fuels used in CHP are
exempted from the Climate Change Levy (CCL). In the April 2002 budget it
was announced that electricity exported from CHP schemes will also be
exempted from the CCL (subject to approval under EU state aid rules).
Enhanced capital allowances (ECAs) are allowed on some items of CHP

(110)

equipment, increasing the opportunities for investment in CHP. However


ECAs are not yet available on all items in a complete installation.
The EU target is to double CHP capacity as a fraction of total electricity
generation capacity from 9% (1994) to 18% in 2010. In order to achieve this
target the EU have proposed a CHP directive[81]. The key points of the
directive are:

The introduction of a common definition of cogeneration;


Targeting of support at schemes with an electrical capacity of up to 50MW
(or at the first 50MW of larger schemes);
Providing a guarantee of the origin of electricity from cogeneration;
The establishment of efficiency criteria for cogeneration;
An obligation on member states to establish their national potential for
cogeneration;
Allowing national support schemes for cogeneration in the short to
medium term (under state aid rules);
The establishment of objective, transparent and non-discriminatory rules
for grid connection and reinforcement;
A requirement for member states to review legislative frameworks with a
view to reducing barriers to cogeneration.

These actions should enhance the European HRSG market.


The Kyoto agreement sets binding targets for greenhouse gas emission cuts for
signatory countries. CHP has been identified as one of the most cost-effective
methods of cutting CO2 emissions. Joint Implementation and Emissions
Trading mechanisms can potentially be harnessed to help develop CHP
schemes. In Annex II countries, the Clean Development Mechanism may
offer opportunities to help develop CHP schemes.
6.4.5

Action to Stimulate the UK Market / Support the UK Industry

A common view was expressed by all UK consultees: there is no problem with


the product but huge problems with the market. No need for DTI funded
research was identified the technology is essentially mature. However the
Government does need to act to stimulate the UK market. The combination of
NETA and a high natural gas price has dramatically reduced the market for
CHP and the climate change levy is not seen as being an adequate incentive to
invest in new HRSGs / CHP schemes. Enhanced support for CHP is required
from government and a CHP obligation seems to be the preferred
mechanism in the industry.
The current pessimistic industry view is supported by a report by Forum for
the Future / Cambridge Econometrics commissioned by the Combined Heat
and Power Association and released in October 2002. This predicted that total
installed capacity would only reach 6.6 GWe by 2010 and 8.6 GWe by 2020
under current conditions, compared to the government target of 10 GWe by
2010.

(111)

The Combined Heat and Power Association has therefore identified the
following steps as means of improving the situation for the development of
CHP projects:

6.5

Introduction of a CHP obligation, similar in form to the Renewables


Obligation but with a lower buy-out price, to provide an underpinning
market mechanism
A proactive planning and communication strategy
Adequate resourcing within government
Full implementation of existing support mechanisms for CHP
Action to address existing and emerging barriers to CHP
Delivery of effective support to emerging technologies
North American Market

The North American market at present is considered by some to be expanding


significantly, in contrast to the opinions expressed by some consultees (see
Section 6.3 above). In the year 2001 according to Power Magazine [61] an
installation record of 48.6 GWe in total was established for new gas turbine
based power plants. This figure surpasses the 1974 record of 43 GWe for new
installations. However, the 2001 record is envisaged as being short lived as a
result of the 66 GWe for 2002 and 69 GWe for 2003 announced or under
construction [61]. Whilst mid-sized aero-derivative gas turbines are part of the
sales surge, the vast majority of this increased capacity has been, and no doubt
will continue to be in large, utility scale machines.
When it is considered that just 27 GWe of new capacity came on line in the
year 2000, the extent of the rapid growth within the North American market is
clearly apparent.
Whilst the size of the North American market is clearly large, there are
obvious signs of the high level of competition existing in this potentially
lucrative business sector. Evidence of this rivalry can be found from the
undisputed fact that in recent years the market has attracted the attention of
several new competitors, all contending for their share of the profits.
Furthermore this additional competition has emerged from both international
manufacturers winning American based contracts and American based
companies with previously limited interests in the HRSG market suddenly
expanding their involvement.
From what was essentially a US supplier only market, there has been a recent
spate of key contracts awarded to non-US manufactures such as Toshiba
(Tokyo, Japan), CMI (Brussels, Belgium) and Hitachi (Tokyo, Japan). Foster
Wheeler is a prime example of an American company expanding its activities
in the HRSG market. Prior to 1997 its HRSG interests were considered not to
be a prime business within the company. However, the company claims that
having recognised the huge potential for rapid combined cycle power plant
expansion, it strategically expanded into this area from its traditional business
of solid-fuel boiler fabrication. The end result is that in just 5 years its HRSG
business has now grown from a minor business sector to its largest one.

(112)

6.5.1

Possible Non-Technical Barriers to Development of Combined Cycle


Technology within the USA

Some analysts [82, 83] are sceptical about the future of the North American
market and predict as much as 50% of the announced future projects in the
states will not make it further than the drawing board. The main reasons cited
are:

Inadequate supplies of natural gas: Natural gas is the fuel of choice for at
least 95% of the projects. The US Energy Information Administration
estimates that natural gas production and delivery would have to rise 4050% over the next 15 to 20 years to supply the projected combined cycle
developments. This is viewed as a somewhat heroic task based on the aged
state of the nations gas fields and requirement for extensive new pipelines
for gas delivery.
Price of Gas: Market forces are also at play with the high price of gas
dissuading investors in gas turbine technology and looking towards coal,
nuclear and hydro power for their energy solutions. The use of gasification
combined cycles is still however a possible lucrative market that should
not be ignored.

6.5.2

The US CHP Market

In the US, the Assistant Secretary for Energy Efficiency and Renewable
Energy has issued a CHP challenge calling for industry and government to
work together to double the capacity of CHP in the US by 2010. According to
the US Combined Heat and Power Association (USCHPA), the current
installed capacity is around 65GW, on the way from 46GW in 1998 to the goal
of 92GW in 2010. However measures to support this target are limited. There
are federal programmes supporting awareness raising, research and
demonstration. A 10% investment tax credit for CHP was included in
proposed legislation passed by both houses of Congress in the last two years,
but it was not enacted into law due to disagreements over other provisions of
the law. In order to avoid revenue loss, the same provision would have
stretched out CHP asset depreciation for tax purposes, so it was considered
something of a neutral measure by the CHP industry. There are federal
requirements that CHP operators achieving certain levels of efficiency can
compel utilities to purchase their power at the utilities' avoided costs. This was
a strong incentive for CHP in earlier times when incremental power generation
costs were quite high, but motivates fewer projects now, according to the
USCHPA. Few states have introduced mechanisms to support CHP or
equitable rules to govern the connection of CHP schemes to the grid.
6.6

Chinese Market

In terms of future world markets the Peoples Republic of China is significant.


Currently the installed power generation capacity in China is the second
largest in the world, with the United States being in first position. However the
per capita electric power utilisation level in China is still low. China expects
its economy to grow at an average rate of 7% or more per year over the next
decade. Therefore if a constant ratio of primary energy to gross domestic
(113)

product is assumed for this period the primary energy consumption would
nearly double. Thus the electric generating capacity for the nation will be
required to increase dramatically.
China is a large coal production country with coal as its main source of power
generation. This situation is envisaged as remaining unchanged in China for
the foreseeable future. Of the ~300GWe of installed capacity at the end of
1999 fossil fuel capacity was almost 75%. Coal fired units account for more
than 95% of the fossil fuel fired capacity. China is therefore actively pursuing
Clean Coal Technologies (CCTs) as a means of meeting their future energy
requirements with integrated gasification technology (IGCC) and supercritical
boilers attractive options for the 21st Century.
China has made a decision to build a large-scale IGCC demonstration power
plant and is currently conducting preparatory research for such a project.
Yantai power plant in Shandong province has been proposed as the host site
for this demonstration for which two 400MWe IGCC units are being
considered.
Assuming that the Yantai IGCC project proceeds, it could be in commercial
operation by the end of year 2005. Wider deployment of IGCC could,
therefore, be forecast for the period beyond 2005-2010. The potential rise of
IGCC within the market place over a 15-year period is predicted as resulting
in a 17% share in the coal-powered generation market by the end of 2025.
With the HRSG being an integral component of the IGCC plant (as outlined in
Section 4.4.1) the knock on effect of HRSG sales in the Chinese market place
may well be significant. However, it is apparent that the final market size will
depend entirely on the success of the demonstration and the cost reductions
achieved.
In addition, the air blown gasification combined cycle (ABGC) has been
proposed as another possible contributor to Chinas future energy generation.
This cycle also is dependent on the use of the HRSG to enhance overall
performance. Essentially the ABGC is a hybrid combined cycle power
generation technology based on the partial gasification of coal [84]. The
combustion of the fuel-gas is undertaken within a gas turbine. The combustion
of the remaining gasifier char is carried out in a circulating fluidised bed
combustor where steam is generated to drive a steam turbine. A key feature of
the ABGC process is its potential to achieve high cycle efficiencies with low
environmental emissions.
ABGC development is estimated at being around five years behind IGCC.
However, its main attractiveness stems from the fact that it is well suited to
poor quality high sulphur coal which is in abundance in the developing
countries of China and also India.
Predictions indicate that on the basis that a working plant could be established
within the period 2005-2010, then within 15 years some 10% of the market
share of coal fired power generation could be supplied via ABGC.

(114)

Detailed studies of CCTs within the Chinese market place have been
undertaken by Mitsui Babcock [70, 84, 85] along with various Chinese research
institutions. This previous work has been undertaken with the financial
support of the DTI.
6.6.1

Possible Non-Technical Barriers to Development of Combined Cycle


Technology within China

The specific market for HRSGs within China is clearly dependent on the
Chinese Authorities adoption of a positive policy on cleaner coal
technologies. In general the Chinese market is seen as being hindered by seven
factors listed below.
6.6.2

Complex Administrative Procedures

China is in the process of government and administration reform and


enormous changes have been made in recent years. In 1998, the State Power
Corporation (SPC) replaced the Ministry of Electric Power and the
government's administrative responsibility for the power industry was
transferred to the State Economic and Trade Commission (SETC).
Traditionally, the State Development and Planning Commission (SDPC) is the
top authority responsible for approving new power plant projects. The State
Economic and Trade Commission (SETC) is the top authority responsible for
approving renovation projects. These two commissions are the most powerful
government agencies in terms of applying for and receiving approval for
cleaner coal technology projects.
The first step in influencing the SDPC and SETC is to inform them of the
technology, the history of development, the current situation, technical and
economical features, advantages and disadvantages. Providing them with
documents, inviting them to attend a workshop, or visit research facilities or
demonstration sites therefore allows this interaction to take place. Secondly, if
a project is being prepared, a feasibility study report with favourable financing
arrangements such as a soft loan or a grant from international organisations
will certainly have a positive influence on the approval process.
6.6.3

Low Institutional Capability

The lack of collaboration between design institutes, research institutes and


manufacturers acts as a key barrier to international technology transfer. Most
R&D for cleaner coal technologies requires a multidisciplinary approach. In
addition, Chinas state-owned manufacturing enterprises have not developed
commercial or innovative skills and there is a lack of market pressure on
Chinese enterprises. With the deepening of economic reform and system
restructuring, however, all state-owned enterprises and research institutes will
accelerate the process of upgrading management and technology in order to
improve competitiveness.

(115)

6.6.4

Environmental Emission Controls

With China being a developing country, the standards relating to


environmental protection are still much lower compared to those in fully
industrialised countries. The regulations on emissions from thermal power
plants, for example, are not so stringent. This situation does not put enough
pressure on industry to create a demand for cleaner coal technology hardware
and services. In addition, the implementation of these standards is sometimes,
and in some places, poor and inconsistent. The lack of enforcement and
monitoring therefore also has a negative influence on environmental
investment. Environmental protection, however, is one of Chinas basic
national policies for sustainable development. With the rapid economic
development and improvement of living conditions environmental policy is
being given a higher priority and becoming more stringent.
6.6.5

Financial Issues

Lack of finance is often an important barrier to cleaner coal technology


transfer. The following measures will enhance the possibilities for technology
transfer:
(i)
Both government and international organisations will devise more
favourable policies and offer concessional finance for the introduction of
advanced cleaner coal technologies in the form of soft loans, capital subsidies
or grants.
(ii)
Cleaner coal projects will become economic if the issue of pollution
costs is addressed. This issue is linked to the reform of the pollution levy
system.
(iii)
The cost of cleaner coal equipment manufactured in China is much
lower than the cost of imported equipment. Hence, there is a strong economic
and financial incentive to maximise the local manufacture of equipment. This
can only be realised with technology transfer.
6.6.6

Maturity of the Technology

As end users, power companies will only employ mature technologies. It is


generally deemed to be crucial that at least two reference plants of the same or
comparable size should be operating. For newly developed technologies a
demonstration project of relevant size and parameters is important.
6.6.7

Issue of Intellectual Property

Gradually, the move to commercialise state-owned industries is strengthening


respect for intellectual property rights. Furthermore, the move to a competitive
market will eventually bring about a situation in which companies in China
will have less incentive to share information with each other.
6.6.8

Long-Term Collaboration

Joint ventures between Chinese and foreign firms or involving technology


licensing agreements can potentially facilitate the transfer of the wider

(116)

knowledge, expertise and experience necessary for managing technological


change. Joint ventures in particular have one important feature which can help
collaborative relationships in China to be successful: a relationship that gives
both sides a stake in the future success of the product or service concerned,
and allows the build up of trust.
6.7

Conclusions

Whilst the utility scale HRSG market has been healthy in recent years,
there is a predicted sharp downturn in the HRSG market in the shortmedium term due to plant over capacity. The situation is not expected to
pick up again until around 2007-2011. Key future HRSG markets are seen
as the USA and China (via IGCC). Non technical barriers in these two
markets include the price/availability of gas in the USA and
administrative/financial issues in China.

For industrial scale HRSGs, the European market is depressed due to


falling electricity prices and rising gas prices. However potential markets
include Russia, Central and Eastern European (CEE) countries, Turkey
and the Middle East. In the USA, despite problems on the utility scale,
there are still opportunities for development of CHP schemes on industrial
sites, largely driven by security of price and supply issues in the volatile
deregulated electricity market.

The current surplus of generating capacity in the UK and fluctuations in


the price of natural gas have led to a low requirement to build large scale
power generation plant within the UK. The only plus side is that with the
effects of flexible operation becoming more apparent, plant performance
upgrade opportunities are present.

The combination of NETA and a high natural gas price has dramatically
reduced the UK market for CHP and the climate change levy is not seen as
being an adequate incentive to invest in new HRSGs / CHP schemes.
Enhanced government support for CHP is required if the target of 10 GWe
by 2010 is to be met.

(117)

UK ACTIVITIES

This chapter reviews the prospects of UK manufacturers in the global HRSG


market, and their capabilities. Areas of current research, development and
demonstration (RD&D) which are being undertaken in the UK are indicated
alongside recommendations of areas of significant future focus.
7.1

Prospects of UK Suppliers and Manufacturers in the Global Market

Large, new build HRSG manufacture, as with many other heavy engineering
manufacturing business within the UK, is finding it increasingly difficult to
compete with the low costs associated with both European competition based
on the continent (e.g. in Italy and Spain) and the significantly reduced outlay
incurred by manufacturing in East Asia.
With more modular HRSG designs becoming commonplace, the future of UK
HRSG companies in the new-build market lies in the ability to sell, supply and
assemble HRSGs, based on their own designs and advanced technologies,
albeit that the standard components may not necessarily have been
manufactured inside the country.
Within such an environment, licensing agreements and collaborative
partnerships are therefore deemed essential in order for companies to maintain
the ability to compete in the global market.
In order to form such alliances, UK companies must be in a position to offer
something in return. Under such conditions the requirement to be continually
developing new technologies becomes vital. Therefore to guarantee future
long-term prospects for UK suppliers and manufacturers in the global market,
its knowledge and development of leading edge technologies must be
maintained.
7.2

UK Capabilities in HRSG Design, Manufacture and Supply


Utility-Scale

In terms of UK based large utility HRSG manufacture and design, Mitsui


Babcock has a significant presence within the UK.
Mitsui Babcock is a major energy engineering company incorporated in the
UK and since 1995, a wholly owned subsidiary of Mitsui Engineering &
Shipbuilding of Japan. The company is a technology leader in large fossil fuel
steam generating plant, and specialises in the design, engineering,
manufacture, construction, commissioning and after sales servicing of high
efficiency, high availability coal, oil, and gas fired boilers for the power
stations of electricity generating companies world wide. The company is also a
major manufacturer and supplier of heat recovery steam generating plant,
industrial fluidised bed and other clean burn coal fired boilers, coal milling
plant, flue gas desulphurisation plant and low NOx technologies.

(118)

Up until the mid 1990s and operating at the time as Babcock Energy, Mitsui
Babcock was one of the worlds foremost suppliers of HRSGs. In 1993 Mitsui
Babcock were nominated as the worlds leading supplier of HRSGs and were
winners of the coveted Power Engineering International project of the year
award. To date the company has won contracts in some 23 different
countries. However by the mid 1990s the HRSG market suffered a severe
downturn with limited opportunities and reduced margins. To compound this,
the market was migrating from assisted circulation to natural circulation
designs and the companys specific technology offering became less
competitive. However with a recent upturn in the fortunes of the HRSG
market particularly in America, Mitsui Babcock have enhanced their product
range by concluding a licensing agreement with Babcock Hitachi KK for
natural circulation designs. Significant orders for utility plant HRSG supply
such as the natural circulating HRSG for Naco Nogales power plant in Mexico
and the assisted circulated HRSG employed at Blackpoint power station in
China have been completed.
Mitsui Babcock employs approximately 3000 people in its various operations
in the UK and abroad. Its headquarters are at Crawley in the UK but operates
locally with regional operations elsewhere in the UK and around the world.
The companys strengths lie in its depth of engineering capabilities, its
technology base, its extensive manufacturing facilities, its considerable
experience in site erection, commissioning and servicing of major power plant
in countries across the world and its ability and experience in managing very
large multi-disciplinary projects. The companys combination of
technological, financial and skill resources enable it to deliver projects in a
range from 10 million to 400 million, anywhere in the world.
Thermal Engineering International Ltd Greens is the largest independent
manufacturer of utility HRSGs in the UK. Originally known as E Green &
Son the Wakefield based company has amassed over 150 years of experience
in the field of heat recovery since its founder Edward Green invented and
patented the worlds first economiser which he patented in 1845.
TEI Greens has manufactured Utility HRSGs for most of the worlds leading
boiler designers/makers as many no longer support their own manufacturing
facilities. TEI Greens has been successful in manufacturing HRSGs for
domestic and export projects and has a wide experience of different designs
including vertical and horizontal gas types and once through designs.
Currently, around 40% of all the UKs utility HRSGs have been
manufactured by TEI Greens.
TEI Greens are holders of the ASME S & U stamps and have a large facility
of over 100,000m2 with extensive workshops. It has 3 x High Frequency
Finning machines in its Wakefield Factory (10 worldwide) for the
manufacture of high frequency welded helical fin tubes as used in modern
HRSGs. The facility is capable of producing over 120,000 tubes and 10 major
HRSGs per annum.

(119)

There is also a number of major utility-scale HRSG turnkey contractors


operating within the UK, although their headquarters and manufacturing
facilities are generally based overseas or manufacture is sub-contracted.
Alstom Power, Foster Wheeler Energy Ltd, Mott Macdonald, Nooter/EriksenCCT Ltd and Siemens KWU all fall within this category.
7.3

UK Capabilities in HRSG Design, Manufacture and Supply


Industrial-Scale

Wellman Robey and BIB Cochran both manufacture and supply smoke tube
(shell boiler) type HRSGs. Wellman Robey is owned by the Wellman Group
of the UK and BIB Cochran is owned by the Mechmar Corporation of
Malaysia. Wellman Robey supply units in the 5-10MW range for use in
exhaust heat recovery behind GTs; gas and diesel engines; incinerators, kilns
and furnaces; and process integrated units in petrochemical, other chemical
and iron and steel industries. It has its own manufacturing capabilities at its
factory in Oldbury, and also offers contract manufacturing services. Besides
HRSGs, Wellman Robey supplies a range of fired package boilers and offers
after sales support and maintenance. BIB Cochran similarly supplies units for
a range of exhaust heat recovery and process integrated applications. It
manufactures its products at its factory in Dumfries and Galloway and its
product range also includes a range of fired package boilers. It has
representation in many counties in eastern and western Europe, the Middle
East, south east Asia, India and the Americas. It also provides various after
sales services.
M E Engineering is owned by the Thermax group of India. It only supplies
bespoke units rather than package units. It has a range of water tube designs
for exhaust heat recovery and process integrated applications. In GT exhaust
heat applications, the range of GTs served is from 5MWe to around 70MWe.
Besides heat recovery systems the company can supply boilers for a wide
range of solid, liquid and gaseous fuels. The company does not have its own
manufacturing facilities, but uses either the factory of its parent company in
India or sub-contract manufacture.
Industrial HRSGs are designed and manufactured in house by TEI Greens.
These may be of a water tube or smoke tube design and may incorporate
supplementary or auxiliary firing where required.
The UK arm of Nooter/Eriksen also supplies industrial HRSGs although its
design capability is based in the US and manufacture is sub-contracted.
Details of each of the companies described above are summarised in Table 10
below.

(120)

Company

Ultimate
Parent
Company

BIB Cochran

Mechmar
Corporation
Malaysia

Wellman
Robey

Wellman
Group, UK

ME
Engineering

Thermax
Group, India

HRSG
Business
Annual
Turnover
(range)
-

HRSG Designs

Scale

Applications

Smoke tube,
Package,
Optional
supplementary
firing

Up to 35 tph steam at
up to 35 barg

CHP, GT exhaust,
reciprocating engine
exhaust, incinerator
flue gas,
petrochemical, other
process industry

$1-$5M

Smoke tube,
Package,
Optional
supplementary
firing

5-10 MW or up to
20MW if
supplementary fired.
Up to 40 tph steam at
up to 35 barg and
360C, saturated or
superheated

CHP, GT exhaust,
reciprocating engine
exhaust, incinerator
flue gas,
petrochemical, other
process industry, iron
& steel

$0.5-$1.0M

Smoke tube,
Water tube,
Optional
supplementary
firing

Up to 55 tph steam at
up to 60 barg and
450C, saturated or
superheated

CHP, CHP & cooling,


GT exhaust,
reciprocating engine
exhaust, incinerator
flue gas (clinical,
municipal),
petrochemical, other
process industry,
biomass IGCC, iron
& steel, offshore oil
production

(121)

Capabilities

Site surveys,
Design,
Manufacture in house,
Installation,
Commissioning,
Training,
Service,
Repair and maintenance
Spares supply
Design,
Manufacture in house,
Commissioning,
Training,
Operation,
Service,
Repair and maintenance,
Spares supply
Design,
Contract manufacture local
to project or at parent
company factory in India,
Installation,
Commissioning,
Training,
Repair and maintenance

Other Products

Gas burners,
Package units:
Thermax, Clansman and
Calpac hot water boilers
Wee Chieftan, Thermax
single and double furnace,
Borderer and Coalmaster
steam boilers
Pressure vessels,
Sub-contract manufacture,
Package units:
Robey low NOx boilers,
STONE steam generators,
Ygnis hot water and steam
boilers
Biomass & fossil solid fuel
boilers (travelling , dumping
and pinhole grates, fluidised
beds) up to 100 tph evaporation,
Liquid / gas fired single / bi
drum water tube boilers up to
300 tph evaporation,
Fired once through coil boilers
up to 50 tph evaporation and
200 barg
Re-tubing, air-preheaters,
economisers,
Heat recovery to water,

TEI Greens

Thermal
Engineering
International

Nooter/Eriksen
CCT Ltd

CIC Group
Inc.

$1-10M

Smoke tube,
Water tube,
Optional
supplementary
firing or auxiliary
firing

Various from industrial


to utility scale.

Smoke tube,
water tube,
optional
supplementary or
auxiliary firing.

Various from industrial


to utility scale.

Waste heat recovery


from boiler and
process flue gas
streams in power
generation (CHP &
CCGT), refining,
chemical, process and
general industries.
Power generation
(CHP & CCGT),
designs for waste
incineration and
process industry
applications.
Inclusion of catalysts
possible.

Design,
Manufacture in house,
Unit build,
Erection,
Commissioning

Design (overseas),
Manufacture (subcontracted),
Unit build, erection,
commissioning.

Table 10: Capabilities of UK industrial HRSG suppliers.

(122)

water/glycol, thermal oil,


Absorption cooling
Helical finned tube manufacture
in both solid and serrated fin
profiles.
Utility-scale HRSG
manufacture in house, unit
build, erection and
commissioning (but not design)
Optimised designs for cycling
and constructability.
Enhanced Oil Recovery
OTSGs for 80% quality steam.

7.4

Research, Development & Demonstration Activities undertaken in


the UK

The following section lists areas of significant R&D activity in the UK. Areas
in which long term future RD&D efforts should be focused are
recommended:7.4.1

HRSG Once-Through Technology

One area of significant development within the UK with respect to HRSG


design is once-through technology. This is highlighted by the fact that a sub
critical, utility-scale, once-through demonstration plant has been built by
Babcock Borsig (ex-Deutsche Babcock) at Cottam Development Centre in the
East Midlands. The advanced Siemens V94.3A gas turbine exhausts into a
HRSG of horizontal gas-flow, vertical tube, triple pressure design with reheat.
The plant is now operating at close to 58% thermal efficiency [86]. The HP and
IP circuits are once through, using Benson technology licensed from Siemens
KWU (the turnkey contractor), whilst the LP circuit is of conventional, natural
circulation design. The HP/IP systems operate at sliding pressure down to 60%
load, whilst the LP operates at fixed pressure. The HRSG was actually
designed for the larger WEGA S9 gas turbine, but this has not been installed
to date. The salient outlet conditions for both gas turbines are shown in Table
11 below:

Temperature
Pressure
Flow

C
bar(g)
kg/s

V94.3A
HP
IP/RH
545
530
133.9
29.8
73.9
85.8

LP
244
4.7
10.1

HP
580
160
87.5

WEGA S9
IP/RH
560
35
99.3

LP
246
4.9
17.2

Table 11: Steam outlet conditions of the once-through, sub-critical Benson


design HRSG at Cottam Development Centre for the existing V94.3A gas
turbine and the proposed WEGA S9 gas turbine.
Due to the relatively high steam conditions, the HRSG design makes extensive
use of P91 in the HP superheater (75% of components) and high temperature
reheater (50% of components). The relatively low heat flux within the HRSG
(compared to a PF fired boiler) allows a low mass flux, natural circulation
once-through design (i.e. without re-circulating pumps). There are two steamwater separator vessels and one water reserve vessel for the HP system and a
single combined steam-water separator/reserve vessel for the IP system.
During start up, these vessels re-circulate water back to the evaporator inlet
and can also overflow to the flash vessel and/or the LP drum (if water quality
is high enough) on start-up. If and when the WEGA gas turbine is installed,
the IP circuit will provide cooling of GT turbine bypass air via a heat
exchanger.

(123)

The HRSG pressure parts have a 200,000 hour design life. The HRSG is also
designed for flexible operation, a fatigue assessment of critical pressure parts
being carried out by Deutsche Babcock to TRD 601 at the design phase. The
HRSGs ability to operate flexibly is improved by the use of full penetration
welds on all pressure parts with all headers, tubing, pipework and steam/water
separator vessels being of seamless construction.
The technology employed in once-through boilers will act as a stepping-stone
towards supercritical HRSGs. These are predicted to become commercially
available within the next 5-15 years depending on the development of gas
turbines and the achievement of greater exhaust temperatures.
7.4.2

Effects of Cyclic Loading on HRSGs

Another immediate focus of research is the requirement to design HRSGs


suitable for flexible operation. Existing plants have been largely designed for
base loading with only limited flexible operation calculations performed
during the design stage. However with plants now operating in a cyclic
manner under NETA the future challenge is to establish which components are
most at risk and identify how this risk may be reduced. Thus engineering
analysis is being undertaken by a variety of companies, with operators
undertaking trials and inspections to establish the components most at risk [87].
7.4.3

Various Novel HRSG Designs

Currently conceptual power cycles are in development which endeavour to


radically reduce harmful emissions from utility power plant and achieve near
to zero emissions status [88]. Many of these incorporate the use of HRSGs in a
variety of ways to improve efficiencies within the novel cycle. Within the UK
research is continually underway to assess the feasibility of such designs.
For example Mitsui Babcock are currently undertaking a conceptual design
and feasibility study for an appropriate HRSG suitable for the GAS-ZEP
concept. The project involves collaboration with Alstom Power, Innogy and
various universities and consultants. The basis for the GAS ZEP process is the
combustion of natural gas in an almost nitrogen free atmosphere. The cycle is
shown in Figure 39. Essentially, pure oxygen is separated from air and used
for combustion in a specially designed gas turbine. The energy from the
combustion is used to generate electricity and heat. The flue gas consists
mainly of water vapour and CO2. A large proportion of the flue gas is cooled
and circulated back to the compressor. The water is separated from the
recirculated gas by cooling and condensing. CO2 is separated from the process
and taken to a separate CO2 compressor for injection or storage. The primary
feature of the combustion system is that it is closed and has no emissions to
the atmosphere. The primary technical challenges associated with the project
are associated with combustion, turbomachinery components and materials.

(124)

Figure 39: GAS ZEP concept (courtesy of Mitsui Babcock Energy Ltd).
Another conceptual HRSG design previously undertaken by Mitsui Babcock
was on behalf of the Carbon Capture Project (CCP) [89]. This assessed the
feasibility of a HRSG fired on natural gas and oxygen (oxyfuel) via in-duct
burners with various degrees of flue gas recycle.
NovelEdge Technologies, LLC has recently patented a process that is
applicable to combined cycle power plants. It specifies a plant that has a
relatively large steam turbine capacity (the low cost power-producing
machine) and less gas turbine capacity. Utilising a unique heat recovery
method (HRSG) in conjunction with a more efficient steam cycle, the
NovelEdge Technologies' cycle offers more high-efficiency power from a
compact arrangement of equipment. The technology lends itself especially
well to the re-powering of conventional steam plant, because the NovelEdge
steam cycle is much better adapted to the existing steam turbine and balance
of plant equipment.
7.4.4

Ongoing Gasification Development

Mitsui Babcock alongside Ulster University has been involved with gasifier
development studies of both the IGCC system and the Air Blown Gasification
Cycle (ABGC) [70, 84, 85]. Whilst the technology behind IGCC is proven and
plants are currently being built and operated around the world, the ABGC is
still considered some way behind at a pre-demonstration stage.

(125)

ABGC essentially combines two proven technologies of IGCC and fluidised


bed combustion. However, claims that ABGC can achieve greater efficiency
than any other coal based technology (47% compared with a ~43-45% for
IGCC) make it an attractive option. In addition, a further benefit is that ABGC
is generally considered better for poor quality, high sulphur coal, as found
commonly in China.
Gasification, albeit at a current loss in efficiency, allows the benefits of gas
turbine technology to be utilised with another fuel type. Thus the stranglehold
that the price of gas has on gas turbine and combined cycle auxiliary
equipment manufactures (including HRSGs) is relaxed by the broadening of
the fuel base. Furthermore developing countries such as China and India,
which form the future energy markets, are abundant in coal resources and
limited in gas. Therefore in order that UK companies maintain a long-term
future, UK research has to continue to be proactive within this area.

(126)

7.5

Conclusions

There are a number of utility-scale HRSG turnkey contractors operating


within the UK (e.g. Alstom Power, Mitsui Babcock, Foster Wheeler
Energy Ltd, Mott MacDonald, Nooter/Eriksen-CCT Ltd and Siemens
KWU). Whilst Mitsui Babcock has its headquarters and manufacturing
facilities in the UK, the remainder have their headquarters and
manufacturing facilities overseas or subcontract the manufacture. TEI
Greens is the largest independent manufacturer of utility-scale HRSGs in
the UK and has manufactured utility-scale HRSGs for most of the worlds
leading boiler designers/makers on both domestic and export projects.

Wellman Robey, BIB Cochran, ME Engineering and TEI Greens are UK


companies of UK origin with the capability to design and supply industrial
HRSGs. Wellman Robey and BIB Cochran manufacture shell type boilers
and have UK manufacturing facilities. ME Engineering supply water tube
or shell boilers manufactured outside the UK. TEI Greens design and
manufacture industrial HRSGs of water tube or smoke tube design in the
UK. The UK arm of Nooter/Eriksen also supplies industrial HRSGs but its
design capability is based in the US.

In such a competitive HRSG market, licensing agreements and


collaborative partnerships have been necessary in order for companies to
maintain the ability to compete. Under such conditions the requirement to
be continually developing new technologies is vital.

Areas of current research include once through technology, the effects of


cyclic operation, new forms of gasification and the use of HRSGs in novel
low emission power cycles.

(127)

OVERALL CONCLUSIONS

Current status

Current state of the art utility scale HRSGs operate at HP steam conditions
of up to 124 bar/565C and allow the associated CCGT to deliver
electrical power at a claimed net efficiency of up to 60%. They are
generally two pressure or three pressure with reheat, and may be of either
vertical or horizontal gas flow.

The capital cost of new-build CCGT plant is around 425/kW, with the
HRSG accounting for 10-15% of this total. The estimated delivered energy
cost for a CCGT in the UK is around 2.2p/kWh.

Current state of the art industrial HRSGs generally operate at lower steam
conditions than utility scale plant, and are usually of single pressure
design. They often include provision for supplementary or auxiliary
(stand-alone) firing. Highly fired units may incorporate a water-cooled
furnace. Lower pressure industrial boilers are usually of shell rather than
water tube design.

The recent trend has been for CCGT plant to be built under turnkey
contract. Whilst this does have advantages to the user in terms of
accountability, it does tend to mean that the user has less influence on the
detailed HRSG design.

Operational experience with HRSGs indicates that inclusion of specific


design features and attention to detail during fabrication are just as
important as the overall HRSG design, and that non pressure parts can be
as problematic as pressure parts.

Key areas for improvement include build quality, access for in service
inspection & maintenance, control & instrumentation and capability for
flexible operation. Overall cycle chemistry philosophy also needs to be
more thoroughly considered at the design stage.

The current challenge for operational HRSGs, particularly in the UK, is


the need to cycle plant which has been designed for and/or previously
operating at base load. Many users are currently carrying out
investigations/trials and plant modifications.

New & Developing Technologies

Future increases in HRSG operating conditions will largely be dictated by


increases in GT exhaust temperature.

One area of significant interest is once through design. The main benefit of
this technology at present is its suitability for flexible operation. In the

(128)

long-term future, it should pave the way for supercritical cycles with even
higher thermal efficiencies.

Another area of significant interest is the use of HRSG steam for GT blade
cooling in the latest class of GTs. This presents significant challenges for
HRSG design in attaining the high steam qualities required.

The use of HRSGs within IGCC plant has now reached the status of
commercial operation, although the costs still remain relatively high.

Other development areas include modular design to reduce build costs,


improving reliability, and improving access. The latter two items address
specific problems experienced by plant operators. Improvements in these
areas are perceived to provide product differentiation in an extremely
competitive market place.

Industrial scale HRSG technology is relatively mature. Most development


comes from the integration of HRSGs within new processes, and the
trickling down of technology from utility scale HRSGs. One exception is
the use of once through technology which is already standard practice for
one HRSG supplier, albeit on a small scale.

World Wide Activities

Over the past ten years (1992-2001), the biggest buyers by far of utility
scale HRSGs have been in the USA, with 48% of all worldwide purchases.
Next are the United Kingdom and Japan with 4% of the total sales.

Key manufacturers in the above period were Alstom Power (14.2%),


Nooter/Eriksen (12.6%), Deltak (9.5%), NEM (7.7%) and Aalborg
Industries (7.5%). Companies outside this top five in the 1-7% share of the
market included Foster Wheeler Energy (6.4%), Mitsubishi Heavy
Industries (4.3%), Doosan Heavy industries (4.0%) and Mitsui Babcock
Energy Limited (1%).

For industrial scale HRSGs, there were around 33% of sales in each of the
USA and Europe, with the other leading market being Asia and Australasia
(excluding China) with 19%.

Market Potential

Whilst the utility scale HRSG market has been healthy in recent years,
there is a predicted sharp downturn in the HRSG market in the shortmedium term due to plant over capacity. The situation is not expected to
pick up again until around 2007-2011. Key future HRSG markets are seen
as the USA and China (via IGCC). Non technical barriers in these two
markets include the price/availability of gas in the USA and
administrative/financial issues in China.

(129)

For industrial scale HRSGs, the European market is depressed due to


falling electricity prices and rising gas prices. However potential markets
include Russia, Central and Eastern European (CEE) countries, Turkey
and the Middle East. In the USA, despite problems on the utility scale,
there are still opportunities for development of CHP schemes on industrial
sites, largely driven by security of price and supply issues in the volatile
deregulated electricity market.

UK Activities

The current surplus of generating capacity in the UK and fluctuations in


the price of natural gas have led to a low requirement to build large scale
power generation plant within the UK. The only plus side is that with the
effects of flexible operation becoming more apparent, plant performance
upgrade opportunities are present.

The combination of NETA and a high natural gas price has dramatically
reduced the market for CHP and the climate change levy is not seen as
being an adequate incentive to invest in new HRSGs / CHP schemes.
Enhanced government support for CHP is required if the target of 10 GWe
by 2010 is to be met.

Wellman Robey, BIB Cochran, ME Engineering and TEI Greens are UK


companies of UK origin with the capability to design and supply industrial
HRSGs. Wellman Robey and BIB Cochran manufacture shell type boilers
and have UK manufacturing facilities. ME Engineering supply water tube
or shell boilers manufactured outside the UK. TEI Greens design and
manufacture industrial HRSGs of water tube or smoke tube design in the
UK. The UK arm of Nooter/Eriksen also supplies industrial HRSGs but its
design capability is based in the US.

There are a number of utility-scale HRSG turnkey contractors operating


within the UK (e.g. Alstom Power, Mitsui Babcock, Foster Wheeler
Energy Ltd, Mott MacDonald, Nooter/Eriksen and Siemens KWU). Whilst
Mitsui Babcock has its headquarters and manufacturing facilities in the
UK, the remainder have their headquarters and manufacturing facilities
overseas or subcontract the manufacture. TEI Greens is the largest
independent manufacturer of utility-scale HRSGs in the UK and has
manufactured utility-scale HRSGs for most of the worlds leading boiler
designers/makers on both domestic and export projects.

In such a competitive HRSG market, licensing agreements and


collaborative partnerships have been necessary in order for companies to
maintain the ability to compete. Under such conditions the requirement to
be continually developing new technologies is vital.

Areas of current research include once through technology, the effects of


cyclic operation, new forms of gasification and the use of HRSGs in novel
low emission power cycles.

(130)

REFERENCES

1. Powerplant Technonlogy, M.M.El-Wakil, McGraw Hill, 1984


2. D Foster-Pegg, Cycles for Power Generation (Paper 1), Seminar notes:
The Design of Gas Turbine Combined Cycle and Cogeneration Systems,
5th 7th July 1999 London.
3. ABB, Combustion: Fossil Power, Published 1991.
4. Powergen plc Environmental Report 2000, www.pgen.com
5. Babcock and Wilcox, Steam: Its Generation and Use, 40th Edition 1992.
6. B J Davidson and K A Keely, The Thermodynamics of Practical
Combined Cycles, Seminar on CCGT, 12th Oct 1990.
7. Digest of UK Energy Statistics, 2002, HMSO, Ch6
8. A Quality Assurance Programme for Combined Heat and Power at
www.defra.gov.uk/environment/consult/chp/response/pdf/chp.pdf

9. Personal Communication between M E Engineering Ltd. and Innovative


Steam Technologies.
10. Private communication between Power Technology and Siemens KWU,
19th September 2002.
11. M Pearson & R Anderson, Reliability and Durability from Large
HRSGs. Presented at CCGT-5, CCGT Plant Components Development
and Reliability, 23rd October 1998.
12. Private communication between Power Technology and Innogy plc, 9th
October 2002.
13. Private communication between Power Technology and Nooter/Eriksen,
11th September 2002.
14. Private communication between Power Technology and NEM, 4th
December 2002.
15. Private communication between Power Technology and Standard Fasel
Lentjes, 9th December 2002.
16. D Bogert, Dealing with the Rigours of Cyclic CCGT Operation: An
Operators Perspective. Modern Power Systems, October 2001, pp 33-35.
17. J M Pearson & R W Anderson, Root causes of transient tube temperature
anomalies measured in horizontal gas path HRSGs. Published in the

(131)

Boiler Tube Failures and HRSG Tube Failures and Inspections


International Conference, Phoenix, Arizona, 6th-8th November 2001.
18. A C Jones & S T Simpson, Power Technology Internal Report
PT/01/BB1182/M.
19. R B Dooley, T A Kuntz, W McNaughton, S R Paterson, M Pearson & K J
Shields, EPRIs HRSG Tube Failure (HTF) Manual. Published in the
Boiler Tube Failures and HRSG Tube Failures and Inspections
International Conference, Phoenix, Arizona, 6th-8th November 2001.
20. L B Lim & C Dometakis, Effects of cycling on the remaining life of
HRSGs. Modern Power Systems, July 2001, pp25-27.
21. M Pearson, Warning: Cycling HRSGs can be dangerous to your health.
Power, Jan/Feb 1997, pp 49-52.
22. D Bogert, HRSG Flexibility. Presented at the ETD/EPRI Two-Shifting
Seminar, 25th-27th June 2001.
23. Power Technology, Boiler Section File Reference 900/38/3, 900/38/3v2 &
900/38/3v3.
24. Vogt-NEM, Cyclic Operation of HRSGs. Technical Bulletin, 27th July
1999.
25. R R Harries & D A Smith, Power Technology Internal Report
PT/98/PA882/M.
26. K C G Bindemann, Power Technology Internal Report PT/00/BH1121/R.
27. J Davison and S Simpson, Power Technology Internal Report
PT/98/PA437/M.
28. D Maddison, Power Technology Internal Report PT/98/NA394/M.
29. J Davison, G Furnival, P Gallagher, R Harries, D Jarvis, K Manning, B
Nicholls, P Roberts and S Simpson, Power Technology Internal Report
PT/01/BC1018/M.
30. D M Blood & S T Simpson, Power Technology Internal Report (in draft).
31. B W Hirons, Power Technology Internal Report PT/00/BB930/M.
32. J McMorrow & S T Simpson, Power Technology Internal Report
PT/00/BB1174/M.
33. A C Jones & S T Simpson, Power Technology Internal Report
PT/99/PA491/M.

(132)

34. S T Simpson, Power Technology Internal Report PT/00/BB1077/R,


Section 4.
35. D M Blood & S T Simpson, Power Technology Internal Report
PT/00/BB1265/M.
36. Power RE-View, Volume 1, Issue 3, October 2001.
37. S T Simpson, Power Technology Internal Report PT/00/BB1077/R,
Section 10.
38. D M Blood & S T Simpson, Power Technology Internal Report
PT/01/BB1250/M.
39. A C Jones & S T Simpson, Power Technology Internal Report
PT/97/PA1525/M.
40. R G Adams & D Snow, Power Technology Internal Report
PT/01/BB1284/M.
41. British Standard BS2486:1997. Recommendations for Treatment of Water
for Steam Boilers and Water Heaters.
42. Guidelines for Boiler Feedwater, Boiler Water and Steam of Steam
Generators Operating at Pressures of 68 bar and Greater. VGB-R-450L,
1998. VGB, Essen, Germany.
43. Interim Cycle Chemistry Guidelines for Combined Heat Recovery Steam
Generators (HRSGs). EPRI Report TR110051 (1998). EPRI, Palo Alto,
California.
44. Food and Drug Administration (USA), Department of Health and Human
Services. Part 173 Secondary Direct Food Additives Permitted in Food
for Human Consumption. Section 173.310 Boiler water Additives.
(Commonly known as FDA Regulations.)
45. B Chexal, J Horowitz, et al. Flow Accelerated Corrosion in Power Plants.
EPRI Report No. TR-106611, (1996). EPRI, Palo Alto, California.
46. R R Harries and M J Willett. Flow Accelerated Corrosion in HRSG:
Interdependence of Cycle Chemistry and Design. Proc 6th International
Conf. On Fossil Plant Cycle Chemistry, Columbus, Ohio, June 27 29,
2000. Published as EPRI Report No. TR-1001363. EPRI, Palo Alto,
California.
47. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for
Drum Units. Epri Report No. TR-103665, December 1994. EPRI, Palo
Alto, California.

(133)

48. A Whitehead and R G Rowe. Water and Steam Use in Gas Turbines in
GE Combined Cycle Plants. Proc 6th International Conf. On Fossil Plant
Cycle Chemistry, Columbus, Ohio, June 27 29, 2000. Published as
EPRI Report No. TR-1001363. EPRI, Palo Alto, California.
49. S J Geary and J C Bellows. System Monitoring For Chemistry Control.
Proc. 61st International water Conference, Pittburgh, Pennsylvania,
October 2000. Engineers Soc of W Pennsylvania.
50. R Svoboda, F Gabrielli, E Liebig, H Hehs, and H Sandmann. Combined
Cycle Power Plant Chemistry Concepts and Filed Experience. Proc. 6th
International Conference on Cycle Chemistry in Fossil Plants. Columbus,
Ohio, June 27 29, 200. EPRI Report 1001363, April 2001.
51. M Pearson, Design and Operation of Combined Cycle Units to avoid
Cracking in Superheaters of Heat Recovery Steam Generators. Presented
at the Thermal Plant Repowering Conference, Ontario, 1995.
52. Powergen Internal Source, 2002.
53. Performance and Innovation Unit Report, The Energy Review. Crown
Copyright, February 2002.
54. Department of Trade and Industry Low Carbon Workshop, 21st June,
2002.
55. EPRI report AP-5772 1988
56. P Thame, Power Technology Availability Database.
57. C J Davis, Power Technology Boiler Tube Failure Database.
58. Personal Communication between M E Engineering Ltd. and Bowman
Power Systems Ltd.
59. Thermax marketing literature
60. J F Galopin, HRSG Technology: Going Supercritical - Once Through is
the Key, Modern Power Systems 1998, Vol 18 (12) P39-43.
61. R V Peltier, Special Report- Avanced Gas Turbines And HRSGs, Power
Magazine Sept / Oct 2001
62. Maher Elmasri, Triple Point and reheat Combined Cycle design Point
thermodynamics (paper14), Seminar notes: The Design of Gas Turbine
Combined Cycle and Cogeneration Systems, 5th 7th July 1999 London.
63. D J Moeller and G O Ganowski, Steam Injected Gas Turbines Enhance
Cogeneration Plant Performance, Pulp and Paper Sept 1987

(134)

64. D Foster Pegg, Steam Injected Gas Turbine Cycles (Paper 22),
Seminar notes: The Design of Gas Turbine Combined Cycle and
Cogeneration Systems, 5th 7th July 1999 London.
65. GT Pro, Thermoflow Inc.
66. http://www.power.alstom.com
67. http://www.gepower.com
68. DTI Publication, Technology Status Report, Supercritical Cycles for
Power Generation Applications, Jan 1999.
69. D Foster-Pegg, The Use of Coal, Seminar notes: The Design of Gas
Turbine Combined Cycle and Cogeneration Systems, 5th 7th July 1999
London.
70. G B Welford, D J Dillon and Xu Shisen, Introduction to China of
Supercritical and Boilers and emerging Clean Coal Technologies, Mitsui
Babcock Report 30/01/039, Oct 2002.
71. J D De Graaf, P Zuideveld and F G Van Dongen, Shell Pernis Netherland
Refinery Residue Gasification Report, Seminar: Gasification For The
Future, 11-14 April 2000
72. Thermal processing of Biomass for Fuels and Chemicals,
A.V.Bridgwater, in proceedings of Renewable Bioenergy, Technologies,
Risks and Rewards, ImechE conference, London, 29-30.10.2002
73. R W McCoy, Power Reports : Heat Recovery Steam Generators, Dec 31
2001
74. Personal communications between M E Engineering Ltd. and the
following: Esscano Power Services Ltd., Centrax, Innovative Steam
Technologies, CHPA, Nedalo, Bowman Power Systems, Wellman Robey,
ALSTOM, BIB Cochran, Advantica Technologies Ltd., Siemens Power
Generation, ABCO Boilers, Vogt-NEM, Rentech Boilers, USCHPA.
75. National Grid Company, Seven Year Statement, May 2002.
76. Energy and Environment in the European Union, Environmental Issue
report No. 31
77. DTI statistics at www.og.dti.gov.uk/information/bb_updates/chapters/Charts2_42_8.htm

78. OFGEM statistics at www.ofgem.gov.uk/neta/prices_neta.htm

(135)

79. Onwards from 7% - new survey from WADE quantifies decentralised


energy world-wide M.Brown, Co-generation and On-site Power
Production, Vol.3 Issue 6 2002
80. Future Cogen Study @ www.cogen.org/projects/future_cogen.htm
81. Directive of the European Parliament and of the Council on the promotion
of cogeneration based on a useful heat demand in the internal market,
COM(2002) 415 final, Commission of the European Communities
82. Salomon Smith Barney Power curve
83. WEFA Inc, Banking on Advanced Gas Turbines: Prospects for a
Financial Meltdown
84. G B Welford, P J Melling, A Martin, S Lu, F Wood and W Livingston,
Gasifier Developments and the Air blown Gasification Cycle, Mitsui
Babcock Report 37/00/090, Sept 2000.
85. D J Dillon, A Bartleman and B Xu, Heat Recovery Steam Generators For
IGCC Demonstration Plant In China, Mitsui Babcock Report E/01/039,
July 2002.
86. Modern Power Systems, July 2000, pp33-35.
87. I Mech E, Speakers Presentations, The Heat Recovery Steam Generator
User Group, 29-30 Oct 2002.
88. Vision: The Newsletter of the Foresight and Link Initiatives, Down to
Zero Autumn 2001: Issue 1
89. P I Moore and S A Hume, The Carbon Dioxide Capture Project: Zero or
Low Recycle Oxyfuel Steam Generator (In-Duct Burner) Pre-Study,
Mitsui Babcock Report E/01/039, July 2001

(136)

10 ACKNOWLEDGEMENTS
Thanks to all those who helped us in carrying out the review and those who
responded to the HRSG questionnaire.
In addition, special thanks to: The Department of Trade and Industry
G Brown, D Bogert and M Lee of Innogy plc.
E Reijns of NEM.
P Parker and D Stean of Nooter/Eriksen UK.
J Franke and W Heese of Siemens KWU.
F Baltussen and R van den Bosch of Standard Fasel Lentjes.
ABCO Boilers
Advantica Technologies Ltd.
ALSTOM
BIB Cochran
Bowman Power Systems
Centrax
CHPA
Esscano Power Services Ltd.
Innovative Steam Technologies
Nedalo
Rentech Boilers
Siemens Power Generation
USCHPA
Vogt-NEM
Wellman Robey

(137)

11 APPENDIX A
Company Name

Country of Website
Origin

Aalborg
Engineering

Denmark

www.aalborgengineering.com

Aalborg Industries

Denmark

www.aalborgindustries.com

Ultimate Parent
Notes
Company and Country
of Origin
Aalborg Engineering, Supplier of industrial and small utility HRSGs
Denmark?
Highly fired HRSGs for use with GTs in range 4-50MWe
producing steam flows at up to 200 tph, 20 - 120 bara and
saturated - 540C
Vertical gas pass HRSG for use with GTs in range 10130MWe producing steam flows 15 - 250 tph, 3 - 120 bara
and saturated - 540C
Bespoke HRSG designs for exhaust gas streams including
those with high dust loads and aggressive compositions
Aalborg Industries
Industrial HRSGs
A/S, Denmark
Manufacture and supply HRSG products: Water tube HRSG for diesel engines of 0.5MW and
above;
Forced and natural circulation HRSGs up to 40 barg for
gas engines >1MWe
Hot water boilers for microturbines 75-250kWe
HRSGs for gas turbines in the range 2 23 MW with
exhaust flow up to 60C, 50kgs-1 to produce steam flows
up to 50 tph
Exhaust gas HRSGs of smoke and water tube design to
produce steam flows 0.5 50 tph at pressures up to 40

(138)

Aprovis Energy
Systems GmbH

Germany

Erie Power
USA
Technologies, Inc.
(formerly Aalborg
Industries Inc. until
01.10.2002)
AB&Co Thermal
Denmark
Transfer Ltd., TT
Boilers
Abco Boilers
USA

Alstom Power
(Industrial boiler
business)
Alstom Power
(Large HRSG
business)

www.aprovisgmbh.de

Aprovis Energy
Systems GmbH,
Germany
DaeKyung Machinery
& Engineering Co.
Ltd., Korea

www.abco.dk/

www.abcoboilers.c
om

AB&Co Thermal
Transfer A/S,
Denmark
Peerless Mfg. Co.,
USA

Alstom, France

USA

www.power.alstom
.com

Alstom, France

(139)

barg
Industrial
HRSGs for industrial scale CHP in range 20-6000kWth
Water and smoke tube designs and designs optimised for
biogas and landfill gas
Utility
HRSGs for gas fired combined cycle power plant

Industrial HRSG
Small package HRSGs with steam output up to 5 tph at 0.5
190 barg.
Industrial HRSGs
Capability to design and manufacture a wide range of HRSGs
for industrial applications. Fire tube and water tube designs
including designs with membrane water walls.
Industrial
HRSGs for GT < 50MWe
HRSGs for process applications
Utility
GT > 50 MWe
OCC optimised for cycling and constructability
Horizontal and vertical gas pass utility HRSG designs
165 bar (2400 psi ) once through steam generator design

Babcock & Wilcox

www.babcock.com

McDermott
International, USA

Babcock Borsig
Power

Germany

Babcock Borsig,
Germany (filed for
insolvency on
04/07/2002)
www.bbpwr.com
Babcock Borsig,
Germany (filed for
insolvency on
04/07/2002)
www.bhk.co.jp/eng Babcock Hitachi,
lish/product/prd
Japan
main.htm

Babcock Borsig
Inc.

USA

Babcock Hitachi

Japan

BIB Cochran

UK

www.bibgroup.co.uk

Mechmar Corporation
(Malaysia) Berhad,
Malaysia

Chanute
Manufacturing
Company
CMI \ Larsen &
Toubro

USA

www.optimustulsa.com

Optimus Corporation,
USA

www.cmi.be/utility
-boilers/index.htm

CMI Group, Belgium

www.babcockborsi
g.com

(140)

Utility / Industrial
Utility scale combined cycle plant
Industrial scale HRSGs for waste incinerators, paper and
pulp mills etc.
Utility and smaller GT HRSGs
Variety of desing features for HRSGs behind 4 200
MWe GTs, giving up to 500tph steam at up to 175 barg
and 565C.
Utility and industrial
Natural circulation water tube HRSGs
Benson type once through steam generators
Utility
Utility scale HRSGs for combined Cycle power plant
Catalyst modules for HRSGs
In house manufacture of units and finned tube
Industrial
Smoke tube designs to produce steam at up to 35 tph and
up to 35 barg behind small GTs, waste incinerators and
other process exhausts
Contract manufacture of HRSG modules

Utility and industrial


Variety of design features, inc. once through and
supercritical
CMI Utility HRSG, GT > 80MWe 260MWe

Deltak

USA

www.deltak.com

Doosan Heavy
Industry and
Construction
Company

Korea

www.doosanheavy. Doosan Heavy


com/default.htm
Industry and
Construction
Company, Korea

Energy Recovery
International

USA

www.hrsg.com

Aqua-Chem Group,
USA

IHI (IshikawajimaHarima Heavy


Industries Co.)
Innovative Steam
Technologies

Japan

www.ihi.co.jp/ihi/p
roducts/productse.html
www.otsg.com

Ishikawajima-Harima
Heavy Industries Co.
Ltd., Japan
Aecon Group Inc.,
Canada

M E Engineering

UK

www.meengineering.co.uk

Thermax Group, India

Canada

Deltak LLC., USA

(141)

CMI Industrial HRSG, GT <80MWe and behind process


In house manufacturing capacity
Utility and industrial
HRSGs with a variety of designs for GTs 1MWe250MWe.
Specialist water and fire tube designs for industrial and
process applications
Utility
Took over Samsungs power plant business in 1999
Horizontal gas path, natural circulation and vertical gas
path, forced circulation designs for utility combined cycle
power plant
Industrial
Water wall, highly fired packaged HRSG for GTs 1 15
MWe to produce 9 90 tph steam
Water tube packaged HRSG for GTs 1 15 MWe to
produce 4.5 90 tph steam
Single / multiple pressure modular natural circulation
HRSG for GTs 15 50 MWe
Utility
HRSGs for Utility scale combined cycle power plant
Coal IGCC power plant
Industrial and utility
Once through steam generator designs, behind GTs only
Mainly industrial scale, but some larger units.
Industrial
Wide range of water tube, fire tube and once through

Mitsubishi Power
Systems

Japan

www.mhi.co.jp/po
wer/e
power/product/boil
er/hrsg/index.htm
www.mitsuibabcoc
k.com

Mitsubishi Heavy
Industries Ltd., Japan

Mitsui Babcock

UK

NEM

Netherlan
ds

www.nem.nl

Babcock Borsig,
Germany, (filed for
insolvency on
04/07/2002)

Nooter / Eriksen

USA

www.nootereriksen CIC Group Inc., USA


.com

Rentech Boiler
Systems Inc.

USA

www.rentechboiler
s.com

Mitsui Engineering
and Shipbuilding,
Japan

Rentech Boiler
Systems Inc., USA

(142)

designs
HRSGs behind GTs<70MWe
Exhaust heat boilers behind reciprocating engines
Waste heat boilers behind incinerators, oxidisers etc.
Process integrated boilers
Utility
HRSGs for Utility scale combined cycle power plant
Natural and forced circulation, vertical gas flows
Natural circulation, horizontal gas flow
Utility
large fossil fuel steam generating plant, environmental
control systems and equipment.
HRSG manufacture and supply

Mainly large scale utility HRSGs

Natural and forced circulation designs


Utiltity and industrial
Mainly large HRSGs behind GTs up to 250 MWe
Variety of designs forced / natural circulation, fired and
unfired etc.
Also boilers behind waste incineration systems
Utility
HRSGs behind utility scale GTs
Favour horizontal gas path, natural circulation design
Industrial
HRSGs behind GTs in range 3-50MWe
Wide variety of water and fire tube designs for

Siemens

Germany

www.seimens.co.u
k

Siemens, Germany

Standard FaselLentjes BV

Netherlan
ds

www.mitco.co.kr/
m101.html

Modern Industrial
Technology Co.,
Korea

TEI-Greens

UK

www.teigreens.co
m

Thermal Engineering
Holdings

Thermax
Thermax Babcock
and Wilcox

India

www.thermaxindia. Thermax Group, India


com

Vogt-NEM

USA

www.nem.vogt.co
m

Was owned by NEM,


Netherlands, in turn
owned by Babcock

(143)

applications in waste incineration and process industries


Inclusion of catalysts possible

Optimised for cycling and constructability


Enhanced Oil Recovery OTSGs for 80% quality steam
OTSG HRSGs with Benson Technology
Utility
Developed and license Benson once through design, but
do not manufacture
Industrial
Natural and forced circulation water tube designs
For GT, reciprocating engine, FCC units, process furnaces
etc.
Utility and industrial
Power, petrochemical, chemical and offshore industries
Supply finned tube and HRSG components to other HRSG
suppliers
Industrial
Wide range of water tube, fire tube and once through
designs
HRSGs behind GTs<70MWe
Exhaust heat boilers behind reciprocating engines
Waste heat boilers behind incinerators, oxidisers etc.
Process integrated boilers
Utility and industrial
Mainly large HRSGs behind GTs 25MW up to largest
available

Borsig, Germany,
(filed for insolvency
on 04/07/2002)
Now owned by
Babcock Borsig
Capital Co.
STF Spa, Italy

STF Spa

Italy

www.stf.it

Wellman Robey

UK

www.welllmanrobe Wellman Group, UK


y.com

(144)

Variety of designs forced / natural circulation, fired and


unfired etc.
Access to once through design
Some industrial applications e.g. steam generation for
enhanced oil recovery

Utility
Manufacture of Mitsui-Babcock design HRSGs for utility
combined cycle application under licence
45 450 tph steam
Industrial
Fire tube design only
Typically behind GTs around 5MWe size
Also for some process applications