REVIEW OF STATUS OF ADVANCED MATERIALS FOR POWER GENERATION

Report No. COAL R224 DTI/Pub URN 02/1509

by J E Oakey, Cranfield University L W Pinder, Powergen UK UK plc R Vanstone and M Henderson, ALSTOM Power S Osgerby, National Physical Laboratory

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 February 2003

REVIEW OF STATUS OF ADVANCED MATERIALS FOR POWER GENERATION

J E Oakeya, L W Pinderb, R Vanstonec, M Hendersonc and S Osgerbyd

SUMMARY
The construction of economically viable and durable power plant components is dependent on the selection of the most appropriate materials and fabrication methods. The implementation of advanced materials in power generation applications has resulted in significant advances in plant performance and hence emissions reductions. Advanced materials are now being deployed in new and existing power plant to improve operating performance and reliability, availability, maintainability and operability. These materials, from ferritic alloys for boilers to ceramics for filter elements or coatings for the protection of gas turbine blades, are the result of extensive research programmes. In some cases, it can take over ten years to take a new material concept from initial trials to implementation, supported by the necessary long term mechanical and environmental performance data. This review addresses advanced materials and their applications in current and future coal power technologies relevant to the UK and UK industry exports. All the key power technologies are covered from supercritical pulverised coal boilers, through steam turbines and gasifiers to gas turbines and advanced cycles involving fuel cells and CO2 management. From the review, the R&D priorities for the UK can be divided into three overarching clusters: high temperature materials for boilers, steam turbines, gas turbines, gasifiers, high temperature heat exchangers, as well as functional materials such as sorbents, catalysts and membranes protective systems/coatings for the same technology areas as high temperature materials, and modelling of materials processing, component manufacture and life assessment. The review concludes that materials technologies are critical to achieving significant improvements in power plant efficiency. If ignored, they represent a major cost factor and a key constraint that will limit both: the implementation of new generation technologies and; the adaptation of power plant technologies to new fuels, thereby reducing opportunities for improved environmental performance. Materials research is a major part of any initiative to produce cleaner/cheaper energy systems. Innovative materials from concept to implementation can take decades to develop. UK materials programmes must, therefore, have the foresight and resources to pursue the necessary technology through to practical demonstrator outcomes. As a result, this review recommends that an integrated long term materials R&D core programme is needed for power generation covering three key elements - high temperature materials, protective systems and modelling. A long term view is essential. Low risk, near term development will not deliver the necessary benefits.
a

Cranfield University, b Powergen, c ALSTOM Power, d NPL

TABLE OF CONTENTS

SUMMARY
2 1.

Pages

INTRODUCTION............................................................................................................................... 1

1.1
1.2
2.

Background ........................................................................................................ 1 Aims and Objectives of the Review ....................................................................... 2 Boilers and Related Technologies .................................................................... 6
Combustion Systems ........................................................................................................... 6 Pressure Part Design........................................................................................................... 8 Operating Environments and Failure Modes: PF-fired Plant...................................... 11 Operating Environments and Failure Modes: FBC Plant ............................................ 16 Flue Gas Desulphurisation ............................................................................................... 21 Component Integrity and Life Assessment..................................................................... 21 State of the Art PF-fired Boilers ...................................................................................... 21 State of the Art FBC ......................................................................................................... 27 Future R&D Priorities...................................................................................................... 27

3.

IDENTIFICATION OF PLANT OPTIONS .................................................................................... 4 MATERIALS REVIEWS .............................................................................................................. 6

3.1
3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.1.7 3.1.8 3.1.9

3.2
3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6

Steam Turbines ................................................................................................ 48

Steam Turbine Market Trends ........................................................................................ 48 Steam Turbine Development Trends .............................................................................. 48 Steam Turbine Design and Operation............................................................................. 49 Steam Turbine Materials State of the Art ................................................................... 50 Steam Turbine Materials R&D........................................................................................ 52 Future R&D Priorities...................................................................................................... 54

3.3
3.3.1 3.3.2

Gasification Systems........................................................................................ 60
Gasification and Gas Coolers........................................................................................... 60 Gas Cleaning Systems ....................................................................................................... 64

3.4
3.4.1 3.4.2 3.4.3 3.4.4

Gas Turbines .................................................................................................... 75

Gas Turbine Market Trends ............................................................................................ 75 Gas Turbine Design, Operation and Materials ............................................................. 77 Review of UK Opportunities ............................................................................................ 88 Future R&D Priorities...................................................................................................... 90

3.5

Associated Technologies ..................................................................................... 99

4. 5. 6. 7. 8.

3.5.1 Introduction ....................................................................................................................... 99 3.5.2 Fuel Cells............................................................................................................................ 99 3.5.3 High Temperature Heat Exchangers .............................................................................. 99 3.5.4 CO2 Capture Technology ............................................................................................... 101 REVIEW OF UK OPPORTUNITIES .......................................................................................... 105 FUTURE R&D AND TECHNOLOGY TRANSFER.................................................................. 107 CONCLUSIONS ............................................................................................................................. 108 RECOMMENDATIONS................................................................................................................ 108 ACKNOWLEDGEMENTS............................................................................................................ 108

ANNEX A......................................................................................................................................1 Literature review..........................................................................................................................1

A.1 Introduction ..................................................................................................................................... 1 A.2 Low Alloy (1-3% Cr) Steels............................................................................................................ 1 A.3 9-12% Cr Martensitic Steels .......................................................................................................... 3 A.4 Austenitic Steels............................................................................................................................... 6 A.5 Ni- and Co-based alloys .................................................................................................................. 7 A.6 Intermetallics ................................................................................................................................. 12 A.7 Oxide Dispersed Strengthened (ODS) Alloys ............................................................................. 13 A.8 Ceramics......................................................................................................................................... 15 A.9 Fuel Cell Materials........................................................................................................................ 17 A.10 Corrosion Resistant Coatings................................................................................................... 18 A.11 Thermal Barrier Coatings (TBCs).................................................................................................. 24 A.12 Erosion Resistant Coatings.............................................................................................................. 25 A.13 References .................................................................................................................................. 27 ANNEX B................................................................................................................................................... 11

Listing of organisations involved in the preparation of the review .......................................11

1. 1.1

INTRODUCTION Background

The major challenges facing the Power Generation Industry in the 21st Century are focused primarily on achieving the difficult targets of increased efficiencies, increased reliability and stringent environmental standards. The challenge for coal, in particular, is to develop new, and evolve existing clean technologies, in order to achieve these goals. One of the core activities essential to the attainment of these future requirements is that of Materials Technology. Many of the required enhancements in plant performance are restricted by materials limitations and for this reason, a coherent and co-ordinated strategy outlining the future requirements for Materials R & D is essential. There are many activities currently ongoing worldwide relating to the Materials R & D requirements for the Power Generation Plant of the future. Recent publications from the Institute of Materials1 and the European Commission have addressed in part some of the key issues. The activities of collaborative programmes such as COST 501/5222 and Framework IV/V Joule/Thermie or Brite/Euram also include many materials programmes related to future generation requirements. Within the UK Foresight Programme, the Energy and Natural Environment Panel is continuing the work of the previous Energy Panel, which prepared reviews on Clean Coal3 and Gas Turbine4 Energy Technologies. These reviews highlighted Materials R & D as one of the key development areas for future advanced power plant. They also recognised that the long lead times required from material development to implementation in service, and the dependence of the UK on overseas suppliers, means that a well defined co-ordinated strategy is required in the very near future to retain our competitiveness into the 21st Century. An industry-led grouping, the Advanced Power Generation Task Force (APGTF)(recently renamed as the Advanced Power Generation Technology Forum), was formed in 1998 as a Foresight Associate Programme, to take a sector-wide view of the issues facing the power generation industry in the UK and overseas5. The review, presented in this report, has been prepared by the Materials Technical Group, which provides advice on materials issues to the APGTF. The Materials Group is based on the former Institute of Materials Task Force on Power Generation which produced an overview report on the UKs activities in materials for the power generation sector in 19971. The Group represents the complete power generation supply chain from materials manufacturers through to electricity producers. ALSTOM Power Gas Turbine and Steam Turbine Corus Group Cranfield University Innogy Institute of Materials (IOM) NPL PowerGen Special Metals Wiggin Chromalloy UK Howmet Mitsui Babcock Energy Ltd Siemens Power Generation UK

The above organisations are involved on a day to day basis in materials issues relevant to all aspects of current and future power generation plant. This involves short-term problem solving (failure investigations), through plant integrity and remnant life assessments, to R&D on new materials and coatings to withstand the ever more onerous duties expected of power plant components. In addition, the organisations are major participants in all recent/current UK/European research projects of relevance to the review, for example:-

1 of 108

Clean Coal Power Generation Group (CCPGG) for the development of Air Blown Gasification Cycle, COST 522 European Collaborative Programme on Power Generation in the 21st Century: Ultra-Efficient Low-Emission Plant, THERMIE Advanced 700C Ultra-supercritical Power Plant Project, and Practical Improvements in Power Plant Efficiency through Materials Engineering (PIPPE).

A core team from the Materials Technical Group, comprising Cranfield University, ALSTOM Power, Powergen and NPL (with assistance from the Institute of Materials) will undertake the review, with active input from all the other APGTF Materials Technical Group members. Figure 1 illustrates the many sources of information and advice available to the team. For the purpose of this review, the term 'Advanced Materials' relates to materials or materials systems (e.g. combined alloys with coatings), which are being used today to improve the efficiency and reduce the environmental impact of power plants. These materials may be newly developed and suited to new plant design or retrofit applications to extend component lives or broaden a plant's operating envelope. Alternatively, they may be traditional materials in a new application for which little relevant performance data exist (e.g. the use of Ni alloys (developed for aerospace applications) in steam turbines). So the term 'advanced', relates to the application as much as to the materials themselves, which for power generating equipment have been developed in an evolutionary way over past decades. An example of an advanced application would be the effects of advanced process or engineering concepts on the materials selected on the basis of current best practice (e.g. steam cooling of gas turbine blading). 1.2 Aims and Objectives of the Review

The principal aims of the review were to assess the current global state of development of advanced materials for coal-based power generation and to identify and prioritise areas where further UK R&D and technology transfer activities need to be focused to enhance the market potential for UK materials products and services. The specific objectives were: to assess the global state-of-the-art and application of advanced materials in coal-based power generation systems to identify opportunities for UK developers, suppliers and service providers, taking into account current and future market influences to critically review current activities and capabilities of UK organisations and to identify priority areas in which future R,D&D activities need to be focused, to meet current and future demands and to enhance the market potential for UK products and services. The publication of the Review is expected to provide for: clearly directed areas of materials R & D in industry and academic/research establishments through DTI, EPSRC and similar initiatives. focused goals for alloy producers and processors to achieve the required material property targets, size, shape and fabricability requirements. utilisation of the above by the component and plant manufacturers to achieve the requirements of service conditions in terms of operating temperatures, environments and 2 of 108

stresses. attainment of target efficiencies and emissions by the plant operators. ensuring competitiveness of UK manufacturers products in the worldwide market. spin-off advantages for non-coal-fired power generation plant.

References for Section 1 1. Institute of Materials Task Force Report, Materials R & D Requirements for Power Plant: Into the 21st Century, 1997. 2. J E Oakey, D H Allen and M Staubli, Power Generation in the 21st Century The New European COST Action, Proc. 5th Int. Charles Parsons Conf: Parsons 2000 Advanced Materials for 21st Century Turbine and Power Plants. Edited by A. Strang, R. D. Conroy, G. M. McColvin, C. Neal and S. Simpson. IOM Communications, London, Book 736 (2000). 3. Foresight Research, Development and Demonstration Priorities for Cleaner Coal Power Generation Technology, DTI - OST, May 1999. 4. Foresight Research, Development and Demonstration Priorities for Gas Turbine and Advanced Combined Cycle Technology, DTI - OST, November 1999. 5. Advanced Power Generation Task Force, Annual Report 1999/2000.

3 of 108

2.

IDENTIFICATION OF PLANT OPTIONS

Before reviewing the status of advanced materials in coal-based power systems, it is necessary to identify the range of systems to be considered. There are several new initiatives that are currently ongoing which are looking at the strategic R and D requirements for global power generation over the next 20-30 years. Examples of these are: UKs current Foresight initiative; ECs Framework 6 consultations; USAs Vision 21; Japans New Sunshine programme. The APGTF issued a UK technology strategy out to 2030 in September 20011 to assist the identification of those power plant technologies worthy of UK development for home or export applications. This strategy focused on the research and development needs for fossil-fuelled power generation and associated technologies, including those for waste and biomass. Most future, global predictions2,3 still see the large, central power station as being dominant for new build over the next 10-15 years. The dominant technology for new coal plant is likely to remain as pulverised fuel (pf). For gas turbines (GTs), large plant will continue to dominate but there will be a growing market for the micro, small and mid-size GTs. Gradual growth of Distributed Generation (DG) is expected to continue and up to 20% of generation in EU could be from Distributed Generation by 2020. The major impact of renewables on the electricity market is not expected to be felt until after 2020. So, there is expected to be a continued drive for higher efficiency in fossil-fuelled plant which could act against the large central power station for some applications in the future and in favour of distributed generation, which may achieve higher efficiencies through combined heat and power applications. For fossil-fuel technologies, CO2 capture and sequestration is a future possibility. According to a recent US Department of Energy report4, Scientific experts are optimistic that (large scale) CO2 capture and sequestration could be implemented on a scale that would mitigate climate change ... and if successfully developed could allow the continued use of fossil fuels in the presence of carbon emission constraints. However, this is unlikely to be cheap; current estimates by the US Department of Energy are that it will add up to 50% on the electricity price to the consumer and this would still leave the problem of CO2 disposal. A more recent publication from the IEA5 gives estimates of the cost of capture as 50-70% of the cost of generation, depending on the technology. However, despite its expense, some utilities are carrying out studies on how to implement CO2 capture and sequestration on some existing coal plants. If greenhouse gas trading or a system of credits comes into force in the future, then this could provide an extra incentive for CO 2 capture or reduction technologies; it could also provide an additional strong impetus for the development of higher efficiency power generation technologies. In order to be competitive with the US and Japan out to 2030, then a UK/EC R and D programme must have a set of comparable targets. The scenarios presented in the APGTF Strategy suggest that post 2020, new power plant will need to have very low emissions as well as being commercially cost effective. The following would be required by 2030: efficiencies (in electricity) > 60% on coal > 70-75% on gas > 85-90%

thermal efficiencies near zero pollutant emissions cost-effective management of CO2 emissions cost-effective hydrogen technologies. 4 of 108

Other plant performance parameters such as reliability, availability and maintainability will also need to be better than present values. This puts a strong emphasis on the materials used for system components. Future plant options for the period up to 2030 must be based on those technologies which are either fully developed or close to market today. Therefore, the APGTF strategy identifies the following plant technologies as significant for the timescales considered and can use coal as a fuel (or be involved in a coal-based system): Pulverised Fuel (PF) Combustion. Gasification air and oxygen-blown Fluidised Bed Combustion Pressurised PF or Fluidised Bed Combustion Fuel Cells Gas Turbines

Following on from this listing of the main plant technologies the Strategy identifies the following key technologies: Hybrid or novel cycles Co-fuelling with biomass Hot gas clean-up Fuel flexible gasification Combustion High temperature heat exchangers Reformers and fuel cell fuelling Fuel cell electrochemical processes Membranes Plant manufacture Component integrity CO2 Management Plant Modelling Control Systems. Power Electronics (Hydrogen Storage and Production if Hydrogen Fuelled is included) All of the above have materials implications to greater or lesser extents; those highlighted in italics have been identified as priority technologies for this review. It should be noted that CO2 management is included as it can have important impacts on the selection, operation and performance of the main plant technologies in the medium to long term and may change significantly the operating environments of the materials used. References for Section 2 1. The Transition to Zero Carbon Emissions, a Technology Strategy for Power Generation from Fossil Fuel and Associated Technologies, 2001, APGTF. 2. World Energy Outlook, 1998, International Energy Agency. 3. European Union Energy Outlook to 2020, 1999, European Communities. 4. Carbon Management: Assessment of Fundamental Research Needs, 1997, US Department of Energy. 5. Carbon Dioxide Capture and Storage, 2000, IEA/DTI.

5 of 108

3.

MATERIALS REVIEWS

The status of materials technologies etc. given in this section is divided into four parts. The first deals with boilers and related technologies, the second with steam turbines, the third with gas turbines and finally the fourth associated and new technologies, which also deals with the possible effects of the introduction of CO2 capture.

3.1 3.1.1

Boilers and Related Technologies Combustion Systems

Pulverised Fuel Combustion Pulverised fuel (PF) has been the mainstay of coal-based electricity generation in the developed world for many decades. Sub-critical cycles, where the boiler and turbine operate below the steam/water critical pressure, represent the dominant technology to date. Over the years, there has been a progressive increase in the size, operating temperature and pressure of this plant, leading to increased thermal efficiency and a concomitant reduction in fuel costs. The constant drive to reduce fuel costs led to the development of super-critical designs (where the pressure exceeds the steam/water critical pressure) in the 1950s and 1960s. However, technical difficulties, largely related to the use of thick section austenitic steel components, led to unreliable plant with poor operational flexibility. Consequently, for plant built throughout the 1970s and 1980s, the steam temperatures have been limited to the 540-566C range. With increasing pressures on fuel costs, now being reinforced by ever more urgent environmental imperatives to reduce SO2, NOx and CO2 emissions, the energy industry has increasingly returned its attention to the development of super-critical steam cycles. Consequently, there has been a clear trend towards increasing steam temperatures and pressures in plant being designed and built throughout the 1990s. The potential improvement in heat rate achieved by increasing steam temperature and pressure in single and double reheat cycles from a base case of 535C/18.5MPa is shown in Figure 3.1.11. The higher steam parameters operative in advanced supercritical boilers necessitate higher component operating temperatures and stresses. A number of plant design and operational procedures, advocated by the boiler manufacturers, can reduce the impact of these conditions on pressure parts. However, the continuing upward trend towards progressively higher thermal efficiencies has identified that better materials will be required in a number of critical components if reliable and flexible plant operation is to be achieved. The contribution that some materials and other cycle developments can make to improvements in thermal efficiency, according to data provided by Siemens, are shown in Figure 3.1.2. These improvements are very significant in terms of the near future requirements to reduce worldwide CO2 emissions, since it is estimated that a 1% improvement in efficiency would result in a reduction of around 1M tonnes of CO2 emissions during the lifetime of an 800MW machine2. ELSAM have conducted study projects into IGCC, PFBC and USC plant for the construction of power generating plant in the range of 300-400MW. For the present and near future, they consider that USC offers the highest efficiency coupled with the cheapest investment3. Limits to the development of USC boiler plant are seen to be: the creep properties and weldability of low alloy steam generating tubes, the weldability, creep and thermal fatigue properties of materials for thick section steam separators, headers and steam pipes and, 6 of 108

the creep, fireside and steamside corrosion properties of superheater materials.

Fluidised Bed Combustion Fluidised bed combustion is a relatively young technology but is being used in most areas of the world to generate electrical power, provide district or process heating and combust industrial and municipal waste. Its particular advantages in relation to many other forms of fuel firing are: Ability to utilise a wide variety of fuels, including solids, liquids and gases, and fuels with a high fraction of inert ash or water. Sulphur originating from the fuel can be captured in the combustor bed by means of a sorbent, usually limestone or dolomite. Generation of the regulated nitrogen oxides (NOx - NO and NO2) is inherently lower than, for example, pulverised fuel (PF) firing due to the low combustion temperature. Additionally, the NOx emissions can be reduced, by at least 80%, by the relatively simple measure of introducing ammonia at the combustor outlet.

A disadvantage of FBC systems compared to PF plant with wet limestone/gypsum flue gas desulphurisation (FGD) is the lower utilisation of the limestone. Hence, the FBC will produce a greater amount of solids waste, by about 10 to 30% based on a typical coal, due to the presence of unused and partially reacted sorbent. Combustor systems take two general forms: bubbling and circulating beds. Bubbling beds occupy the lower part of a combustor vessel and have a distinct bed surface. They are fluidised at gas velocities, which give rise to upward flowing streams of bubbles which promote mixing. In circulating beds, fluidisation velocities are greater. Most of the bed material is entrained in the gas stream and passes through the full height of the combustor, to be separated out by a cyclone and returned to the bed vessel, either directly or via external fluidised bed heat exchangers. Both forms of FBC can operate at atmospheric pressure or pressurised conditions. With atmospheric systems, the combustor represents an alternative to other forms of boiler, with steam raised in the combustor walls, convective back-pass and in some cases, in-bed tube banks or external heat exchangers, being used to drive a conventional steam turbine plant or supply heat. Pressurised FBC typically forms part of a combined cycle, with the hot, pressurised gas from the combustor driving a gas turbine and generating power, which supplements that produced in a steam plant. In all variants of FBC, the bed is made up of a large mass of hot inert particles (fuel ash, spent sorbent etc) and the amount of solid fuel particles present in the bed is small, typically of the order of 1%. Atmospheric Fluidised Bed Boilers In atmospheric pressure plant, circulating bed combustors (CFBCs) have largely superseded bubbling beds due to a much longer contact time between the solid particles and the gas stream. This increases the efficiency of sulphur capture and improves carbon burnout. The fuel flexibility of CFBC units is particularly wide and has allowed the utilisation of materials, such as mining wastes and sewage sludges, which before the advent of FBC technology would not have been considered to be fuels at all. Additional advantages of CFBC over bubbling beds are that NOx emissions can be reduced (by means of air staging), less fuel feed points are required and load following is simpler. The increase in scale of CFBC plant from much smaller units, 50MWe or less, has occurred in the last ten or so years. Consequently, operating experience at the larger scale is restricted and 7 of 108

materials and operational problems are still being resolved. All operating plant use sub-critical steam conditions, in many cases not very advanced, though super-critical units are in-design. The thermal efficiency of an optimised CFBC boiler should be the same as a PF boiler. However, greater air fan power requirement may lead to a unit efficiency about 0.4% points less than PF with FGD at the same steam conditions. Typically, operating plants remove 90-95% of the fuel sulphur. NOx emissions, without any NOx reduction technology, are generally three to seven times smaller than the emissions from PF boilers in the UK fitted with low-NOx burners. Pressurised Fluidised Bed Systems Pressurising a fluidised bed results in a large reduction in combustor size: thus, for example, the amount of machined steel used in current PFBC units is about one-quarter to one-half of that required for PF or atmospheric CFBC plant. The ABB Carbon PFBC uses a bubbling bed combustor with, typically, two stages of cyclone to remove most of the entrained particles from the combustor exhaust gas before it is admitted to a 'ruggedised' gas turbine. The gas turbine generator provides about 20% of the total plant output, the remainder coming from a steam turbine. Steam is raised by means of tube bundles submerged in the bed, by membrane-tube combustor walls and an economiser section in the gas turbine exhaust. Gaseous emissions are similar to atmospheric plant but sorbent utilisation is better. The ABB units have achieved sulphur capture efficiencies of 90-99%. NOx emissions at low load increase substantially due to increase in excess air. As with atmospheric units, the NOx emissions can be significantly reduced by injection of ammonia at combustor outlet, although at low load this needs to be combined with firing a small amount of additional fuel into the bed freeboard to maintain effectiveness. The ABB combustor design appears tolerant of coal quality, an inherent advantage of all the FBC technologies. PFBC systems based on circulating beds are being developed. Due to the high dust burden of the gas stream, these units will have a ceramic filter as the last stage of particle removal. While this allows the use of fairly standard un-ruggedised gas turbines, the filter represents a major technical risk at the proposed operating temperature of about 850C. Circulating bed systems are expected to show advantages over bubbling beds in the same areas as atmospheric plant (e.g. sorbent utilisation and low load operation) but the benefits may not be so extensive. One reason for this is that the contact time between solids and gas in pressurised bubbling beds can be made much longer than that for the atmospheric equivalent.

3.1.2

Pressure Part Design

Metal Temperatures For design purposes it is necessary to know the mid-wall temperature of boiler tubes, pipes and headers. For boiler tubes, the difference in temperature between the furnace gases on the outside of the tube and the water or steam flowing within the tube is determined by the summation of a number of temperature gradients through various components, as shown in Figure 3.1.3. In order to commence a metal temperature calculation it is necessary to determine the heat flux falling on the tube and the steam or water temperature within the tube. In most calculational methods, the internal surface of the tube is assumed to be clean. Typically, an iterative calculation is used to determine the inside wall temperature and this is then used, along with a knowledge of wall 8 of 108

thickness and the thermal conductivity of the chosen tube material to determine the mid wall temperature. For thick walled components such as headers and steam pipes, there is no significant heat flux falling on the outside of the component. Hence, mid wall metal temperatures are typically taken to be around 5C above the temperature of steam delivered to the turbine. However, significantly higher metal temperatures could develop under upset conditions, such as during starts, and these have to be accounted for in the design. Codes The design, manufacture and quality of pressure parts for boilers in the UK has typically been in accordance with BS1113, which in turn refers to other British Standards for specific requirements. Similar design codes have been used in other countries according to their needs. Hence, examples of Design Codes historically applied in Europe include TRD (Germany), AFNOR (France) and Stoonwezen (Holland). A draft European code, bringing together the requirements of the different member states, is currently awaited. Japan (JIS) and America (ASME) currently have their own Codes. The American ASME Codes tend also to be favoured in Asia. The purpose of the Design Codes is to apply appropriate materials parameters and design methods to produce an economical, safe and reliable system for the required duty. They have generally developed by taking into account past experience of type and quality of construction, ranges of operating conditions and applied loadings, variability in materials properties and possible modes of failure. Therefore, the design methods have been derived, and the allowable design stress limitations have been specified and modified as necessary, over many years. Materials Properties A table of allowable design stresses, for a number of approved materials, operating over a range of temperatures, is given in BS1113. The allowable design stresses are related to temperature in the proof range and both temperature and time in the creep rupture range as follows.

Material

Design Stress Proof Range Creep Rupture SRt 1 .3 SRt 1 .3

carbon manganese and low alloy steels

Re T 1 .5 Re T 1.35

austenitic steels

Where: ReT SRt = the minimum 0.2% proof stress for carbon, carbon manganese and low alloy steels or the 0.1% proof stress for austenitic steels. the mean value of the stress to produce rupture in time t.

Figure 3.1.4 shows the 0.2% proof stress and creep rupture stress for 100,000 hours as a function of operating temperature for a plain carbon boiler steel and the corresponding allowable stresses for design. Hence, it is clear that the allowable design stresses effectively incorporate a safety factor to accommodate for, e.g. reduction in strength due to welding, variability in materials 9 of 108

properties, etc. Below the crossover point in the curve, the design stress is considered to be independent of time, although creep can still occur to a limited extent and may lead to failures in particularly highly stressed components at, for instance, structural discontinuities. Code scantling calculation allows stresses to be greater than design at such discontinuities with, for instance, a stress concentration factor of up to 2.25 being allowed for at branch openings to drums or headers. No allowances are included for thermal stresses, weld residual stresses or stresses induced during boiler erection. All pressure parts operating at temperatures above the crossover point are subjected to creep deformation and likely to fail in creep rupture. Consequently, all such components are designed for a specified minimum life. The life will be dependent upon the creep stress to produce creep rupture failure within a given time at a given temperature. Data is usually gathered from long-term creep rupture tests in the laboratory and all such data typically show scatter of around 20% about the mean. Hence, the factor of 1.3 used to derive the design stress from mean creep rupture data allows a small safety margin on minimum data. Component Dimensions Rules and design equations are given in the codes to calculate the required thicknesses of pressure parts within the boiler, including drums, headers, tubes and pipes. For plain, thick walled, cylindrical components such as boiler tubes, the principle life limiting stress is normally given by the hoop stress: t= PD 2f + P

where: t P D f = = = = wall thickness internal pressure outside diameter, and allowable hoop stress.

For headers and drums, a ligament efficiency () term is included in the above equation to take account of the tube penetrations through the wall as: t= PD 2 f + P

Various formulae for are given in the design codes for different layouts and sizes of the tube penetrations. Further rules are given to account for additional stresses due to, e.g., self weight and weight of contents, whilst local loads at supports are given in BS5500. The methods for calculating additional loads due to thermal expansion and construction mismatch are covered in BS806. Manufacture and Assembly As well as considering the materials used, BS1113 describes the methods of fabrication, manufacturing tolerances, welding, stress relief and pre- and post-welding heat treatment requirements. Limits are placed on the acceptable levels of wall thinning and ovality which arise from tube and pipe bending. Special attention is drawn to austenitic to ferritic transition joints, which are always made under workshop conditions and located at butt welds clear of geometrical discontinuities.

10 of 108

Modern power stations use a range of materials from carbon steel through to highly alloyed austenitic stainless steels. Component sizes range from a few mm to over 100mm. A number of welding processes is available and required to cover this range of materials and sizes. The current route for approval of weld procedures requires the manufacturer to demonstrate that the chosen procedure is technically correct and that sound welds can be produced to the acceptance standard. Further, welder qualification tests are carried out to show that the welder is competent to carry out the weld procedure. These are generally contractual requirements during power plant construction and repair, but must satisfy the statutory approval process carried out by the inspecting and insurance authorities. Inspection and Testing Inspection during construction is required to ensure that materials, design, construction and testing complies with the standard codes. This is normally carried out by an Inspecting Authority who is appointed by the insurers. The inspection will also include approval of welding procedures, inspection of weld approval tests and welder's certificates. Hydraulic testing is carried out to demonstrate the strength and integrity of individual components and the completed boiler.

3.1.3

Operating Environments and Failure Modes: PF-fired Plant

Steam Generating Walls Furnace Wall Corrosion Furnace wall corrosion has been a well-established and reported phenomenon since the 1940's. It has been responsible for numerous forced outages in most UK coal-fired plant, incurring considerable costs; not least those associated with lost generation. Numerous examples of plant suffering from furnace wall corrosion have also been found throughout the world4. Enhanced wastage is generally found on the radiant face of the furnace wall tubes, over localised sections of the furnace wall, up to 3m in height and extending typically over 60 tubes wide. The wastage patterns are highly dependent upon the firing configuration. For front wall fired boilers, wastage typically occurs on the side or rear walls, although simultaneous attack can occur in both locations where the furnace chamber is relatively small for its rated capacity. On corner fired units, wastage is often confined to the front wall, where the flame vortex is deflected forward by the nose on the rear wall5. In areas of persistent attack, gas analyses have revealed virtually zero oxygen partial pressures, >0.5% CO and uncombusted matter accounting for up to 50% of the solids recovered6. Whilst strictly the result of sub-stoichiometric combustion associated with impingement of the flame envelope, such environments are frequently referred to in the published literature as reducing conditions. Recently research has shown that negligible corrosion occurs under predominantly oxidising local flue gas conditions7. The corrosion rates are found to be dependent only upon metal temperature, with no effects of coal composition or heat flux being discernible. Under reducing conditions, the corrosion mechanism shifts progressively from oxidation towards sulphidation as the CO content of the local flue gas increases. There is no measurable effect of heat flux on the corrosion rates in the absence of chlorine, but increasing the CO content of the local flue gas increases the corrosion rates via the formation of progressively more sulphide within the corrosion scale. In the presence of even small quantities (<0.1%) of chlorine in the coal burnt, metal chlorination, via the HCl/FeCl2 mechanism, comes into play and dominates the overall metal loss. Linear reaction kinetics ensue and these are strongly influenced by a close interrelationship between coal chlorine content and heat flux7. Low NOx Technology 11 of 108

Environmental legislation in the UK has demanded significant reductions in emissions from coal-fired power plant. The principal route towards meeting this requirement in existing UK plant has been the retrofitting of low NOx burners. The initial low NOx systems were fitted in the late 1970's on corner-fired Units firing coals with a weighted mean chlorine content of approximately 0.3% and which were experiencing corrosion rates in excess of 600nm/h on the front wall5. At the first survey after the low NOx installation, and in the absence of any significant changes in operating regime or fuel chemistry, marked reductions in corrosion rate were found. Similar results have been reported on a corner fired super-critical boiler where the secondary air has been offset to effect a partial low NOx system, although this improvement (1000nm/h to 100nm/h) coincided with a reduction in load factor. Based upon current understanding of fireside corrosion, the beneficial impact of low NOx retrofits on furnace wall corrosion has been attributed to: o containing the flame vortex within an oxygen excess blanket, mitigating the opportunity for furnace wall tubes coming into contact with hostile environments, o promoting a more diffuse heat pick-up in the burner zone, reducing the incident heat fluxes to which tubes are exposed.

Early experience with low NOx burner installations in front-fired boilers proved less encouraging, with the detection of high CO levels on the rear walls and evidence of increased metal wastage on tube samples taken between surveys. Resetting the angle of secondary air vanes to promote greater swirl has subsequently rectified this problem by shortening the flame length and these sites have now reported reduced corrosion rates in prior corrosion-affected areas of the furnace wall. Burner retrofits at other Stations firing coals with a broad range of fuel chlorine contents have yielded comparable improvements in wastage rates. The retrofitting of deeply staged low-NOx combustion systems in the USA has led to an upsurge in furnace wall corrosion problems with metal losses as high as 2mm per year being reported8. Super-critical units are generally more severely affected than sub-critical units and corrosion is generally limited to coals with more than 1%S. The highest corrosion rates have been reported in that part of the furnace where H2S-rich sub-stoichiometric flue gas mixes with air from the overfire air ports. It is postulated that FeS-rich ash deposits out of the reducing flue gas onto the tube surfaces. Because of the turbulent flow in boilers, these FeS-rich particles may be carried up and deposited on boiler tubes which are normally exposed to oxidising conditions. Under these oxidising, or mixed oxidising/reducing, conditions these FeS deposits are thought to be oxidised, creating a highly sulphidising environment immediately adjacent to the tube surface. Waterside Corrosion The primary reaction involved in the waterside corrosion of high temperature boilers9 is the formation of magnetite, via the intermediary formation of ferrous hydroxide: Fe2+ + 2OH- = Fe(OH)2 3Fe(OH)2 = Fe3O4 + 2H2O + H2. In approximately neutral conditions, a magnetite scale forms on the tube, the growth of which is dominated by solid state diffusion through the scale. In the absence of sufficient hydrogen, the solubility of magnetite in the boiler water is markedly temperature dependent and the corrosion rates of steels approach minimum (acceptable) levels around neutral pH values4-9. Excursions from neutral values lead to increasingly severe attack. In commercial drum re-circulating boilers, significant transport of corrosion products may occur in the water circuit and these tend to be deposited in areas of high heat flux. Hence, a porous overlay of magnetite crystals, often incorporating nickel or copper derived from the feed system, 12 of 108

generally develops on the outer surface of the as-grown magnetite scale. The overall rate of accumulation of this deposit follows approximately linear growth kinetics and, under normal circumstances, the rate of magnetite accumulation typically ranges from 1 to 2m/1,000 hours. However, under the decidedly more flexible operating regimes now practised in the UK following privatisation, higher rates of magnetite growth and deposition (circa 4m/1,000 hours) are becoming more common. In normal boiling conditions, steam generated at the tube surface is immediately replenished by adjacent liquid and no appreciable concentration of solutes is experienced. However, in areas of local dry-out, under thick porous deposits and in crevices, significant accumulation of solutes may occur. Concentration factors of up to five orders of magnitude may be generated, leading to severe on-load corrosion. Hence, common boiler water practice is to minimise corrosion, deposition and carryover of boiler water solutes in the steam. This is often achieved by adding hydrazine to the water to scavenge the oxygen and by removing some of the boiler water on a regular basis by blow-down. Further, regular acid cleaning of the boiler is often carried out to limit the total oxide (both grown and deposited) thickness. In once-through units, any dissolved salts present in the boiler water remain in the system10. Hence, the concentration of contaminating salts in the feedwater must be maintained at a very low level to ensure very low boiler water conductivity. A water treatment regime gaining in popularity is the neutral, oxygenated, low conductivity (NOLC) regime. Significant oxygen is deliberately added to the water, either directly or as hydrogen peroxide, which results in the formation of very protective oxide films (magnetite/haematite) on the tube bore. Further, maintaining very low conductivity reduces the deposition which might otherwise lead to an increase in tube wall temperature with time. In order to prevent the risk of hydrogen induced stress corrosion cracking of ferritic steels from the internal fluid, the hardness in the heat affected zones of welds in boiler tubes should not exceed 350 Hv1011. Further, austenitic steels are susceptible to stress corrosion cracking when exposed to typical boiler water chemistries in high temperature plant. Final Superheater/Reheater Banks Superheater/Reheater Corrosion Superheater and reheater tubes may fail well inside their intended design life by a combination of fireside corrosion and creep5. Excessive corrosion attack of the leading tubes in a pendant stage is generally characterised by the formation of wastage flats on the tube surface at approximately the 2 and 10 o'clock positions in relation to the gas flow. This is associated with the preferential accumulation of ash debris on the radiant crown, sculpted by the gas flow around the tube. The accumulated debris thermally insulates the crown section from the flue gas, displacing the hottest operating section of the tube surface towards the 2 and 10 o'clock positions where the deposits are thinner and, consequently, the transmitted heat flux is greatest. Metal loss on tubes deeper within the bank is invariably less pronounced, by virtue of their more sheltered position, and any wastage flats that form tend to be located at the 3 and 9 o'clock positions relative to the gas flow. The fundamental mechanism of high temperature fireside corrosion is well established, being first proposed in the mid 1940's12 and it is quite distinct from that operating on furnace walls, involving molten sulphatic phases derived from deposited ash. These molten sulphatic phases, specifically those of alkali metals, exacerbate metal loss by a combination of direct sulphidation and scale fluxing to form alkali metal tri-sulphates. The primary factors which dictate the rate of metal loss, either in isolation or synergistically, have been identified as tube metal temperature, incident heat flux (determined by the gas 13 of 108

temperature (Figure 3.1.5), and fuel chemistry, together with the tube material and the position and alignment of the tube within an element. The influence of tube metal temperature on molten salt corrosion was characterised by classic work in the late 1950's and early 1960's13,14. At relatively low tube metal operating temperatures, typically below 550C, the ash deposits in contact with the tube substrate and/or corrosion scale are present as a porous solid layer. These deposits allow relatively free access of the bulk flue gas to the metal substrate/scale and the corrosion proceeds according to the temperature dependent rate for the gas phase oxidation-sulphidation of the metal. As the tube metal temperature increases, sodium and potassium sulphates, formed earlier in the furnace section, gradually accumulate on the tube surface facing the gas flow by condensation at the base of the porous alumino-silicate deposit. As the temperature increases through the melting range of the deposit, the chemical activity of the melt is heightened, causing a rapid increase in the corrosion rate. However, as the temperature increases, the stability of iron sulphate/irontrisulphate under the prevailing SO3 partial pressure decreases. Eventually, a temperature is reached at which the instability of sulphated iron species becomes dominant, causing the rate of attack to diminish back to that associated with corrosion by gaseous species in the bulk flue gas. The dominance of melt activity, or corrosion product instability with temperature, gives rise to the classic 'bell shaped' temperature dependence of corrosion under molten sulphates which starts at 600C and reaches a peak around 670C10. Low NOx Technology The retrofitting of staged low-NOx combustion systems in the USA has raised concern regarding carburisation of austenitic stainless steel superheater and reheater tubes8. The C-content of the fly ash is increased which may lead to locally reducing conditions in the superheater and reheater areas, leading to carburisation of the tubes. Measurement of the increase in carbon content of superheater steels in boilers fitted with staged low-NOx burners has suggested that 6% of the chromium content of the steel can be tied up as carbides15. In addition, delayed combustion of the fuel may result in higher metal temperatures of the superheater and reheater tubes which could also lead to an increase in corrosion rates16. Steamside Oxidation The exposure of all classes of steels to high temperature steam leads to the formation of an oxide scale, accompanied by some metal loss. For plain carbon steels, the oxide scale which forms is largely magnetite (Fe3O4) which offers some protection to the underlying substrate as it thickens. Hence the incremental increase in scale thickness declines with time, with total scale thickness following a parabolic relationship. The underlying metal loss typically equates to around one half of the overlying scale thickness. For low alloy steels, a duplex scale, comprising an inner inward growing spinel phase, overlying a magnetite layer of similar thickness, is formed. With increasing chromium content in the steel (>9%Cr), the first scale to form may comprise a highly protective chromium sesquioxide (Cr2O3), the rate of thickening of which may be several orders of magnitude lower than that found on the lower alloy steels. However, the highly protective nature of this scale may eventually break down, leading to the more rapid growth of the more familiar duplex scales seen on lower alloy steels (Figure 3.1.6). The tendency for this scale reversion to occur can be inhibited by increasing the rate of chromium supply to the growing oxide scale. This may be facilitated by increasing the chromium content of the steel, chromising the surface or by enhancing chromium diffusion in the substrate by surface cold working or significantly reducing the grain size of the steel. For instance, shot blasting the bore is claimed to improve the steam oxidation resistance of austenitic boiler tubes by about an order of magnitude10. For most conventional PF-fired plant, metal losses due to steamside oxidation are relatively minor and scant attention is taken of the effects of oxidation on, e.g. the creep rupture life of the components. Figure 3.1.717 shows that the anticipated scale thickness on the bore of 1 to 2%Cr 14 of 108

steels after 200,000 hour exposure at 560C is only anticipated to be around 450m, which equates to a metal loss of approximately 0.25mm. This equates to a loss of around 4% of the wall thickness of a 6mm thick superheater tube and 0.3 to 0.4% of that of a header or main steam pipe. However, for higher temperature plant, the growth of oxide on the bore of superheater tubing progressively insulates the tube and restricts the heat flux through it. This can increase the midwall metal temperature of the tube (Figure 3.1.8), resulting in higher rates of both steamside and fireside corrosion and shorter creep rupture life11. In addition, the issue of scale spallation from the bore of low alloy steel tubing and the subsequent solid particle erosion of turbines has reportedly been a significant problem in the USA8. Increasing the volume of steam grown oxide in the bore of the tubes is likely to exacerbate this problem. Headers/Pipes/Steam Separators Thermal Fatigue Fatigue due to cyclic loading can occur at local stress concentrations and thermal stresses due to temperature differentials. The latter is of particular relevance to thick walled vessels which may see rapid changes of temperature due to start-up and shut down cycles. Hot water or steam entering a cold, thick walled vessel can lead to steep temperature gradients through the wall thickness. Expansion of the inner regions of the wall, close to the bore, may be constrained by the outer, cooler material. This imposes compressive stresses near the bore which may be sufficient to cause local yielding. As the temperature gradients equalise, and the outer regions of the wall expand, the material at the bore may then go into tension. With continued operation at temperature the residual tensile stress relaxes until, on cooling, the process is reversed. Principal factors in limiting the thermal fatigue life of plant components are the material properties (thermal conductivity, coefficient of expansion, mechanical properties), plant start-up and shut-down rates and the presence of stress concentrating factors, such as tube penetrations into a header. Ferritic steels generally have higher thermal conductivities and lower thermal expansion coefficients than austenitic steels and are thus less prone to suffer from thermal fatigue (Figures 3.1.9 & 3.1.10)18. Problems Associated with Welding of Pressure Parts Fusion welding is by far the most important process used in the fabrication of modern boilers. Components are joined by the formation of a molten pool of metal between them. The production of high quality welds with a high degree of consistency is readily achievable. However, defects are more likely to occur in welds than in wrought material. Whilst there are several modes of weld failure, these are generally associated with inadequacies in weld procedure or practice and should readily be detectable by non-destructive testing during manufacture. Nontheless, welding of pressure parts can lead to other failure modes that only reveal themselves with time in service. Examples of such problems can arise due to stress corrosion cracking of weldments in low alloy steels exposed to boiler water and Type IV cracking in thick walled components. The need to limit the hardness in steam generator tube welds to less than 350Hv10 has thus far limited the choice of materials to low alloy steels such as plain carbon steel, 15Mo3 and T11. The well known T22 steel cannot be used since this, and more highly alloyed steels, require post weld heat-treatment (PWHT) to reduce the hardness in the heat-affected zones of welds. PWHT presents no problems for the shop manufacture of membrane wall panels. It does, however, pose great difficulties during erection and repair in the boiler. Type IV cracking typically arises in the extremities of the weld and spreads into the inter-critically annealed regions. It manifests itself late in component life (typically after 35,000 hours in _%Cr_%Mo_%V steel hot reheat pipework) and propagates rapidly. Cracking appears to be a direct consequence of exhaustion of creep ductility in the soft Type IV zone at the extremity of the weld HAZ19 in these materials. This leads to a stress reduction factor of 20% and a cyclic life reduction factor of approximately two under low cycle (i.e. thermal) fatigue. 15 of 108

3.1.4

Operating Environments and Failure Modes: FBC Plant

Atmospheric Pressure Fluidised Beds Corrosion in Bubbling Beds In view of the low combustion temperature (< 900C) and the capture of sulphur by added sorbents such as limestone or dolomite, it was originally felt that there would be few, if any fireside corrosion of in-bed tubes in AFBCs. However, early pilot plants demonstrated that the atmosphere in the bed could be aggressively sulphidising in the combustor if limestone or dolomite was present20. Enhanced sulphidation/oxidation was even found to occur in the absence of limestone if the coal had a high, naturally occurring, CaO content21. In an AFBC burning a range of different coals in a bed of sand with no limestone added the Coal Research Establishment identified that negligible corrosion occurred on the cooler (up to 450C) in-bed tubes, even when burning UK coals with up to 0.73% Cl22. For the hotter tubes (550 and 650C) the extent of corrosion was found to vary markedly with coal type. Only slight oxidation occurred when burning a low Cl, Illinois coal but corrosion was accelerated by mixed oxidation/sulphidation, though not to unacceptable levels, when burning high Cl, UK coals. In all cases, corrosion was greater on the in-bed tubes than those tubes exposed to the same temperature above the bed. However, the most severe corrosion was observed when burning coals with higher than average sodium contents when scale fluxing, analogous to the fireside corrosion experienced in the superheaters and reheaters of PF boilers, was observed. Corrosion in Circulating Beds High temperature corrosion, separate from mechanical wastage, was not previously reported as a problem in CFBC's23. Ahlstrom Pyropower use a straightforward 1%Cr 0.5%Mo steel for most of their superheater tubing, reverting to standard austenitic boiler tube steels when a higher creep strength is required (i.e. in reheaters). However, with CFB units now being used to fire high chlorine-content fuels such as agrobiofuels, straws, grasses and domestic waste, the high temperature chlorine-enhanced corrosion of superheater surfaces is becoming increasingly commonplace. Biofuel combustion results in the formation of gaseous alkaline sulphates and chlorides. At metal temperatures in the range 300-550C and flue gas temperatures of approximately 750C, such as those found in the superheater, any alkali metal chlorides carried over from the combustion process are likely to partially melt and adhere to heat exchanger tube surfaces. Deposits on superheater tubing in a CFB boiler in Sweden were found to contain both sodium and potassium chlorides24. Cl, Na and K in biomass fuel have resulted in corrosion beneath deposits on tube surfaces and subsequently accelerated erosion-corrosion wastage of the furnace wall surface25. The effects of high-temperature chloride corrosion have been reduced by the application of protective coatings at particularly exposed points, operational optimisations i.e. limiting temperatures in the superheater to below that of the melting points of the alkali chlorides, and the development of new superheater materials. At Gren, evaporator wings have been retrofitted to the combustor to aid temperature control, and external fluid-bed heat exchangers account for final superheating from 470-505C. The intermediate superheaters in the convective pass are operated at safe steam temperatures of less than 450C to prevent chlorine-induced corrosion. The highest loaded banks in each section are designed for periodic replacement and the hottest 16 of 108

convective heat surfaces are designed for excessive fouling by deposits with low ash melting points26. Erosion in Bubbling Beds Sethi and Wright27 have made the general observation that steel tubes operating at temperatures of above ~400C tend to from a hard, thick and adherent oxide scale that leads to relatively little wastage by erosion in the bed of an AFBC. However, at lower temperatures, the steel cannot form a protective oxide scale and severe erosion can be experienced. Identical units burning different feedstocks have shown markedly different responses to erosive wear23 but there is no clear understanding of what factors are involved. Bakker, Liebhard and Brekke28 point out that correlations between bed-material properties and erosivity are none existent. In terms of particle dynamics, Sethi and Wright27 have listed a number of potential causes of erosion in FBC systems. These are: the general flow of particles past the tubes, the presence of localised in-bed jets arising, for instance, near re-injection nozzles and close to bubble caps, the presence of long range flow patterns such as the general downflow of particles near combustor walls and the net upflow at the centre of the combustor, local flow regimes which arise as a result of geometrical irregularities, fast-moving particles ejected by bursting bubbles in the splash zone just above the bed, fast moving particles in bubble wakes, bubbles tracking along vertical or inclined tubes which can travel faster than those in the general bed, intrinsically fast particles, at the top end of the range of general particle velocities, which may be travelling at up to five times the superficial velocity. blocks of particles impacting onto tubes due to the presence of pressure pulses, in the frequency range of 0.5 to 5 Hz, within the bed.

Given such complex particle dynamics, modelling of the local particle velocities and flow direction in different areas of different bed designs has proven impractical. Consequently, much of the knowledge of in-bed erosion problems has accumulated with experience and many of the ameliorative techniques have been developed on a trial and error basis. Erosion in Circulating Beds As with bubbling FBCs, wastage processes within the combustion chamber of circulating bed FBCs are dictated by the particle flow patterns. These are dominated by the general flow pattern, with additional flows being induced by gas injection, solids injection, bed recycling and other design features, which may induce turbulence. Wastage rates are also thought to be a function of feedstock with, for instance, increased sodium in the ash increasing the erosivity. However, again, there is no clear understanding of what the dominant factors are. 17 of 108

The distribution of solids at the distributor plate has proven to be a significant problem. This distribution is non-uniform, owing to various design factors such as solids feeds and recycle injection ports. Any asymmetry in flow may lead to enhanced erosion rates and this is likely to become a bigger problem as unit size increases. Injector nozzles in this area are subject to heavy erosion and require frequent replacement. The bulk of the particle flow within the combustion chamber takes place downwards around the walls. As the particles fall, they tumble and rotate. This particle motion can lead to severe wastage at any discontinuity or disturbance to the flow pattern. The termination of the refractory coating on the waterwall tubes acts as such a discontinuity and appears to be the area in CFBC's most consistently affected by severe erosion. Following the patterns of behaviour adopted to deal with erosion of in-bed tubes in bubbling beds, the operators of the CFBC's have tried a number of essentially empirical 'fixes' to alleviate erosion at the refractory termination of existing plant. These include the attachment of shelves, plates or scalloped bars, to disrupt the downward flow pattern at the walls29,30, modification of the shape of the refractory at the termination, the use of coatings and weld overlays25,30,31 and redesign of the tube profile by the use of kick-out tubes32. In order to abstract as much heat as possible from the gas flow in the larger units, manufacturers such as Ahlstrom are now installing pendant tubes into the combustion chamber, prior to the cyclone. This exposes the tubes to a more erosive environment than they would otherwise see in the back passes. To alleviate this problem, Ahlstrom have introduced 'Omega' tubes into the boiler. These tubes are designed to eliminate any unnecessary discontinuities in the gas path that would change the direction of upward or downward flowing particles. They are extruded as a standard product by Sumitomo Industries and come in a variety of standard sizes. They have flat sided extensions which, when welded together, form a rectangular panel of tubes with straight sides aligned parallel to the general particle flow. It is claimed that these tubes have shown no signs of erosion damage, other than at the base of the lowest tube. Even this is claimed to have been rectified by welding a flat plate of iron into the lower channel to develop a backpressure which deflects the particle flow away from the tube underside. Double Omega type tubes are used in the primary superheater of the 230MWe circulating fluidised bed (ACFBC) plant at Turow in Poland33. Pressurised Fluidised Bed Combustion Bubbling Beds The pressure vessel for the ALSTOM Power (formerly ABB) PFBC is constructed from low carbon steel. This pressure vessel contains a deep bubbling bed FBC operating at a bed temperature of 850C. Normally, the tube bank is totally immersed in the bed and comprises the evaporator, superheater and reheater (if required). The boiler chamber is of refractory coated, membrane wall construction; the steamside is conventional with 600C metal temperatures. Circulating beds Ahlstrom Pyropower describe their pressurised circulating FBCs as being of identical construction to their atmospheric CFBC's. The water walls of the combustion chamber are of membrane construction with the lower section being refractory lined. The fluidising velocity is 5m.sec-1 with a bed temperature of 870C. The superheater and reheater are located high in the furnace and are constructed from Omega tubes linked together to form a flat plate. Minimum erosion has been experienced on the side faces since they are aligned parallel to particle flow.

18 of 108

With the exception of hot gas filter materials, there are no materials problems specific to PFBCs which have not been experienced in atmospheric FBCs, Erosion- and Corrosion-Resistant Coatings Protective coatings are seeing increased usage as plant operators seek to reduce the number of forced or planned outages necessary to repair or replace components susceptible to erosion and/or corrosion by extending the components life. Indeed, a flame spray coating of Al/80Ni20Cr now has a lifetime of at least 3 years when applied to waterwall tubes30. A number of coatings have been used in both operational and pilot plant, and a number have undergone, or are currently undergoing, laboratory testing, with variable performance in all cases. The choice of erosion-corrosion resistant coatings should be specific to each individual application. Often a coating will perform well in one FBC environment, only to fail in another. This is because the erosion-corrosion behaviour of a coating is not only related to the morphology of the coating, but also to the characteristics of the erodent particles34. Weld overlay was initially the preferred protective coating, typically forming layers 3mm thick using 1.85% Cr hardfacing material25. Durable layers of Alloy 625 have been built up to approximately 7mm thick on waterwall tubes in the USA30. However the application of weld overlay is restricted due to the low productivity of the process, erosion of the resulting ledge and thermal shock 35. Repeated applications at successive outages leads to embrittlement of the old overlay and to cracks that could propagate into the tube proper36. This had led to more attention being focused on the application of thermal spray coatings. A range of materials and application techniques (arc spraying, flame spraying, high velocity oxygen fuel spraying [HVOF], sintering) are available37. A variety of alloys have been employed in efforts to improve on parameters such as the adhesion strength of the coating to the base material, resistance to spalling, hardness and oxidation in the bonding zone, heat transferability and resistance of the coating to corrodent penetration37. It is important to recognise that the coating performance is sensitive to surface preparation and post-processing, with the elimination of surface irregularities, and the critical nature of the contour of the coating from its bottom to top edge. The coating should also be applicable on site at a reasonable cost25,38,39. Refractories Refractories in the combustion chamber of CFBCs may experience rates of temperature change of the order of 550C in a matter of a few minutes. Refractories in the combustor roof and cyclone may experience even more damaging temperature excursions greater than 1100C. These temperature excursions can lead to cracking, spalling and subsequently enhanced erosion of refractories23. In their earlier plants, ABB-CE used to use SiC bricks to cover the wall tubes in the lower, bed area, of the combustion chamber. These were held in place by spigots attached to the boiler tubes. However, oxidation of the spigots led to their ultimate failure, with the SiC bricks becoming detached from the chamber walls. Thin refractory linings on studded water walls are now the norm for utility size CFBC's. These refractories comprise SiC plastic or gun mix applied over densely spaced studs and castable steel anchors. Repairs are carried out using phosphate-bonded plastic. The lower, sloping portion of the combustion chamber of ABB plant is studded and covered with a 25mm thick layer of plastic refractory, Alumina Paste. Following an internally-funded review of refractory experience in FBC plant, Powergen concluded that:

19 of 108

Many of the early refractory installations in FBCs suffered significant degradation in the first few months of operation as a result of failure of support structures, cracking and spalling due to thermal cycling or thermal shock and erosion by circulating solids. Most of these failures could be attributed to the installation of appropriate refractories using inappropriate attachment methods, including faulty gunning techniques or improper dry-out and curing procedures. Others were largely attributable to the failure to adopt operational procedures (e.g. control of heating and cooling rates) commensurate with refractory properties. Application of second-generation refractory technology, such as low cement castables in uncooled ducts, thin phos-bonded plastic on water-cooled walls and shop-fired bricks in cyclones, has largely been successful in combating refractory failures in modern CFBC plant. Refractory lives commensurate with total plant life are being predicted for many installations. The cyclone target area is subject to the most intense erosion within CFBC plant and is the one area where refractory lives of just two years are still anticipated.

However, recently published experience shows that problems with refractories still include fabrication quality control, mechanical design when anchoring refractory components to metals with different expansion coefficients and also the operating time required to heat up or cool down these materials40. Refractories within the barrel and cone of the cyclone see the most onerous conditions23. The efficiency of particle separation from the gas requires high centrifugal force which is dependent upon a high gas velocity and a tight turning angle. Erosion has proved to be a major problem on the cyclone target area which can be reduced by careful design. The gunned linings of cyclones on the first generation of CFBC's suffered from severe crack formation and movement of large sections adjacent to major openings and along the sidewalls41. Eventually, large lumps broke away and clogged the cyclones. Within the first twelve months many systems had to be completely replaced. Now, the only castable refractories are in the cyclone roof, inlet roof and outlet duct to the bypass. The nose section of the cyclone is typically lined with phosphate-bonded plastic. The refractory system on the barrel often comprises a multi-component lining of dense, high-alumina, abrasion resistant, fibre-reinforced, super-fireclay brick, (or 90% Alumina or SiC) overlaying calcium silicate insulating block adjacent to the barrel. The target zone of the cyclone may be covered with a silicon-oxynitride-bonded, silicon carbide. The thermal cracking of refractories in cyclones has been mitigated by recent improvements in design, utilising thinner layers of refractory. The thermal stress resistance of a thin refractory layer makes the cyclone more suitable for frequent turn-down operating conditions42. As an example, the water-cooled Kvaerner Cylindrical Multi Inlet Cyclone (CYMIC) uses less refractory than previous designs, leading to shorter start-up and shut-down periods and tolerance of wide load variations. Water-cooling means that light refractories can be used. Two commercial CYMIC boilers are in operation, the largest of which is 185MWth at Rauma, Finland. An Integral Cylindrical Cyclone and Loopseal (ICCL) assembly has been developed for larger units, with a 500t/h installation operating in the USA30,43. In the Foster Wheeler Compact CFBC design, a water- or steam-cooled centrifugal separator (a square cyclone) is used, with the entire separation unit fabricated with flat walls constructed from conventional water-cooled membrane panels. To counter erosion, the interior walls of the separator are covered with a thin layer of refractory held in place by a dense pattern of metal studs. This lining has apparently proven very durable. The cooled construction minimises differential expansion between the furnace and separator, thus minimising both the number and 20 of 108

the relative movement of expansion joints. A number of such plants are either operating or under construction30,44.

3.1.5

Flue Gas Desulphurisation

Flue Gas Desulphurisation (FGD) is widely used to control the emissions of SO2 and SO3 from PF-fired plant. A variety of processes are available, most of which use an alkali sorbent, such as limestone, to capture the sulphur compounds from the flue gas. The most widely used FGD systems comprise the limestone gypsum process, producing either saleable gypsum by-products, or disposable wastes such as sludges or mixed solids. The only FGD plants on large PF-fired power stations in the UK are fitted at Drax and Ratcliffeon-Soar. These both comprise the limestone wet scrubbing system and were both ordered in 1990. At the time, this system could be described as a mature technology, with many of the early corrosion and scaling problems having been cured by the application of suitable materials and coating regimes. These included the use of rubber, glass flake vinyl ester resins and high nickel alloys. These materials have performed very well so far, giving little impetus for further materials developments. However, some improvements have been made with slightly improved, or lower cost, alloys and the addition of mica, rather than glass flake, to the vinyl esters. The latter can withstand slightly higher flue gas temperatures. Further, scrubbers manufactured from glassreinforced plastic, which are cheaper than metal absorbers and offer better corrosion and weatherresistance and easier maintenance, are becoming available in increasingly large sizes. FGD recycle pumps are probably the main area of plant where materials improvements are still required. Rubber coatings on these pumps can suffer damage from contact with solids and the metal castings and impellers are susceptible to corrosion and cracking, particularly in high chlorine environments.

3.1.6

Component Integrity and Life Assessment

Operating conditions and environments represent an ongoing threat to component integrity in steam raising plant. There is no prospect of any utility being able to adopt a fit and forget philosophy within the design life of a boiler. Indeed, current life extension means that much of the PF-fired plant in the UK is approaching, or has already exceeded, twice the original design life. Further, with changes in commercial markets and regulation, existing plant is currently having to operate in a much less certain and much more flexible regime than its original design intent. Accordingly, components are increasingly being subjected to ever more onerous operating conditions under ever more aggressive environments. Hence, all utilities are obliged to develop, adopt and continually improve methods for assessing the integrity and future safe life of components in service. These life assessment methods incorporate the most up to date NDE techniques and metallographic assessment procedures to determine materials degradation in service. These are supplemented by extensive databases of materials properties, fracture mechanics, knowledge of microstructural evolution with time at temperature, measurement and analysis of corrosion mechanisms and finite element stress analysis, in order that risks to component integrity can be minimised.

3.1.7

State of the Art PF-fired Boilers

Table 1 lists the chemical compositions of the boiler materials referred to in this section. Only recent and near-term developments are discussed below.

21 of 108

Steam Generating Walls Current Without exception, the steam generator tubes in the furnace walls of supercritical plant are of membrane wall construction. In conventional supercritical plant (250bar, 540C) the maximum temperature of the water/steam fluid in the water-wall panels is 420C at the outlet. Because of the high heat fluxes in the furnace chamber, the mid-wall metal temperature is approximately 450C on entering service. Growth and deposition of magnetite on the bore of the tubes increases the temperature difference through the walls such that, at 100,000 hours, the estimated mid-wall temperature rises to 455C. Under such conditions 1%Cr 0.5%Mo steel has adequate mechanical properties for a 100,000 hour life. Stork Ketels use a 0.5%Cr steel in the evaporator wall tubes and claim an acceptable metal temperature at the furnace outlet of 460C. Temperatures of up to 500C can be tolerated for short periods (5 minutes) at low pressures during start-up. ABB favour the use of T11 or T22 steels and quote a maximum operating temperature of 538C. Then, C-Mn and lowalloy steels can continue to be used for main steam conditions of up to 580C and 290 bar. All of the boiler manufacturers offer some form of staged combustion to limit the NOx emissions. In all cases this implies sub-stoichiometric combustion in the lower portion of the furnace chamber. Such sub-stoichiometric conditions pose the danger of furnace wall fireside corrosion which can lead to very rapid thinning and, hence, early failure of the membrane wall tubes. ABB claim that they favour the use of T22 steel over the T11 steel because the increased chromium content offers improved oxidation resistance. However, chromium contents in excess of 25% are required before any real improvement in furnace wall corrosion rate is likely to be experienced. None of the materials proposed thus far possess sufficient corrosion resistance to withstand furnace wall fireside corrosion. For furnace wall applications, bimetallic boiler tubes, such as those produced by co-extrusion, are the only tubes likely to possess the required corrosion resistance. However, the use of an austenitic outer layer for furnace wall corrosion resistance would pose tremendous difficulties with differential thermal expansion in membrane walls. Consequently, all boiler manufacturers offer some means of ensuring more oxidising conditions close to the furnace walls. ABB, Steinmller and Stein Industrie all use corner fired units with, for instance, offset secondary air. Stork Ketels use a front wall fired system with the stoichiometry of the burners closest to the sidewalls set up to be more oxygen rich. 620C/325bar For USC plant of 325 bar and 620C, the maximum water/steam temperature at the outlet of the water walls is approximately 475C. This equates to a mid-wall temperature of 497C on entering service which climbs to 513C over 100,000 hours. The metal temperature of the outer surface may rise as high as 524C in general and 539C in the highest heat flux areas (burner zone). The mechanical properties of conventional boiler steels are no longer adequate for this duty and more highly alloyed, creep resistant, materials are required. Hence, UNIPEDE 45 consider that the main difficulty in creating high-grade supercritical steam conditions in the 600C/350 bar range is concerned with the steam generating tube materials available. It will only be possible to exceed 45% thermal efficiency if materials for construction of membrane walls can be found which can operate above 500C. Preliminary calculations based on the allowable stress data for T23 (Table 3.1.2, Figure 3.1.11) suggest that the maximum permissible water/steam temperature using T23 in the waterwalls of supercritical plant would be 480C. This should be sufficient for use in plant operating at 325bar, 620C3. 7CrMoVTiB10-10 has a slightly higher creep strength than T23 steel at temperatures up to 570C (Table 3.1.3, Figure 3.1.13), whilst above this temperature, T23 would be the preferred choice. However, the long term strength values for the latter may be somewhat optimistic as 22 of 108

they are derived from extrapolation of much shorter duration creep rupture data (~15,000 hours). Further, precipitation of W-bearing Laves phase is expected during exposure, resulting in some loss of solid solution hardening. By contrast, the data for 7CrMoVTiB10-10 steel extend out to nearly 100,000 hours, so that some confidence can be placed in the strength values. Both T23 and 7CrMoVTiB10-10 are readily welded, neither requiring preheat or post-weld heat treatment. Similarly, both steels are reported to have good fabrication characteristics46,47. However, whether either of the materials has the necessary resistance to steam oxidation for operation at such temperatures remains an open question. ALSTOM Power and Mitsui Babcock are currently considering the use of T23 for water wall applications in supercritical plant. Trial panels of T23 and T24 have been installed in Asnaesvaerket in Denmark and Cordemais in France. Preparations are in hand for further trials in Weisweiler, Neckar 2 and Altbach48. Vallourec and Mannesmann have just received their first commercial order for T23 for superheater tubes in a plant in Morocco. If the ongoing tests are successful then problems concerning the water-walls will be solved and it will be possible to operate with water/steam temperatures of 500 to 520C in the water-walls49. 650C/350bar For USC plant with steam conditions above 620C/325 bar, further materials improvements will be necessary to reach steam/water temperatures much above 480C in the water-walls. In view of thermal expansion limitations in membrane wall construction, the materials of choice invariably comprise ferritic steels. In view of this, Mitsubishi Heavy Industries are looking at further improving the creep strength of T23 for these applications by the addition of Re50. However, considerable attention is being directed towards the 9 and 12% chromium steels such as T91/P91, X20CrMoV121 and HCM1251. Notwithstanding the higher chromium contents, it should not be assumed that these steels will have adequate corrosion resistance and, as they are more expensive, their use will inevitably increase constructional costs. Full sized furnace wall panels were made from T91 and HCM12 by a number of manufacturers for the European COST 501 Round 2 Programme51 without reported difficulty. The panels were found to be completely defect-free when examined both visually and by NDT. However, no manufacturer managed to complete a panel fully in compliance with the German TRD code with respect to the hardness criterion of < 350 Hv10 and the weld not more than 150 Hv10 greater than the parent material, without resorting to post-weld heat treatment. Efforts are being made to establish whether reducing the carbon content of these steels will ensure hardnesses of less than the critical value of 350 Hv10 in welds without any reduction in the creep strength of the material11. However, Masuyama et al 26 expect HCM12 to be applicable to water walls because it has better weldability than T91 and a lower susceptibility to SCC. On this basis Franklin and Henry51 recognise that field tests and discussion with local authorities will be needed before either material can be used. As part of this ongoing development, test panels of P91 and HCM12 have been built into the water-walls of plant operating in Germany. Steam Separating Vessels For once-through, super-critical units, steam separator vessels are positioned at the steam generating wall outlet to separate the water from the saturated steam. The steam flows on to the superheaters whilst the water is returned to the boiler water feed train. The separators are large, vertical vessels, which have to be sized to accept large changes in the fluid level without dumping water. At between 40% and 100% of boiler load, the separators are running dry (containing steam) but at lower load, and during start-up and shut down, they pass through the transition from wet (water) to dry (steam)51. This mode of operation imposes severe thermal fatigue stresses on the materials of construction. 23 of 108

The steam separators for Steinmller plant are currently manufactured from a high strength carbon-manganese steel whilst Stein Industries favour the use of the 1%Cr _%Mo steel 13CrMo44. Although the steam separators of advanced super-critical plant could conceivably be manufactured from the above materials, the necessary increase in wall thickness would cause a degradation in operational flexibility, seriously increasing the start-up and shut-down times necessary to limit the through wall temperature transients. Hence, both Steinmller and Stein Industries consider that materials with a higher proof strength will be required for increasing steam conditions. However, materials choice does not necessarily limit boiler operation, since there is the option to use more separators of lower capacity, thus reducing the required wall thickness. Franklin and Henry51, for instance, claim that a large capacity boiler of approximately 800 MW would require 8 separators. P91 or X20CrMoV121 could be used to reduce the wall thickness and thus lower the thermal stress if required. However, long term testing has confirmed the reported tendency for failure in the soft Type IV zone at the extremity of the weld HAZ19 in these materials. Steinmller do not expect the material properties of P92 to be much better than P91 in this application. Whilst the thermal expansion of P91 and X20CrMoV121 are very similar, P91 shows a higher thermal conductivity11. Hence, P91 is expected to possess a greater tolerance to thermal cycling, with a lower tendency to thermal fatigue cracking. Since it is cheaper than the higher chromium X20CrMoV121 steel, the former is expected to be favoured as a steam separator material for more advanced steam conditions11. In terms of further materials development, HCM12 is claimed to demonstrate a low increase in hardness after welding, possesses a high creep strength and a relatively lower sensitivity to Type IV cracking than P91 or X20CrMoV121. It does, however, possess a relatively high delta ferrite content (up to 30%) which could pose toughness problems, particularly for thick walled components, after a long time in service51. Pipes manufactured from P122 are claimed to have a higher creep strength, better corrosion resistance and better weldability than those manufactured from P91 steel52. Superheater Tubes Current Superheater tubes must be designed to operate at temperatures some 35C above live steam temperature48. For current steam temperatures of up to around 580C, metal temperatures will be around 615C and low alloy steel tubes such as T22 may be adequate. However, not only do the advanced steam parameters for supercritical plant impose higher stresses and temperatures on the superheater tubes, they also increase the potential rates of both fireside and steam side corrosion. Fireside corrosion on the outer surface of superheater and reheater tubes can lead to rapid thinning and, hence, subsequent premature failure by creep. Experience within the CEGB in the 1970's, for plant with final steam temperatures of 565C and burning coal with chlorine contents greater than around 0.15%, showed that these low alloy steels did not possess sufficient fireside corrosion resistance. Consequently, the 500MW units required selective re-tubing with the austenitic steels T316 and T347. If the steam temperatures are to be only moderately increased, and the fireside corrosion potential is low, then the improved medium chromium ferritic/martensitic steels such as P91, P92 and P122 could be considered as possible alternatives to X20CrMoV121. In such circumstances, the steamside oxidation rate assumes increasing importance. Increasing the steam temperature leads to more rapid growth of oxide scales on the bore of the tube17. As the bore oxide scale thickens, the heat transfer to the steam inside the tube is reduced. Hence, the wall temperature of the tube progressively increases with increasing service life51. Such an increase in wall temperature not only leads to the more rapid accumulation of creep damage, it also results in higher fireside and 24 of 108

steamside corrosion rates. Thus, thinning of the boiler tube and increasing wall temperatures are likely to progress ever more rapidly under a self-accelerating process. Steam oxidation rates on 9Cr steels limit practical application to 600C. 620C/325bar For steam conditions of 325 bar at 620C the metal temperature in the final superheater will be around 660C49. Boiler tubes operating under these conditions will need a 105 hour creep rupture strength of about 100MPa and a fireside corrosion rate leading to a maximum metal loss of 2mm in 100,000 hours. Both P91 and P92 are unlikely to possess sufficient fireside corrosion resistance in this application. However, it is hoped that these new, higher creep strength, ferritic steels will limit the cost impact by their selective use in cooler parts of the circuit. Austenitic materials, such as X8CrNiMoNb1616, X8CrNiMoVNb1613 and X3CrNiMoNb1713 exhibit similar creep properties to the currently employed 12%Cr steels, such as X20CrMoV121, at a metal temperature some 70 to 80 C higher 51. Increasing the chromium content of these steels confers a greater resistance to fireside corrosion53. Even with the austenitic steels, the fireside corrosion rate has been found to be very sensitive to the chromium content below approximately 30% (Figure 3.1.13). A 20% chromium steel, for example, shows a fireside corrosion rate of approximately one third of that of a 15% chromium steel under similar conditions54. Whilst fine grained 347HFG and Super 304 both have the required creep rupture strength for operation up to 650C, the inadequate corrosion resistance conferred by chromium contents of around 18% may limit their operating temperatures to 615-620C51. The tubes will also need to be manufactured from a material showing minimal steam side oxidation at these temperatures. This will be necessary in order to limit overall oxide growth and any tendency toward magnetite exfoliation. For both NF709 and HR3C, the creep rupture data relate to tests of relatively short duration (circa 30,000 hours) and hence some caution should be exercised regarding the quoted strength values for operation to 100,000 hours. Also, whilst the steels are likely to have good corrosion and steam oxidation resistance, the final choice of material should be made on the basis of specific data, preferably obtained from tests on panels installed in operating plant. Currently, only data from short-term laboratory tests on material exposed to idealised environments are generally available, these being inadequate for proper evaluation of performance over times relevant to plant. 650C/350bar For tubes with metal temperatures approaching 700C, enhanced versions of austenitic steels such as NF709, AC66, HR3C and HR6W will be required to limit the bore oxide thickness48. Headers and Steam Pipes Current As headers and pipes are situated outside the furnace, fireside corrosion resistance is not a factor in the selection of materials. However, resistance to steam oxidation must still be considered. Since 1980, all new plants built for the ELSAM system have been super-critical49. Ferritic steels, from C-steel up to X20CrMoV121 (12%Cr), have been used for thick section components in plant designed for steam conditions of 250 bar and 540C, with steam lines being manufactured from X20CrMoV121. The same X20 steel is being used for thick section components in plant operating at the higher steam conditions of 250bar and 560C51. Materials with even higher creep strength

25 of 108

will be needed for thick section components under more advanced steam conditions and P91/T91/F91 are claimed to be suitable for such use up to 300 bar and 580C. Keeping the design parameters constant, Figure 3.1.14 shows how the wall thickness can be reduced as the creep properties of the steel increase45. 620C/325bar A temperature of 620C for steam delivered to the turbine implies metal temperatures of approximately 625C during normal operation. However, significantly higher metal temperatures could develop under upset conditions, such as during starts. Such temperatures pose a dilemma as regards the selection of materials, these being at the limit at which currently available ferritic steels can operate. Thus, above 600C the creep strengths of even the strongest currently available ferritic steels begin to fall off rapidly, with steam oxidation rates coincidentally showing a marked rise. Nevertheless, the possibility of using ferritic steels is attractive since they have a low coefficient of expansion and, hence, headers would be expected to have a low susceptibility to thermal fatigue cracking. However, the low creep strength at these temperatures would lead to vessels of relatively thick wall, raising the cost, and possibly limiting the cycling capability of the plant. The choice of materials for superheater headers, manifolds and pipework will depend to a large extent on the philosophy applied to the design of the boiler. P92 and P122 steels appear prime candidates for use where ferritic materials are required. However, welds in thick section headers and pipework are likely to be susceptible to Type IV cracking and allowable stresses will ultimately be limited by the need to account for this phenomenon. Similar consideration will also need to be paid to the performance of transition joints. After successful welding trials, P92 (as NF616) has been tested in a main steam line of the 250bar/560C unit at Esbjerg Unit 3. P92 and P122 were chosen for thick-section boiler components and steam lines for coal fired USC with 325bar 620C/630C steam conditions in an ELSAM design study. Subsequently, as part of COST 501 Round II, superheater outlet headers were manufactured in P92 and P122 and installed in one of ELSAM's 400MWe double-reheat USC plants with 290 bar 580C/580C/580C. Recently, P122 has been chosen by MHI for Tsuruga 2 (24.1 MPa 593C /593C) and Babcock Hitachi have selected P122 and P92 for Tachibanawan (24.1 566C/593C), both for commissioning in 2000. In Denmark, P92 was chosen for headers and steam lines in Avedoere 2. Table 3.1.8 lists trials of these advanced 9/12%Cr steels in plant during the COST 501 programme. ASTM and ASME approval of E911 as P93 will allow construction of thick-section boiler components and steam lines for USC plant with steam parameters of 610CC/325bar for main steam and 625C for hot reheat49. Since they possess higher thermal expansion coefficients and lower thermal conductivities than ferritic/martensitic steels, components manufactured from austenitic steels are expected to exhibit less favourable thermal cycling behaviour54. With regard to the choice of austenitic materials, the somewhat lower metal temperatures expected in headers and pipework mean that less expensive grades with lower chromium contents can be considered than for tubing. A grade suitable for the present application is reported to be X3CrNiMoN1713 steel (Table 3.1.1)55. The allowable stresses for this steel for 100,000 hours operation, derived from the published creep strength values56 are given in Table 3.1.9. The higher creep strengths of the austenitic steels implies that a significant reduction in wall thickness may be realised (Figure 3.1.15). As the required wall thickness of a header increases, the martensitic steels lose their advantage over austenitic steels in terms of allowable rate of temperature change (Figure 3.1.16). Then, at 600C for instance, the allowable rate of 26 of 108

temperature change for a header manufactured from P91 steel is only one half of that of one manufactured from the austenitic 1.4910 steel11. With suitable design, involving a large number of smaller, parallel steam paths and the use of header manifolds to reduce the number of weakening penetrations, then 1.4910 steel could be used at up to 620C and 350 bar or 650C and 250 bar (Figure 3.1.17)11. Four German power stations have decided to install 1.4910 on a large scale for the first time. Boilers planned or under construction include Units R and S in Lippendorf (2 x 800 MW) and Unit 4 at Boxberg (440MW). 650C/350bar For final steam conditions in excess of 620C, the new COST 522 programme will be looking to the development of the 12%Cr martensitic steels for higher strength and higher steamside oxidation resistance. In 1997 the National Research Institute for Metals (NRIM) in Japan started R&D for advanced ferritic steels for large diameter pipes and headers for USC plant with 650C and 350 bar57. This work has been directed towards the development of W-strengthened 9% and 12% Cr steels with long-term creep strengths higher than P92 and P122 and will comprise iridium additions to increase creep strength and silicon additions (especially with simultaneous addition of Ti or Y) to improve oxidation resistance. The even stronger NF12 (12%Cr-W-Co) is also under consideration, since extrapolation of short-term data implies that this steel may have a 105 hour creep rupture strength of around 100MPa at 650C.

3.1.8

State of the Art FBC

A recent development in CFBC is the Foster Wheeler Integrated Recycle Heat Exchanger (INTREXTM) design. This is a bubbling bed heat exchanger located in the lower part of the combustion chamber as part of the furnace, containing one or more tube bundles to cool circulating solids. The immersed tube bundles can replace a superheater in the furnace or backpass. As the INTREX heat exchanger is fluidised with air, the most corrosive elements in the gas stream do not come into contact with high temperature superheater tubes. INTREX units have been installed at a 30MWth plant in Finland and the 77MWth Gren plant in Sweden. Plants under construction using INTREX include a biofuel-fired 157MWth ACFB boiler in Vasteras and a 265MWe project at the Jacksonville Electric site in Florida30,44.

3.1.9

Future R&D Priorities

Much work has been carried out to develop new steels for advanced super-critical plant over the past 20 years. In Europe, where government funding is severely limited, the research work has had to be highly targeted to make the most of limited funding. Government funding for the COST Programme has been very patchy, with projects in Germany and Austria receiving substantial support, whilst that in the UK has previously had little or no financial support. As a consequence, most of the really significant advances in new materials appear to have been made in Japan, where substantial government funding continues to be made available. As a consequence, the Japanese steel makers are already claiming to have developed materials which will allow the manufacture of plant capable of operating up to 650C main steam. Nevertheless, steels with improved mechanical properties are becoming available in Europe. These have allowed a progressive increase in main steam temperatures, such that power plant with final steam temperatures of 600C can now be considered available. Further, main steam temperatures of 620C should be realisable in the near-term future. However, further work will be required to realise some of these goals, particularly with regard to achieving the higher main steam temperatures of up to 650C. 27 of 108

The creep rupture data reported in the published literature for most of these developmental steels extends to only about 45,000 hours, with the 105 hour creep rupture strength being derived by extrapolation. Data obtained from much longer exposures will be required before these 105 hour creep rupture strength values can confidently be used for boiler design purposes. In addition, the 9-12%Cr steels continue to display a 20% or so reduction in strength in the Type IV cracking zone on welding. Either this will need to be accounted for in design, or further work will be required in an attempt to minimise the problem. A joint PowerGen/Mitsui Babcock Programme is underway to examine the mechanical properties and weldability of E911. Perhaps of even greater concern is the fact that most of the fireside corrosion data provided by Japanese steel manufacturers is generally based on 24 to 100 hour exposures of the trial materials to an idealised molten salt mixture in laboratory studies. Whilst these data may be of value for ranking purposes, much more realistic fireside corrosion data, obtained from much longer exposures of the trial materials to relevant environments, will be required. Little attention has been paid to the fireside corrosion of candidate boiler tube steels in the earlier COST programmes, particularly the effects of fuel quality on corrosion rates. This is being rectified to some extent in the new COST 522 programme, where more realistic exposure to high temperature flue gases in a variety of operating plants is underway to elucidate the longer-term performance of some of the newer alloys. Similarly, little attention appears to have been paid to the steamside oxidation resistance of most of the steels. The Japanese have shown that silicon has an important influence on steamside oxidation rates for the 9 to 12% chromium steels (Figure 3.1.18)62, revisiting work carried out by the CEGB in the 1970's. However, the authors noted that increasing the Si-content of the steel above 0.4% was found to be detrimental to toughness. Given that oxidation rates will inevitably increase with increasing temperature, more work needs to be done on determining the best means to limit the steamside oxidation rates on all components in the system. The latest initiative in alloy development comprises computer-aided alloy design. Widespread application of computer programs such as "Thermo-Calc", "MT-Data" and "Dictra" have resulted in major improvements in the ability to understand and predict the equilibrium phase diagrams of multi-component systems, their thermodynamic and diffusion behaviour and diffusion controlled phase transformations, from alloy composition and heat treatment. Such methods, called the "Calphad approach", have been used to calculate alloy compositions for COST 522. The development of improved creep strength in alloys is often obtained at the expense of corrosion resistance. In order to achieve the ever increasing strength levels required it may be necessary to sacrifice some of the corrosion resistance of the base alloy. Then more effort will need to be directed towards the development of coatings or surface treatments that confer substantial resistance to both fireside corrosion and steamside oxidation if ever-increasing cycle temperatures are to be achieved in advanced plant. Developments in life assessment procedures are required to take into account the higher metal operating temperatures, the more arduous thermal cycles and the more aggressive combustion and steam environments that will be experienced in the more advanced boilers burning a wider range of coal and biomass mixes currently being contemplated. These procedures will need to include methods for condition-monitoring and life-assessment of protective coatings. The overall objective of the major European AD 700 project is to develop and demonstrate a new generation of PF plant with advanced steam conditions63. This will be achieved through the application of Ni-based superalloys to allow the use of steam temperatures up to 700C/375bar and an overall thermal efficiency of 55%, compared with the 47% efficiency of state-of-the-art 600C/300bar/300 bar double-reheat plant. This is expected to reduce fuel consumption and thus CO2 emissions of around 15%. Much larger components are required for steam cycle plant than in 28 of 108

gas turbines so these temperature increases represent a significant materials challenge. There will be a need to demonstrate both the manufacturing capability and innovative construction techniques. In terms of materials development, the major needs are seen as creep strength for longterm operation at high temperatures, corrosion resistance in flue gas and steam, thermal fatigue resistance and the ability of components to be manufactured and welded in thick sections. The future R&D requirements for boiler materials may be summarised as follows:1. T23 and 7CrMoVTiB1010 appear to be the most likely materials of choice for the waterwall tubes in super-critical plant operating up to 625C and 325 bar, but stronger materials will be required for higher steam conditions. The candidate materials at the most advanced stage of development at present are P92, P122 and E911, but all three currently require post-weld heat treatment during fabrication. Current materials are available for the manufacture of steam separating vessels for steam conditions of up to 625C and 325 bar. Where limitation of wall thickness criteria apply, stresses in the walls of these components may be reduced to acceptable levels by increasing the number of separators and reducing their duty. Stronger materials will be required for operation above 625C and 325 bar, with T91/P91 being a strongly favoured candidate. Austenitic stainless steels which possess adequate creep rupture strength and fireside and steamside corrosion resistance are available for use in the final superheater tubes of advanced PF-fired plant operating with steam parameters up to 290bar/580C, provided the inherent flue gas corrosivity (Cl content) is low. However, it is unlikely that the fireside corrosion resistance of these steels will be sufficient to operate much above 620C. More highly alloyed steels are under development which may allow operation at steam temperatures of up to 630C. However, more work is required to extend the creep rupture data for these steels out to longer times and longer-term fireside corrosion data, collected under more realistic conditions, will be required before these materials can be used with any degree of confidence. The recent ASTM/ASME approval P92 and P122 should allow construction of thick-section components and steam lines for advanced PF-fired plant operating with steam parameters of up to 325bar/610C. However, welds in thick-section headers and pipework are likely to be susceptible to Type IV cracking and allowable stresses will ultimately be limited by the need to account for this phenomenon. Similar consideration will also need to be paid to the performance of transition joints. The higher creep strength of the austenitic steels and the somewhat lower metal temperatures expected in headers and pipework mean that less expensive grades, such as X3CrNiMoN1713, with lower chromium contents, can be considered than for tubing. Solid Particle erosion presents by far the biggest threat to long-term component integrity in FBC plant. Designs are increasingly being adopted that minimise the likelihood of in-bed erosion. However, as the number of FBC plants burning biofuels continues to rise, so the problem of chlorine-enhanced erosion-corrosion becomes more of an issue. Heat exchangers are especially susceptible to the deposition of alkali metal chlorides and subsequent corrosion, particularly at the superheater. Thermal cracking of refractories continues to be an issue in FBC plant, particularly in the cyclone, and anchoring of refractory components to metals is still problematic. However, recent cyclone designs now utilise thinner layers of refractory, thus reducing the thermal stresses associated with thicker layers. 29 of 108

2.

3.

4.

5.

6.

7.

8.

9.

The development of ceramic tube hot gas filter systems (CTFs) for particulate removal is the most significant recent development in PFBC technology. However, developing and commercial ceramic filter systems are limited by process upsets and process equipment failures, and require improvements in life and temperature capability for use in PFBC. CTFs could limit the development of advanced PFBC plant, as they have not operated reliably at the elevated gas temperatures they would be exposed to.

10. Recycle pumps are probably the main area of FGD plant where materials improvements are still required. Rubber coatings on these pumps can suffer damage from contact with solids and the metal castings and impellers are susceptible to corrosion and cracking, particularly in high chlorine environments. 11. Ongoing requirements for improved materials for advanced boiler plant include: 1. References for Section 3.1 1. K. H. Mayer, W. Bendick, R-U Husemann, T. Kern and R. B. Scarlin (1998), "New Materials for Improving the Efficiency of Fossil-Fired Thermal Power Stations". VGB PowerTech (1998) No.1 pp 22-27. D. V. Thornton and K. H. Mayer (1999), in "Advanced Heat Resistant Steels for Power Generation" Eds. R. Viswanathan and J. W. Nutting, IOM Communications 1999 pp 349-365. R. Blum, (1994), Materials Developments for Power Plants with Advanced Steam Parameters: Utility Point of View in Materials for Advanced Power Engineering 1994 Proceedings of a Conference held in Lige, Belgium, 3-6 October 1994. Edited by Coutsouradis, D. et al. Kluwer Academic Press, (1994), pp15-30 B. Dooley. (1993); "Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report", EPRI Report TR-102433-V1. P. J. James and L. W. Pinder, (1997), "Effect of Coal Chlorine on the Fireside Corrosion of Boiler Furnace Wall and Superheater/Reheater Tubing". Materials at High Temperatures 14(3) (1997) 187. F. Clarke and C. W. Morris, (1983) "Combustion Aspects of Furnace Wall Corrosion" in Corrosion Resistant Materials for Coal Conversion Systems, edited by Meadowcroft, D.B. and Manning, M.I., Applied Science Publishers. C. J. Davis, P. J. James, L. W. Pinder and A. K. Mehta , (2001) "Effects of Fuel Composition and Combustion Parameters on Furnace Wall Fireside Corrosion in Pulverised Coal-Fired Boilers", Materials Science Forum, in Press. 30 of 108 further development, refinement and validation of microstructural modelling procedures, long-term creep rupture and thermal fatigue data, long-term fireside corrosion and steamside oxidation data in real environments, development of protective coatings, further development and validation of life-assessment procedures, including coatings.

2.

3.

4. 5.

6.

7.

8.

9.

10.

11. 12. 13. 14. 15.

16.

17. 18.

19.

20.

21. 22.

23.

24.

25. 26. 27.

R. Viswanathan and W. Bakker, (2001), "Materials for Ultrasupercritical Coal Power Plants - Boiler Materials: Part 1". Journal of Materials Engineering and Performance, Vol. 10, No 1, 2001, pp 81 to 95. D. G. Kingerley, (1994), Chapter 17.4 Boiler and Feedwater Treatment in Shreir, L.L., Jarman, R.A. and Burstein, G.T. Corrosion, Third Edition Butterworth Heinemann, London 1994. A. Fleming, R. V. Maskell, L. W. Buchanan and T. Wilson, (2000) "Material Developments for Supercritical Boilers and Pipework" in Materials for High Temperature Power Generation and Process Plant Applications Edited by Strang, A. Institute of Materials Publication Book No. 728. pp 32-77. A. Kather, G. Helermann, R-U Husmann, P. Hougaard and M. Knizia, (1994), "Steam Generators With Advanced Steam Parameters", Private Communication. W. T. Reid, (1971), "External Corrosion and Deposits; Boilers and Gas Turbines. American Elsevier Publishing Company Inc. New York, 1971. J. G. Koopman, E. M. Marselli, J. Jonakin and R. C. Ulmer,(1959); Proc. Amer. Pwr.Conf, 21, 236. W. Nelson and C. Cain (1960): Corrosion of Superheaters and Reheaters of Pulverised Coal Fired Boilers, J Eng for Power, 82, Ser A No 3, pp194-201, July 1960. D. N. French (1998), "Carburization Corrosion in Superheaters and Reheaters." Proceedings of the 3 rd International Conference on Boiler Tube Failures in Fossil Plants, EPRI Report TR-109938, 1998. W. T. Bakker, J. L. Blough and W. W. Seitz, (2001), Low-NOx Combustion Systems and Superheater Corrosion- Is there a Connection?" Paper 01168, NACE 2001, Houston Texas March 11-16 2001. Sumitomo Metal Industries Ltd (1989), "Steam Oxidation on Cr-Mo-Steel Tubes, Paper No. 805, 1443A. R. Blum, J. Hald, W. Bendick, A. Rosselet and J. C. Vaillant, (1994), "Neuentwicklungen Hochwarmfester Ferritisch-Martensitischer Sthle aus den USA, Japan und Europa" VGB Kraftwerkstechnik, Vol. 74, No. 8 (1994), pp 641-652. N. Naylor, P. Bontempi, G. Maciga and M. Livragi, (1994), Creep and Fatigue Characteristics of Advanced 9Cr Steel- Base and Welds, in Materials for Advanced Power Engineering 1994 Proceedings of a Conference held in Lige, Belgium, 3-6 October 1994. Edt. Coutsouradis, D. et al. Kluwer Academic Press, (1994), pp 341-350 B. Davidson, D. B. Meadowcroft, and J. Stringer, "High Temperature Alloy Requirements for Coal Fired Combined Cycles", CEGB TPRD Report No. TPRD/L/3123/R87, 1987. Report on EPRI Contract CS-3511, 1984. J. Oakey, A. Minchener and N. Hodges, (1991), "The Use of High Chlorine Coals in Industrial Boilers" in "Chlorine in Coal", Eds. Stringer J and Banerjee D B, Elsevier Science Publishers, pp 250-274, 1991. J. Stringer, J. Stallings and W. Bakker, (1991), "Materials' Issues in Circulating Fluidised Bed Combustion Systems", Conference on Fluidised Bed Combustion Systems, London, pp 39-50, December 1991. B-M. Steenari, O. Lindqvist and B. A. Andersson, (1999), "Preliminary Characterisation of Deposits Formed on Superheater Surfaces in an FBC Boiler Fired with Municipal Solid Waste", Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. N. G. Soloman, (1998), "Erosion Resistant Coatings for Fluidised Bed Boilers", Materials Performance, Vol. 37, No. 2, 1998, pp 38-43. I. Rasmussen, I. Hundebol, and L. Hogh Sorensen, (1997), "The ELASM Coal and Biomass CFB Power Plant Project", VGB Kraftwerkstechnik, Vol. 77, No. 6, p429-432. V. Sethi and I. Wright, (1993), "Materials Support for EPRI Fluidised Bed Combustion Program", EPRI Report TR-101804, Vol. 1, May 1993.

31 of 108

28.

29.

30.

31.

32. 33.

34.

35.

36.

37. 38. 39.

40.

41. 42.

43

44.

45.

W. Bakker, M. Liebhard and D. Brekke, (1993), "Erosion of Refractories in CFBC's: Effect of Bed Materials and Refractory Properties: A Laboratory Study", 12th ASME International Conference on Fluidized-Bed Combustion, 1993. S. Brain and E. Rodgers, (1987), "Experience of Erosion of Metal Surfaces in UK Fluid Bed Boilers" presented at EPRI/Argonne National Laboratory Workshop on Wastage of In-Bed Surfaces in Fluidized Bed Combustors, Argonne, Illinois, 1987. A. J. Minchener, P. J. I. Cross, T. N Smith and M. J. Fisher, (2000), "Fluidised Bed Combustion Systems for Power Generation and other Industrial Applications", CRE Group Limited, Report No. COAL R188, DTI/Pub URN 00/743. F. Belin, S. Kavidass, M. Maryamchik, D. J. Walker, A. K. Mandal and C. E. Price, (1999), "Update of Operating Experience of B&W IR-CFB Coal-Fired Boilers", Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. R. Zychal and S. Darling, (1990), in "Proceedings: Workshop on Materials' Issues in Circulating Fluidized-Bed Combustors", EPRI Report No. GS-6747, 1990. W. Z. Nowak, J. Laskawiec, W. Krzywoszynski and R. Walkowiak, (1999), "Design and Operation Experience of 230MWe CFB Boilers at Turow Power Plant in Poland", Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. B. Wang, (1997), "The Dependence of Erosion-Corrosion on Carbide/Metal Matrix Proportion for HVOF Cr3C2-NiCr Coatings", Proceedings of Surface Modification Technologies XI, Paris, Sept 1997 A. Verstak, B. Wang, V. Baranovski and A. Beliaev, (1998), "Composite Coatings for Elevated Temperature Erosion-Corrosion Protection in Fossil-Fuelled Boilers", Materials and Components in Fossil Energy Applications, No. 135, August 1 1998, pp 7-11. D. L. Pearce, (1999), "To the Advanced Coal Steering Group: The 6th International Conference on Circulating Fluidised Beds, Wrzburg August 22-27 1999", PTC Report No. PT/99/BA1350/M. P. Tarkpea, (1998), "Coatings as Protective Barriers in Fluidised Beds", Vaermeforsk Service AB, SE-101 53 Stockholm, Sweden, Report No. SVF-656. S. Arends, (1999), "Thermal Spraying Against Wear and Corrosion", Praktiker, Vol. 51, No. 11, 1999, pp 448-450. Y. Dutheillet and V. Prunier, (1998), "Evaluation of the Erosion-Corrosion Resistance of Coated Metallic Materials for CFBC's Application", Energy Technology, Vol. 5, No. 2, 1998, pp 779-787. J. M. Bursi, L. Lafanerchere and L. Jestin, (1999), "Basic Design Studies for a 600MWe CFB Boiler (270b, 2 x 600C)", Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. T. Elliot, (1994), "Shoot for More Durable, Less Costly Refractory in CFB Units", Power, February 1994, pp 69-71. Z. Junfu Ljiansheng, L. Qing, L. Xiaoma, L. Jiling, Z, Xiaoxing, L. Yi, Y. Long, L. Xudong, and L. Zhengyl (1999), "The progress of the Water Cooled Separator CFB Boiler in China, Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. M. G. Alliston, T. Luomaharju and A. Kokko, (1999), "Improved Water-Cooled Cyclone Constructions in CFBs", Proceedings of the ASME 15th International Conference on Fluidised Bed Combustion, Savannah, May 1999. S. L. Darling, (2000), "Foster Wheelers Compact CFB; Current Status", Foster Wheeler Technical Papers, www.fwc.com/publications/tech_papers/ powgen/compact.cfm. UNIPEDE (1995), Pulverised Coal-Fired Power Plants with Advanced Steam Conditions UNIPEDE Thermal Generation Study Committee 20.01 Thermoprocess, Group of Experts: New and Advanced Processes for Electricity Generation, June 1995, Ref 02001Ren9519. 32 of 108

46.

47.

48.

49.

50.

51.

52.

53. 54.

55.

56. 57.

58.

59.

60. 61.

F. Masayuma, T. Yokoyama, Y. Sawarages and A. A. Iseda, (1994), Development of a Tungsten Strengthened Low Alloy Steel with Improved Weldability, in Materials for Advanced Power Engineering 1994 Proceedings of a Conference held in Lige, Belgium, 3-6 October 1994. Edited by Coutsouradis, D. et al. Kluwer Academic Press, (1994), pp 173-181. R. U. Husemann, W. Bendick, K. Haarmann, J. Heeschen and A. Helmrich, (2000), The New 7CrMoVTiB10-10 Material (Part 2) For Boiler Membrane walls, VGB PowerTech 7/2000. M. Staubli, W. Bendick, J. Orr, F. Deshayes and Ch Henry, (1998), "European Collaborative Evaluation of Advanced Boiler Materials" in "Materials for Advanced Power Engineering 1998" Proceedings of a Conference held in Lige, Belgium, 5-7 October 1998. Edited by Lecomte-Beckers, J. et al. Forschungszentrum Jlich Gmbh (1998) pp 87-104. E. Metcalfe and B. Scarlin, (1998), "High Temperature Forged Components for Advanced Steam Power Plants" in "Materials for Advanced Power Engineering 1998" Proceedings of a Conference held in Lige, Belgium, 5-7 October 1998. Edited by Lecomte-Beckers, J. et al. Forschungszentrum Jlich Gmbh (1998) pp 53-70. F. Masuyama and N. Komai, (1998) "Effect of Re Addition on the Creep Strength of Heat Resistant Cr-W Steels", in "Materials for Advanced Power Engineering 1998" Proceedings of a Conference held in Lige, Belgium, 5-7 October 1998. Edited by Lecomte-Beckers, J. et al. Forschungszentrum Jlich Gmbh (1998) pp 269-276. C. J. Franklin and C. Henry, (1994), Materials Developments and Requirements for Advanced Boilers in Materials for Advanced Power Engineering 1994 Proceedings of a Conference held in Lige, Belgium, 3-6 October 1994. Edited by Coutsouradis, D. et al. Kluwer Academic Press, (1994), pp 89-107. Y. Sawaragi, A. Iseda, K. Ogawa and F. Masayuma, (1994), Development of a High Strength 12Cr Steel (HCM12A),in Materials for Advanced Power Engineering 1994 Proceedings of a Conference held in Lige, Belgium, 3-6 October 1994. Edited by Coutsouradis, D. et al. Kluwer Academic Press, (1994), pp 309-318. Ikezhima, (1983) Bulletin of the Japanese Institute of Metals, Vol 22, No. 5, (1983). R. Blum, J. Hald, W. Bendick, A. Rosselet and J. C. Vaillant, (1994) ."Neuentwicklungen Hochwarmfester Ferritisch-Martensitischer Sthle aus den USA, Japan und Europa" VGB Kraftwerkstechnik, Vol. 74, No. 8, (1994) pp 641-652. B. Rukes, W. Vollmer and E. Wittchow, (1994), Power Plants With High Steam Pressures and Temperatures Matched Design of Steam Turbine and Steam Generator, VGB Kraftwerkstechnik 74 (1994), Number 5, p5. C. W. Wegst, (1998), Stahlschlussel (Key to Steel), p300. Verlag Stahlschlussel Wegst GmbH, 1998. F. Abe, M. Igarashi, N. Fujitsuna, K. Kimura and S. Munecki,. (1998), Research and Development of Advanced Ferritic Steels for 650C USC Boilers" in "Materials for Advanced Power Engineering 1998" Proceedings of a Conference held in Lige, Belgium, 5-7 October 1998. Edt. Lecomte-Beckers, J. et al. Forschungszentrum Jlich Gmbh (1998) pp 269-276. P. Eggerstedt and J. Zievers, (1999), "High Temperature Particulate Removal Using Recrystallised Silicon Carbide Candle Filters", High Temperature Gas Cleaning, Vol. 2, 1999, pp 375-383. T. E. Lippert, G. J. Bruck, R. A. Newby, D. M. Bachovchin, V. L. Debski and H. T. Morehead, (1997), "Westinghouse Hot Gas Particle Filter System", Proceedings of the 14th International Conference on Fluidised Bed Combustion, Vol. 2, Vancouver, May 1997. J-F. Lecostaouec, (1999), "3D Braided Fibre Reinforced Ceramic Hot Gas Filters", High Temperature Gas Cleaning, Vol. 2, 1999, pp 428-441. J. D. Spain and H. S. Starret, (1999), "Characterisation of Monolithic and Composite Filter Elements", High Temperature Gas Cleaning, Vol. 2, 1999, pp 414-427. 33 of 108

62.

63.

T. Tamura, T. Sato and Y. Fukuda (1995), High Temperature Strengths and Steam Oxidation Properties of New 9~12%Cr Ferritic Steel Pipes for USC Boilers 2nd International Conference on Heat-Resistant Materials 11-14, September 1995, Gatlinburg, Tennessee, USA. R. Vanstone, (2000), "Advanced 700oC Pulverised Fuel Power Plant" Proc. 5th Int. Charles Parsons Conf: Parsons 2000 Advanced Materials for 21st Century Turbine and Power Plants. Edited by A. Strang, R. D. Conroy, G. M. McColvin, C. Neal and S. Simpson. IOM Communications, London, Book 736 (2000).

34 of 108

{PRIVATE }Table 3.1.1.

{PRIVATE }MATERIAL C Si Mn P S Al

Chemical Compositions of Boiler Tube Materials

Cr Ni Mo V Nb W N Cu (Ce) Ti B

15Mo3 (1.5415) 13CrMo44 (1.7335) T22 (10CrMo910) (1.7380) T23 (HCM2S) T24 (7CrMoVTiB10-10)

0.12 0.20 max 0.15 0.08 0.15 0.04 0.10 0.05 0.10

0.10 0.35 max 0.50 max 0.50 max 0.50 0.15 0.45 0.20 0.50 0.10 0.30 max 0.50 max 0.50 max 0.70 max 0.50 0.30 0.75 0.30 0.80 0.30 0.60

0.40 0.90 0.30 0.61 0.30 0.60 0.30 0.60 0.30 0.70 0.30 0.60 0.30 0.60 max 1.00 0.30 0.70 max 0.70 max 1.00 5.50 7.00 1.50 2.00 max 1.50

max 0.035 max 0.03 Max 0.02 max 0.02 max 0.02 max 0.02 max 0.03 max 0.03 max 0.03 max 0.03 max 0.03 max 0.035

max 0.03 max 0.01 Max 0.01 max 0.01 max 0.01 max 0.01 max 0.03 max 0.02 max 0.03 max 0.03 max 0.02 max 0.015

<0.03 (sol) Max 0.02 max 0.040 max 0.025 -

0.80 1.25 1.90 2.60 1.90 2.60 2.20 2.60 8.50 9.50 8.50 9.50 8.00 13.00 11.0 13.0 10.00 12.60 10.00 12.50 14.00 16.00 14.00 16.00 15.9 15.50 17.50

0.25 0.35 0.44 0.65 0.087 1.13 max 0.30 0.90 1.10

0.20 0.30 0.20 0.30 0.18 0.25 0.15 0.25 0.10 0.30 0.20 0.30 0.15 0.30 0.25 0.35 0.15 0.40 0.60 0.85

0.02 0.08 -

1.45 1.75 -

max 0.03 Max 0.012

0.05 0.10

max 0.006 0.0015 0.0070 max 0.01 max 0.01 max 0.005 0.003 0.009 0.004 0.008 -

P 91/T 91 (X10CrMoVNb91) (1.4903) E 911 P92 (NF 616) HCM 12 P122 (HCM12A) X20CrMoV121 (1.4922) E1250 12R72 X10CrNiMoTiB1515 17-14CuMo X8CrNiMoVNb1613 (1.4988)

0.08 0.12 0.09 0.13 max 0.15 max 0.12 0.06 0.14 0.17 0.23 0.05 0.15 0.08 0.12 0.11 0.04 0.10

max 0.40 0.10 0.35 max 0.70 0.30 0.80 9.00 12.00 14.00 16.00 14.5 12.50 14.50

0.85 1.05 0.90 1.10 max 1.00 0.80 1.20 0.20 0.60 0.80 1.20 0.80 1.20 1.00 1.40 2.5 1.10 1.50

0.06 0.10 0.06 0.10 max 0.10 max 0.20 0.02 0.10 0.75 1.20 0.43 max 10xC

0.90 1.10 1.50 2.50 0.80 1.20 1.50 2.50 -

0.05 0.08 0.05 0.08 0.02 0.15 0.02 0.10 0.01 -

0.30 1.70 3.10 -

max 0.05 0.30 0.60 0.24 -

35

Table 3.1.1. (Continued) Chemical Compositions of Boiler Tube Materials

{PRIVATE }MATERIAL C Si Mn P S Al Cr Ni Mo V Nb W N Cu (Ce) 2.50 3.50 max 0.75 max 0.20 (0.05) (0.10) Ti B

X8CrNiMoNb1616 (1.4981) X3CrNiMoN1713 (1.4910) AISI T316 AISI T321 AISI T347 AISI T304 Super 304 NF709 Alloy 800 AISI T310 Valinox T310N HR3C (1.6974) HR6W AC66 (1.4877)

0.04 0.08 max 0.04 max 0.08 max 0.08 max 0.08 max 0.08 0.07 0.13 max 0.20 max 0.10 max 0.25 0.05 0.07 0.04 0.10 max 0.10 0.04 0.08

0.30 0.60 max 0.75 max 1.00 max 1.00 max 1.00 max 1.00 max 0.30 max 1.00 max 1.00 max 1.50 0.30 0.50 max 0.75 max 1.00 max 0.30

max 1.50 max 2.00 max 2.00 max 2.00 max 2.00 max 2.00 max 1.00 max 1.50 max 1.50 max 2.00 1.00 1.40 max 2.00 max 2.00 max 1.00

max 0.035 max 0.035 max 0.03 max 0.03 max 0.03 max 0.03 max 0.04 max 0.03 max 0.015 max 0.03 max 0.01 max 0.03 max 0.030 -

max 0.015 max 0.015 max 0.045 max 0.045 max 0.045 max 0.045 max 0.010 max 0.010 max 0.045 max 0.025 max 0.030 max 0.030 -

0.15 0.60 max 0.025

15.50 17.50 16.00 18.00 16.00 18.00 17.00 19.00 17.00 19.00 18.00 20.00 17.00 19.00 18.00 22.00 19.00 23.00 24.00 26.00 24.40 25.00 24.00 26.00 21.00 25.00 26.00 28.00

15.50 17.50 12.00 14.00 10.00 14.00 9.00 12.00 9.00 13.00 8.00 12.00 7.50 10.50 22.00 28.00 30.00 35.00 19.00 22.00 20.60 21.50 17.00 23.00 35.00 45.00 31.00 33.00

1.60 2.00 2.00 2.80 2.00 3.00 1.0 2.0 max 0.20 -

max 10xC 10 x C 0.30 0.60 0.10 0.40 0.40 0.50 0.2 0.6 max 0.40 0.6 1.0

4.00 8.00 -

0.10 0.18 0.05 0.12 0.05 0.20 0.18 0.24 0.15 0.35 -

5xC 0.02 0.20 0.15 0.60 max 0.20 -

0.08 0.002 0.010 0.005 -

36

Table 3.1.2.

T23 - Allowable Stresses for 105h Operation Allowable Stress, MPa 124 117 111 105 87 71 56

Temperature, C 400 450 500 525 550 575 600

Table 3.1.3.

7CrMoVTiB10-10 (T24), Allowable Stresses for 105h Operation Temperature, C 400 450 500 525 550 575 600 Allowable Stress, MPa 131 126 117 112 95 67 39

Table 3.1.4.

P92, Allowable Stresses for 105h Operation Allowable Stress, MPa 132 126 116 110 103 94 70 48

Temperature, C 399 454 510 538 566 593 621 649 Table 3.1.5.

P122, Allowable Stresses for 105h Operation Temperature, C 399 454 510 538 566 593 621 Allowable Stress, MPa 134 127 118 112 104 89 64

37

Table 3.1.6.

NF709, Allowable Stresses for 105h Operation Temperature, C 450 500 550 600 650 700 Allowable Stress, MPa 130 125 120 117 82 57

Table 3.1.7.

HR3C, Allowable Stresses for 105h Operation Temperature, C 500 600 650 700 Allowable Stress, MPa 108 105 75 45

Table 3.1.8. Installation of Trial Header and Pipes in COST 501

Station Vestkraft Unit 3 Nordjyllandsvaerket Schkopau Unit B Staudinger Unit 1 Skaerbaek Unit 3 GK Kiel VEW Westfalen Westfalen Avedoere 2 Material NF616 NF616 HCM12A E911 E911 E911 P92 E911 E911 P92 P92 ID 240 160 550 201 230 480 23.8 159 159 Size Wall 39 45 24 22 60 28 4 27 27 Component Straight pipe Main steam Header Induction bend hot reheat Induction bend main steam Induction bend main steam Header Superheater Steam Loop Steam Loop Headers and steam lines Japan Japan M, V&M M, V&M M, V&M Mannesmannrhrenwerk Dsseldorf-Reisholz M, V&M M, V&M M, V&M Manufacturer Steam 560C/250 bar 582C/290 bar 560C/70 bar 540C/213 bar 582C/290 bar 545C/53 bar 650C 650C/180 bar 650C/180 bar 580/600 300 bar Installed 1992 1996 1996 1996 1996 1997 1998 1998 1998 2000

Table 3.1.9.

X3CrNiMoN1713, Allowable Stresses for 105h Operation Temperature, C 550 600 650 700 Allowable Stress, MPa 147 94 55 34 38

- 10.14 - 7.89 - 4.20 Wnet Improvement in heat rate (%) Wnet -14 -12 -10 -8 -6 -4 -2 -0 - 3.23 - 1.88 535 560 600 650 185 - 5.14 - 7.30 225 Live steam pressure (bar) 250 - 8.62 - 5.74 - 9.24 300 - 8.41

Live steam/reheat temperature (oC)

Figure 3.1.1 Improvements in Power Plant Heat Rate with Increases in Temperature and Pressure to the Turbine1

Double reheat

Reduced boiler pressure drop

52 50 48 46 44 42 40 38
250/540/560 Staudinger 5 Current status 250/580/600 P91 used

315/620/620 austenitics used

48.5 47.7

49

49.8 50.2

44.5 45.2 43.2

45.6

300/600/620 NF616 used

370/700/720 Nickel alloys (Thermie)

Improved boiler layout

Cold end heat recovery

Figure 3.1.2

Potential Power Plant Efficiency Gains from the Utilisation of New Materials and Cycle Improvements

39

BULK GAS TEMPERATURE GAS FILM OUTER OXIDE/FOULING

TUBE WALL

STEAM/WATER FILM BULK STEAM/WATER TEMPERATURE

500 600

Figure 3.1.3

Schematic Diagram of the Temperature Gradients through Boiler Tube Walls for Design

500
MINIMUM TENSILE STRENGTH

STRESS (MPa)

250

MINIMUM 0.2% PROOF STRESS OR LOWER YIELD STRESS MEAN STRESS FOR RUPTURE IN 100,000h

100

DESIGN CURVE INCORPORATING SAFETY FACTOR

50

0 100 200 300 400 TEMPERATURE oC

Figure 3.1.4

Relationship between Measured Proof and Allowable Design Stresses for Plain Carbon Steel

BORE OXIDE

40

200

CORROSION RATE (nm.hour )

-1

150
= Tg
=

12

00

11

00
o

Tg

100
Tg
Tg

10

00

C
C

00 = 9

50

0 580 590 600 610 620 630 640 650 660

SURFACE METAL TEMPERATURE (oC)

Figure 3.1.5

Fireside Corrosion Rate of Austenitic Superheater/Reheater Materials as a Function of Gas and Metal Temperature.

Figure 3.1.6

Breakdown of Highly Protective Chromia Scale (right) on 12%Cr Steel to produce Duplex Scale Nodules (left)

41

800
1% & 2% Cr @ 600oC

700 BORE SCALE THICKNESS ( m) 600

9% Cr @ 600oC

500 400 300 200 100 0 0 20 40 60 80 100 120 140 EXPOSURE TIME (103 hours) 160 180 200

1% & 2% Cr @ 560oC 12% Cr @ 600oC

9% Cr @ 560oC

12% Cr @ 560oC

Figure 3.1.7

Anticipated Bore Oxide Scale Thickness as a Function of Material, Temperature and Operating Hours for Low-medium Cr Superheater Tube Materials17.

50
TUBEWALL TEMPERATURE INCREASE ( C)

9Cr @ 600oC

40

30

12Cr @ 600oC

20
9Cr @ 560oC

10
12Cr @ 560oC

0 0 20 40 60 80 100 120 140 160 180 200 EXPOSURE TIME (103 hours)

Figure 3.1.8

Increase in Tube Wall Tube Wall Temperature as a result of Increasing Bore Scale Thickness17.

42

40
10CrMoV121
THERMAL CONDUCTIVITY (W.m .K
-1) -1

30

P91

X20CrMoV121

20

X3CrNiMoN1713

10

0 0 100 200 300 TEMPERATURE

oC

400

500

600

Figure 3.1.9

Thermal Conductivity of Some Current and Proposed Header and Steam Separator Materials18.

MEAN COEFFICIENT OF LINEAR THERMAL EXPANSION o BETWEAN 20 C AND INDICATED TEMPERATURE

10 o T K

-6

20

18

X3CrNiMoN1713

16

14
10CrMo910

12
P91 X20CrMoV121

10 0 100 200 300

oC

400

500

600

TEMPERATURE

Figure 3.1.10

Mean Coefficient of Linear Expansion for some Current and Proposed Header and Steam Separator Materials18.

43

160
HCM12A

140
Allowable Stress Values (MPa)

HCM12 T91

120 100 80 60 40 20 0 500

HCM2S X20CrMoV121 T22

525

550

575
o

600

625

650

Temperature ( C)

Figure 3.1.11

Comparison of Allowable Stress Values for 2 to 12% Cr Steels

160 140
T22 T23 T24 T91

Allowable Stress (MPa)

120 100 80 60 40 20 0 400

450

500

550

600

650

Temperature ( oC)
Figure 3.1.12 Design Values for T23/T24 Compared to Other Grades

44

50
1 17-14 CuMo 2 AN31 10 800H 11 807 12 617 13 625 14 310S 15 HK4M (35) 16 HR3C 4 5 7 8 9 10 12 13 11 14 15 17 18 19 20 21 8 TP347H (MITI) 9 TP347H (Fine Grain) 17 SZ (36) 18 35 Cr-54 Ni-Nb 19 40 Cr-50 Ni-Fe 20 in-671 21 Chromised

40
WEIGHT LOSS (mg.cm )
-2

3 E1250 4 12R72 1 3 6 2 5 15-15N 6 TP321H

30

20

10

16

0 0 10 20 30 40 CHROMIUM CONTENT (%)

50

60

Figure 3.1.13

Metal Loss Rate as a Function of Cr Content for Higher Cr Materials under Simulated Fireside Corrosion Conditions54.

X20

P91

66 97 NF616

46

Figure 3.1.14

Comparison of Relative Wall Thickness Requirements for Main Steam Pipework for 300bar/580C Duty Manufactured from X20, P91 and P92 Steels45.

45

110
300bar

HEADER WALL THICKNESS (mm)

100
X20CrMoV121 300bar 260bar P91

90

80
260bar

70

300bar

60

1.4910

50

260bar

40 540 550 560 570 580 590 SUPERHEATER OUTLET TEMPERATURE (oC) 600

Figure 3.1.15

Calculated Wall Thickness for Final Superheater Outlet Headers at 260 and 300 bar as a Function of Material and Operating Temperature11.

Figure 3.1.16

-1

260 bar
o

P91 300 bar

260 bar 260 bar X20CrMoV121 300 bar 1.4910 300 bar

RATE OF TEMPERATURE CHANGE ( C.min )

o SUPERHEATER OUTLET TEMPERATURE ( C)

Acceptable Rates of Temperature Change for Header Materials11

46

ID

= 140mm 70mm

Wall =

Range of Application for 1.4910Steel

Figure 3.1.17

Figure 3.1.18

PRESSURE AHEAD OF TURBINE (bar)

STEAM TEMPERATURE AHEAD OF TURBINE o C) (

Application Limits for the Use of 1.4910 Austenitic Steel in the Superheater Outlet Headers of Supercritical Plant11.

2.

25

Cr

1%

0.02%Si

o
0.11%Si 0.25%Si

0.44%Si 0.57%Si 0.78%Si

Steamside Oxidation Rate of NF616 at 650C as a Function of Si-content62.

47

3.2 3.2.1

Steam Turbines Steam Turbine Market Trends

Steam turbines are an essential part of all combustion-based power generation technologies. At some point in the cycle, all of these technologies raise steam whose energy is converted via a steam turbine. The materials issues related to the steam turbine depend largely on the temperature and pressure of the steam generated and, as such, are almost independent of the steam raising technology employed. At the current time steam turbines serve two principal fossil-fired technology markets: Coal-fired plant based on PF or CFB combustion technology. For utility-scale generation this may require steam turbines with outputs up to 800MWe. Gas-fired plant incorporating a combined cycle where the exhaust gas from a gas turbine is used to raise steam. To match the lower power output of the gas turbine, steam turbines for these applications have outputs typically in the range up to 250MWe. At the moment the second of these markets is dominant. The dash for gas in the UK is being replicated to a greater or lesser extent in markets around the world, of which the North American market has recently been the most buoyant. The use of gas has significantly displaced demand for new coal-fired plant over recent years. Nonetheless concerns over security of supply and over the volatility of gas prices have led many utilities, both in Europe and the USA, to begin re-examining the potential for coal-fired generation. In addition to the construction of new plant, an important subset of the first market is retrofit turbines. Advances in the internal efficiency of steam turbines have made it economically attractive to replace elements of older turbines with improved modern turbine components. The advantage can be taken to reduce fuel consumption or increase power output.

3.2.2

Steam Turbine Development Trends

Steam turbine development trends fall into two categories: 1. Improvements in the internal efficiency of the steam turbine. These developments largely focus on improved aerodynamic performance, sealing and cylinder configuration: Improved aerodynamics to reduce losses in the steam path, including valves, blades and cylinder exhausts. Improved sealing to reduce steam leakage flows. Improved cylinder configuration to minimise the number of steam flows and thus minimise losses associated with flow over steam path surfaces. This essentially involves the extension of single flow concepts as far as possible rather than the use of double flow cylinders. It is linked to increasing the exhaust area of LP cylinder flows through the introduction of longer blades. Category 1 improvements bring benefits to the whole range of technologies served by the steam turbine and they have provided the basis for the current retrofit market. Advanced materials have played, and continue to play, a role in this area, for example through: The introduction of stronger or less dense blading alloys enabling more efficient steam path arrangements. 48 of 108

The introduction of rotor materials with the necessary combination of high temperature creep resistance and low temperature fracture toughness to enable the construction of single-flow, single cylinder steam turbines.

2. Improvements in the cycle efficiency. These developments focus on increasing steam temperatures and pressures so as to increase the thermal efficiency of the overall cycle. Category 2 improvements are focused much more on steam turbines for coal-fired plant where there are clear advantages in elevating temperature and pressure to increase thermal efficiency and reduce emissions. In combined cycle plant, the drivers for increased temperature in the steam turbine are low. Here the steam turbine inlet temperature is largely dependent on gas turbine exhaust temperatures and there is no great thermodynamic benefit in elevating this exhaust temperature. Currently these steam turbines operate with inlet temperatures of up to 565oC but there is no great driver for increasing this temperature. Improvements in temperature and pressure capability for coal-fired plant are almost entirely dependent on improvements in materials technology through: Development of alloys with creep resistance at higher temperatures Development of alloys or surface treatment technologies with steam oxidation resistance at higher temperatures The drivers for steam turbine development for current coal-gasification based power generation cycles are similar to those for gas-fired combined cycle plant, where superheating is limited to the exhaust gas from the gas turbine. However, if methods of superheating in the high temperature fuel gas or supplementary superheating stages are provided, then these systems could take advantage of the higher steam conditions being used in PF plant.

3.2.3

Steam Turbine Design and Operation

The first sixty years of the twentieth century witnessed progressive increase in operating temperatures and pressures. In the 1950s the continuation of this trend to temperatures of 540565oC was enabled through the introduction of a generation of creep-resistant low alloy steels, such as the 1%CrMoV rotor forging alloy. A generation of 12%Cr alloys was developed for blading applications and it remains the practice to use these alloys throughout the turbine. This trend reached its peak in the early 1960s when operating temperatures were raised to 593 oC at the Drakelow plant in the UK and even to 649oC in the Eddystone plant in the USA. Austenitic stainless steels were applied to critical turbine components such as the valve chest and rotor forgings. However service experience with these plants was not successful even though some continue to operate to the current day. Austenitic steels have a poor combination of physical and mechanical properties to withstand thermal cycling: their high coefficients of thermal expansion, low thermal conductivities and low yield strengths mean that thermal cycles result in high thermal gradients in thick section components, leading to stresses which can exceed yield. This resulted in poor dimensional stability in thick section components and one of the Drakelow boilers collapsed. Continued operation has only been possible through the use of careful warming of valve chests in preparation for start-up. As a result of this experience, steam turbine conditions retreated to the 540-565oC level and remained at that level for the next 30 years. The majority of this plant operated under subcritical conditions with pressures of around 180bar but in Europe and in the USA several plants with higher super-critical pressures of around 250bar were built. Only with the development of more creep resistant martensitic stainless steels, starting with the development of P91 in the late 70s and early 80s, was a further advance in steam conditions considered. In Japan a series of plants have been built or are under construction with steam conditions of up to 610oC. In Europe the double reheat Danish plants at Skaerbaek and 49 of 108

Nordjlland with conditions of 580oC/580oC/580oC/285bar were followed by the construction of the Lippendorf plant with a reheat temperature of 585oC and the Avedore plant with a reheat temperature of 600 oC. Siemens is carrying out a design study for a plant at Westfalen with even higher temperatures. In Europe the increase in temperature has been coupled with an increase in pressure but in Japan pressures have remained at the level of 240bar (Figure 3.2.1). In Europe, Japan and the USA, development programmes are now in place to elevate operating temperatures to beyond 700 oC through the application of nickel based alloys. Commissioning of the first of such plant is foreseen in the years around 2010. As well as operation under steady load, steam turbines are often operated under daily stop-start cycles (two-shifting) so that an ability to withstand such cycling is required and is specified by customers. Modern plant is generally more efficient than older plant and may operate under a base-load regime early in its life but later, as still more advanced plant is built, it may be operated under cycling conditions.

3.2.4

Steam Turbine Materials State of the Art

A review of materials developments for steam turbines is incorporated in the literature review (Annex 1). In addition excellent review papers have been published in recent years1,2,3,4 An important driver for all areas of steam turbine development is the issue of cost. The steam turbine market is an extremely competitive one and the manufacturers are under constant pressure to reduce costs. This contributes to developments that simplify turbine architecture such as the development of single cylinder machines and also to the application of cheaper materials. An example is the development of cast irons to replace cast steels. Steam Turbine Applications Advanced PF Plant The current generation of ultra-supercritical (USC) PF plant being constructed in Japan and Europe is based on a generation of improved martensitic stainless steels developed for steam turbine applications in the 1980s and early 1990s. Rotor forgings are based on 910%CrMoVNbN steels, the main alloys currently being applied having either a Mo addition of 1.5% or an addition of up to 1.0%W in partial substitution of the Mo content. V and N contents have been optimised to provide precipitation strengthening through a dispersion of VN particles and a low level of Nb is incorporated to control grain size during high temperature heat treatments. Castings for valve chests and cylinder casings exploit analogous alloys, generally with lower C content to provide improved weldability. Blading alloys are similar to the rotor forging alloys. Meeting the requirements of bolts operating at the very highest temperatures has frequently required the exploitation of Ni-based alloys such as Nimonic 80A or Refractalloy 26. The temperature at which 100,000 hour rupture strength is around 100MPa is a reasonable indicator of the maximum application temperature of steam turbine materials. The alloying developments described above have led to increases in this temperature of around 50-70oC (Figure 3.2.2). Even more advanced B containing steels have been developed and demonstrated as full scale forgings in both Japan and Europe. To provide higher oxidation resistance, alloys with higher Cr levels (11-12%) have also been developed: in order to balance the composition of these alloys to avoid delta ferrite formation, additions of cobalt have been made. However, these more advanced alloys have not yet been applied in service.

50 of 108

In support of the development of steam turbines for temperatures greater than 700oC, development of nickel based alloys for application to critical steam turbine components is currently in progress5. Retrofit Plant In some cases the alloys developed for the USC plant have also provided benefits for retrofit plant. For example higher allowable stresses in advanced blading steels have enabled the incorporation of additional stages in retrofit turbines leading to higher efficiency. Steam Turbines for Combined Cycle Plant To a large extent, it has been possible to base steam turbines for combined cycle plant, with their modest steam conditions, on the conventional low alloy steels used since the 1950s. However opportunities have been found for the more advanced steels to improve performance or reduce cost. For example the use of 10%CrMo(W)VNbN steels for rotor forgings can lead to improved cycling resistance4 and the use of newly developed low alloy steels has enabled the construction of single cylinder machines with advantages in performance and cost. Materials Sources The most critical components for steam turbines are manufactured from the following product forms: Large forgings for rotors Smaller forgings for diaphragm construction, valve components and larger blades Bar for blades and bolts Castings for valve chests and cylinder casings The main sources of these materials for UK plant manufacturers are in the UK and in the rest of the European Union. Nonetheless there is a growing trend to source materials from low cost countries such as those of Eastern Europe and for contracts in Asia there is a common contractual requirement to use local material suppliers. In addition Japanese suppliers are occasionally used for supply of large rotor forgings. Despite these trends, UK material suppliers still play a major role in the supply of materials to UK turbine manufacturers and also to turbine manufacturers in other countries. Examples include: Sheffield Forgemasters Engineering Ltd: rotor forgings and castings Hi-Tec: castings Goodwin Steel Castings: castings Corus: blading and bolting bar Allvac SMP: blading and bolting bar Firth-Rixon: small forgings Walter Somers: small forgings International Forgemasters: small forgings Repair, Maintenance, Life Assessment Many steam turbine components have been constructed from steels which are inherently weldable. Therefore ready means of repair through welding are available. Advances in welding technology have extended the range of alloys considered weldable. At one time 1%CrMoV rotor steels were considered unweldable but today welding procedures have been developed to extend the life of these components6. In components where NDE has revealed no defects, the first stage of life assessment is generally based on application of the most recent design and stress analysis techniques. This may be refined by materials testing on the component in question so that analysis is based on specific 51 of 108

component properties rather than minimum properties for the alloy. This may be supported by metallographic examination for evidence of creep or LCF damage. The latter part of this assessment requires a knowledge of the evolution of microstructure as creep damage or LCF damage accumulates. In situations where cracking has been observed, information on creep-fatigue crack growth is required for optimum life management. The application of oxidation resistant coatings in the future will necessitate the development of means of assessing their state of life-exhaustion and of repair procedures.

3.2.5

Steam Turbine Materials R&D

Current Projects The major projects for development of steam turbine materials in which UK participants are active are as follows: COST 522 This programme is a collaborative project involving participants from the UK, other EU countries and other non-EU countries such as Poland, Switzerland, Czech Republic, Slovakia and Hungary. The programme covers a wide range of power generation technologies but one of its major themes is Steam Power Plant (SPP) within which there is a major activity on steam turbine materials7 (Figure 3.2.3). It includes all the major European turbine makers and their material suppliers. The programme started in 1998 and runs to 2003 although there is a possibility that it will be extended. UK participants in this project are supported via the DTIs Cleaner Coal project. The COST 522 SPP activitys main objective is the development of materials technology to enable operation of power plants with steam temperatures of 650C. Development is focussed on improving the temperature capability of martensitic stainless steels and covers: The development of new 9-12%Cr martensitic stainless alloys for rotor forgings and casings, the objective being an improvement in the creep strength of alloys previously developed in COST 501, the forerunner to COST 522, coupled with improvements in oxidation resistance. Testing therefore includes oxidation testing in a range of laboratory environments as well as mechanical testing. The development of new high strength bolting alloys with thermal coefficients of expansion matching those of the 9-12%Cr martensitic stainless steels. Metallographic investigation and modelling of the new alloys. EU Thermie AD700 Project: Advanced 700 C PF Power Plant This project started in 1998 and is a phased project5 extending to 2014 (Figure 3.2.4). The first phase, to run until 2003, is receiving support from the EC and, in the case of the UK participants, from the DTIs Cleaner Coal Project. The project involves 40 participants from 7 European countries. A second phase of the project has recently been launched with additional support from the EC. This phase of the project will run from 2002 to 2005. The first three years of the project resulted in confirmation of the technical and economic feasibility of the concept which fundamentally depends on the application of nickel-based alloys to enable steam temperatures of 700-720C. Turbine related activities included the assessment of a range of candidate alloys and the manufacture of the first prototype components. The second phase will include the extension of the prototype demonstration and characterisation programme. DTI Cleaner Coal Projects 52 of 108

As well as supporting the involvement of UK participants in the COST 522 and Thermie projects, funding has been provided for other turbine-related research projects in the areas of materials and aerodynamics. NPL Life Performance of Materials Programme NPL has recently launched activities under its Life Performance of Materials programme which support turbine development: The High Temperature Degradation project includes research into oxidation mechanisms on a range of steam turbine alloys. The Aqueous Corrosion project includes research into stress corrosion cracking in steam turbine materials. These projects began in 2001 and are planned to run for three years. There are other steam turbine development programmes in Europe from which UK participants are excluded. In Germany in particular there is strong national funding for the MARKCO programme and the VGB is funding the long-term characterisation of advanced 9-12%Cr components under its VGB 158 project. There are also significant development programmes outside Europe. In Japan NIMS is responsible for a major programme (ABE, Parsons) focussed on the continued development of 9-12%Cr steels for 650C steam conditions, the same principal objective of COST 522. A programme funded by the EPDC has also been launched whose objective is 700C steam conditions. In the USA there are fewer programmes to develop steam turbine technology. Nonetheless American turbine makers have access to many of the results of developments in Europe and Japan. Despite the apparently low level of turbine development, there are major initiatives in boiler development, the DOE having just launched a 5- year, $21 million programme to develop boiler materials technologies for steam temperatures up to 870oC. EPRI continue to fund development of pipework materials within its project 1403. UK Strengths and Weaknesses The UK strengths and weaknesses are addressed in terms of the required capacity for materials R&D and in terms of the UKs ability to exploit the results of that R&D. Modelling of microstructural evolution during processing and in service, of mechanical properties, of oxidation and of the behaviour of oxidation-resistant coatings. The UK research community is competitive in this field. Universities such as Cambridge and Loughborough have expertise in microstructure and property modelling and participate in COST 522. Other universities such as Imperial College are also active in creep modelling. NPL have expertise in modelling of oxide and coating behaviour. Test capacity for short and long-term mechanical property characterisation: creep testing, low cycle fatigue and cyclic hold testing, fracture toughness testing, creep-crack initiation and growth testing. UK turbine manufacturers such as ALSTOM and larger material suppliers such as Corus have in-house capability for this testing and can call on support from research institutes such as NPL. Test capacity for environmental (steam oxidation) testing, including the influence of thermal cycling on oxide layers on coatings. NPL and Cranfield University have capacity for laboratory testing. However the correlation between laboratory testing and service performance is weak. The UK lacks a capacity for 53 of 108

exposure in operating power stations. Such facilities exist in Denmark and Germany and place the UK at a disadvantage. Manufacturing capacity for trial melts and more significantly for prototype components including weldments where appropriate. UK forgemasters and foundries have the capability to contribute in this area. SFEL manufactured one of the first COST 501 prototype rotors and Goodwin Steel Castings is active in the Thermie project. Capacity for development of welding consumables. UK suppliers are active participants in COST 522. Capacity for coatings resistant to steam oxidation. UK suppliers exist who are capable of contributing in this area. Some are already active in support of gas turbine coating activities in COST 522.

Although UK universities are strong on modelling, the other areas of research requirements are largely met by industry itself. In other countries such as Germany, universities and research institutes make a much greater contribution to the short and long-term mechanical testing requirements, for example providing creep testing facilities at lower costs than available in the UK, especially since these bodies attract government support for these activities. Many of the strengths of the UK for R&D derive from the capabilities of UK industrial companies. It follows that involvement of those material suppliers, turbine makers, coating and consumable suppliers in R&D projects carries with it an implicit ability for rapid and effective industrial exploitation. The principal weakness identified in R&D lies with the absence of an environmental test facility in operating plant. This reflects the UK electricity generators absence of investment in R&D since the privatisation of the CEGB and under the financial and commercial pressures of deregulation.

3.2.6

Future R&D Priorities

As a result of the trends in steam turbine development, the major emphasis of materials development around the world is focussed on improved high temperature materials for the high temperature cylinders. Nonetheless development continues at a lower level of effort on materials for single cylinder turbines and on materials for the low temperature turbine: Development of rotor materials for single cylinder machines Improved understanding of stress corrosion cracking Substitution of lower cost materials High temperature materials development is a very long-term process. A typical development programme proceeds through several stages: Investigation of trial melts including creep testing to at least 10,000 hours (~2 years) Manufacture of a prototype component in the best trial melt (~1 year). This stage is essential to demonstrate the alloy can be applied to the very large components required for turbine rotors and casings without problems arising from excessive segregation or cracking during manufacture. The inspectability of large components is also demonstrated. Characterisation of the prototype including creep testing to at least 30,000 hours (~4 years). There is potential for significant variation of properties through the section of large components, especially in comparison to the properties achieved in trial melts. Therefore characterisation of the prototype is essential. Characterisation typically includes long-term creep testing, low cycle fatigue and cyclic hold testing and investigation of the influence of 54 of 108

long-term ageing on tensile and impact strength. Investigations of fracture toughness and creep crack initiation and growth properties may also be carried out. Launch as commercial material and gain first order (~1 year) Test commercial products to establish scatterband of properties (~4 years). Cast to cast variation in properties is typically of the order of +/-20%. A knowledge of this scatterband and the position of the first prototype within it is required to fully exploit the properties demonstrated in a single prototype.

It thus takes about 12 years to achieve a completely reliable, mature material. For example in Europe the development of a new generation of martensitic stainless steel material for high temperature rotors began in the COST 501 programme in 1986. After investigation of trial melts, two prototype rotors were manufactured in 1989-90. Long-term characterisation of these materials led to their selection for the Skaerbaek and Nordjylland power plants in Denmark in 1993. These machines were manufactured and commissioned in 1993-97. Test pieces from the rotor forgings used for these turbines were included in a long-term testing programme funded by VGB that was launched in 1998. This programme finished in 2001 with the issue of data establishing a scatterband of properties for commercial rotors manufactured in these alloys. The total time that elapsed from the launch of the development programme to this fully mature condition has been 15 years. Attempts are being made to reduce this development time. Metallographic studies have been carried out and models are being developed to predict microstructure and the long term properties which depend on them. One key area of success in this effort has been the development of models predicting thermodynamic equilibria: MTDATA developed by NPL in the UK and Thermocalc developed in Sweden. An alternative approach is the use of neural network modelling to assess the influence of changes in chemical composition and heat treatment. These tools offer the ability to reduce the number of trial melts to be assessed before finding an alloy meeting the objectives and they also offer support for longer term extrapolation of data, increasing confidence so that alloys may be exploited earlier in the development cycle. Nonetheless the models are not yet sufficiently accurate or robust to remove the need for longterm testing. Even where long-term data already exists on alloys, it is possible that new applications require a cycle of prototype demonstration and characterisation. For example the application of nickel based alloys to steam turbine design and manufacture is supported to some extent by the existence of long-term creep data on some of the candidate alloys. However, this data is derived from product forms such as sheet and small diameter bars whereas turbine manufacture requires components with sections in the range 100-1000mm and weighing many tons. The difference in size has a potentially very significant influence on the microstructures and thus properties which are achieved. A greater potential for chemical segregation, greater difficulty in forging and controlling grain size and the much reduced heating and cooling rates during heat treatment all have potentially significant effects on short and long-term properties. In addition to development and long-term characterisation of the base alloy, development of welding procedures is also necessary, especially for casing materials. Thus development of welding consumables is essential together with long-term characterisation of the weld metal and of welded joints. To date, the development of high temperature alloys has been dominated by the need for improved creep strength. However, as target temperatures increase, the possibility that oxidation will limit application temperatures becomes significant. Current development programmes are now investigating oxidation resistance and it is already clear that some alloys are limited by their oxidation resistance. Alloys are being designed to improve this characteristic but an alternative approach is the development of coatings for protection against steam oxidation. 55 of 108

In summary, the following are required to support materials development: Modelling of microstructural evolution during processing and in service, of mechanical properties, of oxidation and of the behaviour of oxidation-resistant coatings. Test capacity for short and long-term mechanical property characterisation: creep testing, low cycle fatigue and cyclic hold testing, fracture toughness testing, creep crack initiation and growth testing. Test capacity for environmental (steam oxidation) testing, including the influence of thermal cycling on oxide layers and coatings. Manufacturing capacity for trial melts and more significantly for prototype components including weldments where appropriate. Capacity for development of welding consumables. Capacity for coatings resistant to steam oxidation.

References for Section 3.2 1. K-H Mayer, Advanced heat resisting steels, from Advanced Energy Conversion Materials, published by Nova Science Publisher Inc, USA, 2001. 2. R. Viswanathan and W. T. Bakker, Materials for Ultra Supercritical Coal Power Plants Part 2: Turbine Materials, from Power Plant Chemistry 2001, 3(8), pp446-451. 3. B. B. Seth, US developments in advanced steam turbine materials, from Advanced Heat Resistant Steels for Power Generation, published by IOM Communications Ltd, Book 708 (1999), pp 519-542. 4. D. V. Thornton and K-H Mayer, European high temperature materials development for advanced steam turbines, ibid. 5. S. Kjaer, F. Klauke, R. Vanstone, A. Zeijseink, G. Weissinger, P. Kristensen, J. Meier, R. Blum, K. Wieghardt, The advanced super critical 700C pulverised coal-fired power plant, proceedings of Powergen Europe 2001, Brussels, May 2001. 6. K. C. Mitchell, Extending the life of steam turbine rotors by weld repair, proceedings of 9th international conference on creep and fracture of engineering materials and structures, published as Institute of materials book 769, April 2001. 7. M. Staubli, K-H Mayer, T-U Kern, R. W. Vanstone, R. Hanus, J. Stief, K-H Schonfeld, COST 522 Power generation into the 21st century: advanced steam power plant, proceedings of 3rd conference on advances in material technology for fossil power plants, published as Institute of materials book 770, April 2001.

56 of 108