Is 700+°C steam temperature economically viable?21 May 1998
Supercritical steam conditions of 250 bar, 600°C HP, 610°C reheat are now described as state of the art in supercritical steam cycle systems. The prospect of further development towards steam temperatures higher than 700°C has been under active investigation since the early 1990s in Europe, Japan, Russia and the USA. Discussions with key experts at Siemens in Mülheim indicate that while the technology looks feasible, further investigation is needed to demonstrate the economic advantages. Efficiency increases from turbine blade improvements may lower life cycle costs.
Deregulation in the power supply industries around the world does little to advance the cause of supercritical coal fired generating plant systems development, even in countries with a continuing commitment to coal burning.
Countries with limited indigenous fuel resources and a high commitment to environmental protection such as Japan and Denmark, which appear to be dependent on imported coal for the foreseeable future, do have a continuing interest in increasing generating plant efficiency while reducing generating costs.
Australia, Germany, Russia, South Africa, the UK, and to some extent the USA are economically motivated to burn coal from indigenous resources with the highest possible efficiency without sacrificing availability, but the key economic criterion in these days of increasing competition in utility services continues to be the minimisation of lifetime operating costs.
Under these conditions the trade-offs between initial design criteria and materials selection against operation and maintenance costs may be conflicting. It is interesting, on the other hand, that China is increasingly interested in installing supercritical coal fired power plants to gain maximum thermal efficiency.
Current supercritical power plants reach thermal efficiencies of just over 40 per cent, with a few more recent highly supercritical power plants reaching 45 per cent. The leading developers of advanced coal burning power plant technology claim that thermal efficiency as high as 50 per cent, notwithstanding Carnot, is now within reach. But the main competition is from new gas turbine combined cycle plants which are now aspiring to an overall efficiency of 60 per cent, making a huge difference to generating costs and life cycle costs.
On the other hand, the newest gas turbines which are about to be introduced to the market will exhaust into waste heat recovery steam generators at temperatures above 600°C, necessitating the use of the same high chromium steel alloys such as ASTM-A335 P91 (X10CrMoVNb9-1) in the boiler tube bundles as is used in the more highly supercritical coal fired conventional generating plant now being built.
While increasing steam temperature is generally agreed to be the most rewarding route to increasing efficiency in coal fired power plant, there are recognised thresholds of investment costs above which the returns are not just diminishing, but increasingly negative.
This has led companies such as Siemens KWU, which is currently assimilating the technology of Westinghouse Electric in the USA, and Parsons in the UK, to concentrate more on the development of advanced steam turbine blading and shaft design than on the use of the newer austenitic steels and nickel based alloys. Nonetheless, Siemens claims it is able to offer plant designed to work at steam conditions of 300/600/610 and is working towards steam conditions in the order of 300/700/720.
The development of highly supercritical power plants in Germany has been substantially set back since the period of peak interest in 1994/95 with the shelving of the Hessler, Lübeck, and now RWE's newest Niederaussem power plant unit (250/580/600). But if the EU Energy Charter's aspirations towards cost transparency ever materialise, operating experience from recently built supercritical coal fired plants like the four big VEAG lignite stations in former eastern Germany – Boxberg, Lippendorf, Schwarze Pumpe (268/547/565) and Jänschwälde – and also the IPP Schkopau, none of which exceed 40 per cent efficiency by much, will do a good deal to settle the outstanding issues.
Perhaps the most significant reference will be Siemens KWU's latest supercritical plant contract, the 600 MWe Isogo unit being built for Tokyo Electric Power for operation in 2001, which is rated at 250/600/610, since blades of nickel based alloys are generally thought to be necessary for reheat temperatures of 610°C and higher. The steam turbines for this will be supplied directly from KWU's Mülheim works rather than from the local licensee Fuji Electric.
This represents a substantial increase in steam temperature over Chubu Electric's 700 MWe Kawagoe double reheat systems which run on steam conditions of 311 bar, 566/566/566°C. The first two of these units have been running since June 1989 at efficiencies of 45 per cent, an efficiency gain of five per cent over previous conventional plants.
These make strategic use of 12 per cent chrome cast steel in the Toshiba turbines and blades and nine per cent chrome in the pipework.
This was the first use of such alloys in Japan since the standard single reheat steam conditions of 242 bar, 538/566°C were established for large fossil fuelled steam turbines. Since the Kawagoe units demand the high levels of availability that has been essential for large fossil fired middle load power plants in recent years, operating experience from this station is of inestimable importance to advanced supercritical system development.
International development programmes
By 1995 research and development anticipating 50 per cent net plant efficiency within five to seven years was proceeding world-wide with the Brite Euram and COST in the European Community, the VKR Hessler and Preussenelektra Lübeck projects in Germany, the ELSAM "Convoy" double reheat power projects in Denmark, the KEMA work in the Netherlands, EPRI 1403 - 50 in the USA and CRIEPI in Japan.
Much of the attention under these programmes was directed at the metallurgical problems of the steam generators and the tendency for thinning of the superheater tubes in service using the ferritic/martensitic steels in common use. Some parties believe that in practice the tube thinning will not happen, and the new plants now beginning to operate in Germany, Japan and Denmark will give valuable insight and design validation for future projects.
The Schwarze Pumpe 2 x 800 MWe power plant recently commissioned near Dresden in the Lausitz lignite mining area is a good example. These are thought to be largest lignite boilers ever built, with physical dimensions of 161 m in height, 24m x 24 m in cross section. They are once through single pass boilers designed by EVT as 900 MWe units but with a high process steam output of up to 800 t/h to various process plants in the region. Then plants have steam parameters of 268 bar/547°C/565°C for the maximum steam generating capacity of 2420 t/h.
This is said to be the first time that P91 has been used for a power plant steam generator of this size. Used for the live steam and hot reheat piping, this material was selected as having higher creep rupture stress than the X20CrMoV121 normally used, which leads to lower tube wall thicknesses and hence reduction in thermal stresses. The weight reduction in turn reduces stress levels at the boiler connections and at the turbine connections as well as on the structural steelwork.
A seven stage regenerative feed heating system driven by a turbine driven feed pump delivers feedwater at a final temperature of 270°C. The supercritical steam output supplies a single reheat turbine system in sliding pressure mode with main steam at 638 kg/s, 253 bar, 544°C and reheat at 52 bar, 562°C.
Feedwater temperature is becoming an increasingly critical parameter in recent years with the added emphasis on emissions control. There are economic benefits in raising steam pressure to 300 bar since, because of the lower thermal energy transfer required in the evaporator, the heat exchange area required is reduced with a corresponding saving in costly high chromium alloys.
On the other hand, the maximum furnace temperatures have to be retained, even to the point of possibly removing the economiser, but this will then not cater for reheat inlet temperatures above 310°C because the limitations on maximum SCR temperature will be exceeded.
Since with the high process heat output at Schwarze Pumpe a fuel utilisation efficiency of some 55 per cent is claimed, it may be considered that cogeneration of heat and steam could be a more economic approach to beating Carnot, but there are not that that many instances of adequately large heat loads existing close to the mine mouth sites adjacent to the coal resources.
With such constraints and the economic limitations inherent in using P91, P92, austenitic steel alloys, and nickel based alloys for which good creep and manufacturability data is only available from the nuclear industry, Siemens have placed increasing reliance on steam turbine design advances in their quest to break the magic 50 per cent thermal efficiency barrier.
Siemens has been promoting the once through boiler concept since they acquired the Mark Benson patent in the 1920s because it brought the capacity for high pressure boiler potential into the field of power generation for the first time. They claim responsibility for the spiral configuration of the evaporator tubes, and they have since further developed this vertical tube design with a view to reducing investment costs while at the same time introducing boilers operating at supercritical pressures in which the drum type boiler has been the dominant design.
By using rifled tubes with high heat transfer capacities, they were able to reduce mass velocity in the tubes as much as 2500 kg/m2s to less than 1200 kg/m2s. The popular reference for this design is the 550 MW Staudinger Unit 5. This operates at 260 bar, 545/562°C, still using low cost martensitic steels for the final stages of the HP and reheater heating surfaces.
Fully three dimensional, variable reaction blading with compound lean, a Siemens development designated 3DV – the already well publicised curved blades – is the main ingredient in optimising HP and IP turbine efficiencies. Relevant here is a dispute among turbine engineers which goes backmore than 100 years over the choice of stage reaction, i.e. the split of pressure drop and velocity increase between stationary blades and moving blades. Defined more scientifically, stage reaction is the ratio of the enthalpy drop in the rotating row to the enthalpy drop of a whole stage.
A reaction turbine is characterised by a stage reaction of 50 per cent, and the enthalpy drop is equally divided between across stator and rotor rows. The symmetry in enthalpy drop entails a symmetry in flow relative to stator and rotor flows and allows the same profile to be used for both blade rows. In this case, flow velocities and flow directions along the blade path are moderate and profile and secondary losses are quite low. Since half of the enthalpy drop occurs in the rotor row, the pressure differential across the rotor blades is rather high and exerts a large axial thrust on the rotor.
In order to compensate this axial thrust, a dummy balance piston is required in single flow designs. The leakage flow across this dummy piston reduces the turbine cylinder efficiency.
In impulse turbines with zero stage reaction, the total stage enthalpy is converted into kinetic energy in the stator row, while the rotor row merely deflects the steam without further acceleration. In this case different profiles must be used for stator and rotor blades because their flows are asymmetrical.
Flow velocities and flow deflection along the blade path are significantly larger than the equivalent reaction stages, incurring higher profile and secondary losses. Since the pressure differential across the rotor is much less than that for reaction turbines, a much smaller dummy balance piston with lower losses can be used.
Three dimensional blades have the potential to exploit the best of both worlds, and Siemens maintain that by the use of computer controlled five axis milling machines, any three dimensional blade that the design engineer can invent can now be manufactured.
For the first stages of HP and IP turbines, a special three dimensional blade with compound lean has been developed by Siemens.
In these stages, the secondary losses are significant due to the low volume flow rates and short blade lengths. Compound lean and twist both serve to reduce the secondary losses at the root and the tip of blade. The reaction of each stage is set individually and may vary between ten and 60 per cent.
Extensive measurements were performed on a four stage test turbine to confirm efficiency improvements of up to two per cent – a major gain – compared with conventional cylindrical blading.
Another feature of the Siemens KWU supercritical IP turbine is the vortex cooling principle. For rotor stress reduction in the steam admission zone, the rotor is built without an axial through bore. In addition, the temperature and stress conditions in the admission zone are minimised by the design.
There is a double "T" root configuration for the two first stages with tilted first row stationary blades mounted on a heat shield ring with a cooling passage beneath the root. Cooling steam flows through tangential bores to form a vortex which cools the rotor since the steam temperature is reduced by its expansion to a high level of kinetic energy.
The rotor surface temperature distribution shows a temperature reduction of up to 16°C at the critical rotor section, which is enough to greatly improve the stress characteristics of the design.
Every one per cent of efficiency increase gained by the use of more highly supercritical steam conditions is reckoned to increase the output of a 660 MW turbine system in Germany by some 3 MW at normal input conditions.
The higher efficiency saves about 5600 t of coal per year, which in turn saves about 12 500 t of CO2 emissions. On top of this, life cycle costs in Germany are reduced by about $6 million.