Original Date of Article : Early 2007
The 700°C Steam Plant has been promoted as a means of attaining coal to electricity efficiencies in the 53 to 55% range. The development of this plant has to be seen in the context of the recent efforts to increase power plant efficiency by increasing steam temperatures. For about 20 years steam temperatures had been held at 540°C, but from about 1990 temperatures have gradually been increased to about 580°C, through the introduction of strong martensitic steels containing about 9% chromium. The expectation is that over the next 20 years it will be possible to evolve more advanced plants in which the steam temperature will be increased in a series of steps. The purpose of this note is to suggest that while is might be possible to climb part of the way up the “efficiency hill” it will be impossible to reach the summit.
The first question that needs to be asked is what are the benefits of attaining a level of efficiency of 53%, even if this is possible? As efficiency levels increase to the 50% range, a one percent increase in thermal efficiency only gives a 2% fuel saving. Hence the relative improvement between the steps becomes fairly marginal. Will the expenditure on the plant and the development effort be cost effective? Might it be better to stop at a plant in which the steam temperature was in the 650-675°C range?
How feasible will it be to actually get an efficiency of 53-55%? We should immediately discount arguments which depend on the Carnot cycle. This is a purely theoretical concept that has nothing to do with any form of heat engine. In the real world, steam plants are based on a heavily modified form of the Rankine cycle. But in practice no steam plant is based on any of the textbook cycles. Hence process flow modelling must be used to predict plant efficiencies. Any efficiency predictions produced using modelling depend on a large number of factors, which include steam turbine isentropic efficiencies, whether single or double reheat sets of turbines are used, the number of feed heaters, and pressure drops in the connecting pipes boilers and heat exchangers, etc. During the 20 years in which steam temperatures were constant, plant efficiencies went up nevertheless, because these “component” efficiencies were improved. It is doubtful whether there is much chance for much further improvement. For example, the isentropic efficiency of steam turbines is now over 90%. But if there is room for improvement, this can be applied to all future designs of plant, regardless of the steam temperature.
The one real innovation that could be applied to advanced plant is to use double reheat, which is rarely used at the present time. Here the superheated steam, after passing through the high pressure turbine, is passed to back to another superheater called a reheater, in which it is reheated to back to superheat temperatures again. This is normal practice, giving a small increase in efficiency, the main advantage being an increase in output. But to take advantage of reheating, the pressure to the high pressure turbine needs to be increased. However, the original proposal in the 700°C plant concept was to use double reheat, that is, after the steam has left the first reheat turbine it is reheated again before going to a second reheat turbine. The problem is that the steam pressures would have had to be extremely high for this to be worth while. Because of the need for such a combination of high pressures and temperatures, plus turbine design issues, the double reheat proposal has been dropped. But only a few publications seem to acknowledge that it will be difficult to attain the original targets because of this change.
Process flow modelling does show that an appropriate combination of steam pressure and temperature is needed to get the target efficiency in any plant. What is not commonly understood is that as the steam temperature increases, the steam pressure must increase at an exponential rate. This is required to ensure that after full expansion of the steam through all the turbines its temperature will have dropped to just above the temperature of the condenser. For steam temperatures of 700°C pressures will need to be at least 340 bar, even if only one stage of reheat is used. Ideally it needs superheater alloys to operate at higher stresses than the best plants of today, where the pressures are around 250 bar, if tube wall thicknesses are to remain the same. This is a big challenge given the increase in temperature.
This stress/ temperature conditions will require the use of advanced austenitic or even nickel based alloys in the superheater. The actual metal temperature will be around 750°C. The level of design stress in the superheater tubes can be adjusted to some extent by increasing the tube wall thickness, but this will be at the expense of decreased resistance to thermal stress and much increased materials and fabrication cost.
A target rupture stress has been set by the proponents of the 700°C plant of 100 MPa. This is higher than the value for T22 for the older 540°C plants, where the steam pressure was around 160 bar. But the figure is below that for P91 steel which was initially used as a replacement for T22. (The respective values at 580°C are T22: 34MPa and P91:120 MPa). Note that the steam pressure in the 700°C is at least twice as high as in the 540°C plants, so ideally the stress rupture value needs to be around 200 MPa.
But how practical is it to reach even the 100 MPa target at 750°C? Some years ago I pointed out that the ability to obtain high strength in austenitic alloys is limited by the speed of dislocation climb. Since this is governed by the rate of diffusion, stress rupture values drop off rapidly with temperature. The paper was summarised in the EPRI/DOE Materials and Components Newsletter, and included a set of predictions for what was going to be the maximum attainable strength over the range 650-1100°C. My prediction for 750°C was 97.2 MPa. At the present time there is no austenitic alloy reaching this prediction. The best is a modified form of Inco 617 which is really a high nickel austenitic.
I did point out in the paper that improved properties could be obtained if an alloy was available that could be precipitation hardened at a higher temperature. One such has been identified, namely Inco 740, which appears to have a strength of about 150 MPa at 750°C. This is lower than what is really needed, and it also appears that the age hardening temperature is very close to the operating temperature. How stable will be the alloy in service remains to be seen. In addition during welding, parts of the heat affected zone of the alloy will be subject to over-ageing, so there will be a tendency for welds to fail because of Type IV cracking.
In conclusion I am against R&D on the 700°C steam plant, because I do not think it will save much more coal than less ambitious concepts, I think that the thermodynamic estimates of efficiency are incorrect, that the operating temperatures and sophistication of the plant have been downgraded since the initial proposals, and that the proposed materials of construction will be inadequate. Fred Starr: 26th Jan 2007