An externally heated Brayton cycle turbine might be used to generate power from heat recovered from the exhaust of a conventional internally fired Brayton gas turbine. However the Brayton turbine is very sensitive to ambient conditions and to the turbine inlet temperature, which in this application is dependent on the exhaust temperature of the primary turbine. These factors limit its use as a backing device, and this article received little support. This article describes ways in which this limitation can be circumvented, resulting in an ability to produce as much as 50% more power with reductions of up to 30% in heat rate in some cases.
The key to improving the backing cycle is to raise the turbine inlet temperature to a level that allows generation of steam by the exhaust heat and injecting this steam into the compressor delivery duct. To raise steam at a rate of 10% of the compressor delivery and the same temperature and pressure requires a secondary turbine inlet temperature of about 1050F and a primary turbine exhaust temperature of 1100F or more.
The addition of steam allows a reduction in the air supply and compressor energy used for a given turbine mass flow rate. The specific heat of the steam/air mixture is approximately 9% higher than that of air, about the same as that of turbine exhaust gases, so with equal mass flow rates on both sides of the heat exchanger a uniform temperature differential can be maintained throughout the heat transfer process. The turbine temperature drop for a given heat drop will also be reduced compared to that for dry air. Table 1 compares conditions through two backing turbine systems, one using dry air and the other with 10% steam injection. The heat recovery in both cases is nearly the same and is based on a temperature gain of 680F across the heat exchanger. The primary turbine was a 2.814 MW unit with an exhaust temperature of 1080F, turbine inlet temperature 1800F and pressure ratio 6 with 23% efficiency, representing a down rated reference unit (PR12).which is described later.
Table 1 Comparison of Backing Turbines with and without Steam Injection
Parameter Dry Air 10% Steam
Air mass flow rate lbs 37 30
Compressor power (PR=4) kW -3055 -2477
Steam injected lbs/sec 0 3.0
Cp in Heat Exchanger and Turbine 0.28 0.311
Heat recovery from primary turbine Btu/sec 7045 6979
Cp/Cv in turbine 1.36 1.35
Turbine output (ER=1/3.8) kW 4262 4153
Gross power generated kW 1207 1676
Despite a slightly lower heat input the unit with steam injection produces nearly 40% more power. This is subject to a slight correction for the back pressure of the steam generator. Most of the improvement is due to the reduced compressor load.
Application to existing Gas Turbines
A key factor in the performance and economics of a backing system is the temperature of the heat source. For low temperature heat sources a Rankine cycle has the advantage that the principal heat demand is for evaporation of the working fluid at a constant low temperature. For organic Rankine cycles the chemical stability of the working fluid does not allow full advantage to be taken of high temperature sources. The Brayton cycle is a closer match to the heat rejection phase of the primary turbine. In effect the backing turbine extends the turbine expansion process. However the Brayton cycle does not perform well at low turbine inlet temperatures, and as pointed out earlier steam generation by the exhaust becomes less practical with heat sources (primary turbine exhaust) below 950F. For example the Solar Taurus 65 with an exhaust temperature of 1021F would be able to provide about 7.5 lbs of steam per pound of air in the backing turbine and this would allow an estimated increase of 1540kW in output after allowing for the additional back pressure on the primary turbine. Overall the output would increase by 19.6% to 7540kW at 38.5% efficiency compared with the rated 6300kW at 32.9 %.
In many cases the gas turbine exhaust temperature can be raised by increasing the firing temperature or reducing the turbine expansion ratio. Such changes are routinely made in the design of combined cycle units and recuperated gas turbines, and would be beneficial in the use of the steam injected backing system. Another approach would be to introduce duct firing after the heat exchanger without any change to the gas turbine. Duct firing extends the power range with a further increase in efficiency and would be a convenient way to increase the total power output to the limits imposed by the turbine materials. In the case of the Solar Taurus this could raise the total output 30% to 8200kW at 39.7% efficiency.
A comparison can be made between two approaches to design of a compound Brayton cycle backing unit based on a primary gas turbine that would not ordinarily be a good candidate for addition of a Brayton cycle backing unit. The reference turbine in this example is a hypothetical design with characteristics similar to the Maschproekt GT2500 turbine which has a PR of 12 and a low exhaust temperature. For purposes of analysis a number of assumptions were made to produce a realistic reference turbine performance matching that of the GT2500 but these assumptions should not be taken as representative of the GT 2500 itself. The assumptions made it possible to introduce hypothetical but realistic design changes and assess their effects. These changes may not be applicable in practice to the GT2500. The principal design parameters are listed in Table 2. A de-rated version of this engine was produced by omitting the radial compressor stage and one turbine stage. This is described in Table 3.
Table 2 Reference Turbine Parameters
Compressor: axial stages followed by one radial stage with PR=2 to give PR=12 overall 87% efficiency. Mass flow rate 32.5 lbs/sec
Turbine:3 stage, efficiency 88%
Turbine cooling air: 3lbs/sec bypassing combustor and turbine stages
Combustor: Pressure drop 5% delivery temp 1800F, 9760 Btu/sec
Gross shaft power after 1% mechanical loss: 2,944kW
Exhaust: 33lbs/sec at 828F
Efficiency:28.6%
Table 3 De-rated Turbine
Compressor: PR=6, 87% efficiency, mass flow rate 32.5lbs/sec
Turbine: 2 stage, 88% efficiency
Turbine cooling air: 3lbs/sec
Combustor: Pressure drop 5%, delivery temp 1800F, 11464 Btu/sec
Gross shaft power after 1% mechanical loss 2,814kW
Exhaust: 33 lbs/sec at 1,080F
Efficiency: 23%
Combining the outputs of the de-rated reference turbine and the steam-injected backing turbine (Table 1) gives a total of 4349kW at 35.9% efficiency after correction for back pressure of the heat exchanger. This combination gives a higher efficiency than adding a backing system to the unmodified reference turbine except at outputs below about 50% of full power. Addition of the backing turbine raises the maximum power of the de-rated turbine compared to an increase for the reference turbine with supplementary firing. The overall power output of the latter combination is slightly higher. This result is specific to this particular case and is a function of the reference turbine design, but it illustrates the range of performances that can be achieved from a single reference design.
Steam Generation
The heat available from the backing turbine exhaust with the heat demand for steam generation can be compared. For the case presented it shows that for 10% steam/air ratio generous temperature differentials are available It also shows the need to reduce this ratio for lower exhaust temperatures, which is the reason for considering supplementary firing. With an exhaust temperature of 500F only half as much steam could be produced.
Specific Design Configuration
Although the concept outlined above is for application as an add-on it is clearly more readily optimized when conceived from the outset as a compound system. The range of possibilities is illustrated by the two cases compared before, neither of which can be considered as an optimal arrangement. As is the case with steam/gas turbine combined cycles performance and cost can be made more competitive by exercising design flexibility throughout the design process while using the background provided by production gas turbines.
The compound turbine concept should be capable of attaining efficiencies of 38% which is well above that attainable with organic Rankine cycle backing units, and it is applicable to all turbine sizes. The critical design constraint is in the heat exchanger material, but available materials are capable of achieving valuable performance improvements with conservative design margins, and there are less rigorous spatial and size constraints than those presented by recuperative gas turbines. Because steam is used only in the air turbine it does not pose any combustion problems and in this respect there is no technical limitation on the amount of steam that can be used unless duct firing is necessary. With appropriate choice of gas turbine design parameters duct firing is not essential and the output gains are realized with no additional emissions.
Conclusion
This analysis demonstrates that the use of Brayton cycle based backing systems are capable of adding to the output and efficiency of conventional gas turbines with significant reductions in atmospheric emissions for a given unit capacity. They require no cooling system and no major technology breakthrough. A detailed design study is required of a compound system intended specifically to exploit this concept and to produce a realistic economic assessment.
Don Anson is Research Leader, Battelle Memorial Institute
