The International Steam Pages


Researching a Higher Efficiency Single-Turbine Steam Locomotive

Between 1920 and 1958, steam turbine locomotives were operated in Sweden, Germany, the UK and the USA. All suffered the problem of low efficiency at part-load operation. The American steam-turbine-electric locomotives were plagued with additional problems, such as turbine breakage during train coupling operations and electrical system failures. Only near the end of steam locomotive operation did some research appear that shed light on why steam turbine locomotives operated at extreme low levels of part load efficiency.

Research into supersonic compressible fluid low got underway after 1946, when the concept and existence of a "sound barrier" was first discovered. This phenomena not only affected high-speed aircraft. It also occurred when fluids flowed through restrictions in piping systems. When this happened, the mass flow rate of a compressible fluid like air or steam was at its maximum in the restriction. This is the "choked" nozzle. Lowering pressure levels downstream had no effect on either the mass flow rate or the constant ratios that developed between downstream and upstream pressures and temperatures.

In a steam turbine locomotive, the restriction occurred in the throttling valve. At very high power and throttle settings, a high downstream back-pressure prevented the valve from "choking". The maximum turbine power output was restricted by the steam mass flow rate through the smallest area of the inlet stators. The steam flow from the "choked" stator exit was at sonic speed as it entered the first row of turbine blades. The turbine was at maximum efficiency and maximum power.

As power demand declined, the throttle setting was cut back. As the downstream pressure behind the throttle valve declined, the "choking" phenomena would occur at the valve. A locomotive operating at 310-psig boiler pressure would only have 170-psig downstream of a "choked" valve, over a wide range of part-load mass flow rates of steam. The downstream steam temperature would also drop to 600-degrees F, from a boiler/superheat temperature of 800-degrees F. None of the earlier steam turbine locomotives were equipped with gauges to indicate downstream pressures and temperatures.

Brayton Turbine Precedents:

Like steam turbines, gas turbines also operated at low efficiency levels at part-load power settings. A precedent was set during the early 1970's when an open-cycle 2-shift Brayton turbine (gas generator and output turbine separate from each other) was equipped with a turbo-charger. Exhaust from the output turbine energised the turbo-charger, which compressed incoming air (raising its dynamic pressure) before the air entered the compressor of the gas generator. The overall result was optimal efficiency available from 25% of output to maximum power.

A similar result was achieved by the Escher-Wyss division of Sulzer, involving a closed-cycle Brayton turbine which used air as its working fluid. In a closed cycle system, it is possible to vary the internal air pressure to improve overall engine efficiency. This enables a closed-cycle Brayton to deliver optimal efficiency from 20% of output to near maximum, using a single-shaft system (where compressor, turbine and load were all coupled together). Some of the these Brayton system precedents could be transferred to steam turbines, to raise part-load efficiency.

Steam Turbine Power:

Several methods are possible to improve the part-load efficiency of steam turbine powered locomotives. One option is 4-turbines in a 1:2:4:8 power ratio, yielding 15-equal step power settings. Another alternative is to modify the steam pressure regulation system between the boiler and the turbine, while turbo-charging would also be an option. A modern steam-turbine-electric locomotive would have to use a very rugged turbine, a very short turbine that could be mounted transversely so as to avoid the blade breakage problem that plagued an earlier generation of s-t-e locos. The turbine design would need to such that the problem of blade erosion from high-speed droplets of water is either minimised or eliminated.

The radial-flow model CFR5 steam turbine from Kuhnle, Kopp & Kausch (KK&K http://www.agkkk.de/english/index.htm) is rated at 5000-Kw (6700-Hp) at 14,000-RPM's, at a steam pressure of 940-psia (65-bar a) and a temperature of 480-degrees C (900-degrees F). It could drive electrical generation gear is best mounted transversely in a locomotive carbody to avoid turbine blade damage during rough coupling operations. A mechanical transmission may also be used. To manage the problem of "choking", a water-tube boiler of 1800-psia would be used for the turbine. The "choking" pressure ratio for steam is 54.05% and the corresponding temperature is 85% of absolute, that is, downstream of the "choked" valve, steam pressure would be (1800 x 0.54) 972-psia and the temperature would be 367-degrees C (700-degrees F).

A CFR5 turbine would require some 16-lb/sec of superheated steam at maximum output. The water pump (88%-efficiency) and its motor (92%-efficiency) would require 150-Hp (112-Kw) of power to deliver 1800-psia of water at 16-lb/sec to the boiler, or 2.25% of maximum engine power. A variable pressure accumulator tank would be located immediately downstream of the water-tube boiler, acting as a reservoir of superheated steam. Outlet "choking" nozzles (valves) would be located the downstream (exit) end of the accumulator tank. These nozzles would enable a "back-pressure" to exist at the entrance to the accumulator tank, preventing choking at the inlet throttle valve.

The two main nozzles would be sized to operate in "choked" mode and permit a maximum of 10-lb/sec and 6-lb/sec of steam to pass through each, when operating in their on-mode. All "choking" nozzles would feed steam into larger diameter pipes, through a sudden increase in cross-section area. This will ensure that a shock wave remains at the nozzle exits, over a wide range of upstream steam pressures and their resulting mass flow rates, while steam flows through the nozzle throats at sonic speed. Subsonic velocity steam consequently will flow downstream, into a short section pipe of gently increasing diameter (a diffuser section), then into a constant diameter pipe going through the superheater (to raise enthalpy) and into the turbine intake.

The presence of the "choking" valves at the accumulator tank exit would prevent choking from occurring between the boiler (superheater) and the accumulator tank. These valves would create a "back-pressure" in the accumulator tank, over a wide upstream pressure range. Solenoid or mechanically regulated inlet valves with cut-off control would regulate steam pressure leaving the boiler and entering the accumulator tank, from 1800-psi to 1000-psi. Pressure regulation would also regulate steam mass flow rate. At the exit of the accumulator tank and downstream of the "choking" valves, pressures would be at a constant 54% of upstream pressures, between 975-psia to 550-psia and over a wide range of steam mass flow rates flowing to the turbine. The combination of the cross-sectional area of the "choked" valve and the upstream pressure determines the mass flow rate leaving the "choked" valve. To manage material thermal stress levels, the inside surfaces of the "choking" valves would be lined with a ceramic like silicon-carbide or silicon-nitride.

The steam leaving the choking valves would recombine into a single pipe and flow through the superheater, for reheating. This will ensure that steam over a wide range of pressure and mass flow rates will enter the CFR5 turbine at 900-degrees F (480-degrees C). At lower pressure (higher velocity steam) and at the same temperature, the enthalpy (Btu/lb) or energy content of steam increases, as does its specific volume and flow speed into the turbine, contributing to improved part-load efficiency. Maximum turbine efficiency occurs when the steam velocity at the smallest cross section of the inlet stator is at or near sonic speed. This phenomena can occur over a wide range of steam pressures and related steam mass flow rates entering the turbine, giving a range of power output levels.

To further enable high steam flow speed at the turbine inlet and enhance low-power efficiency, "compound-choke" valves may be incorporated into the exit of the accumulator tank. This involves several choking valves in series, one after the other with a small tank between them. For two "choked" valves in series, the combined pressure ratio would be 29% (eg: 1000-psi upstream, 290-psi downstream), a temperature ratio at 73% of absolute and operating pressure range from 540-psia to 285-psia. A triple "choked" valve would yield a downstream pressure ratio of 15.7% (280-psia to 155-psia downstream, from 1800-psia to 972-psia upstream). The steam from these low pressure lines would also be re-superheated to raise enthalpy and increase steam flow velocity (higher specific volume at lower pressures) entering the turbine. All "choking" valves would either be on or off, depending on the required turbine output power levels. These valves would assist in regulating pressure levels and steam mass flow rates.

A wide range of pressures would be available to the turbine from 155-psia to 975-psia, all at 900-degrees F (480-degrees C). At lower pressures and lower mass flow rates, the specific volume and the enthalpy of the superheated steam would be high, allowing for high steam velocities entering the turbine. High-speed superheated steam entering the turbine would enhance its part-load efficiency. Theoretically, the turbine could operate at optimal efficiency levels from maximum output down to below 30% of maximum power, based on precedents in closed-cycle Brayton technology. A computer may control fuel flow to the combustor, inlet valve and "choke" valve operation, thereby controlling power output from the turbine in response to demand from the electrical generation equipment that powers the locomotive.

Turbo-charged Low-Pressure Steam Turbine:

The exhaust pressure of the CFR5 is rated between 1-11 bar a (15-psia to 160-psia). At lower turbine power output levels, turbo-charging of a low-pressure turbine operating in a compound expansion cycle with the CFR5, may assist in raising overall part load efficiency. Some mild re-heating of the exhaust would assist in improving turbo-charger performance, especially at part-loads below 54%of maximum output. Exhaust steam from the CFR5 would flow to the turbo-charger compressor, then to the superheater before entering the low-pressure turbine. The exhaust from the LP turbine would energise the turbo-charger. Efficiency levels at the low end of turbine power output would especially benefit from energy being recovered by the turbo-charger.

The turbo-charger could operate with a pressure ratio in the range of 1.5:1 to 2:1. The pressure entering the LP turbine would be raised by some 50% and the LP steam temperature by 10% to 18% (absolute scale). Additional heat from the superheater would further raise enthalpy and specific volume levels, allowing high velocity steam to flow into the LP turbine. The LP turbine could be specified to operate at its peak output in the range where the CFR5 operates at its lowest usable efficiency levels. This would enhance overall system part-load thermal efficiency.

Lower Pressure Boilers (using a diffuser):

The diffuser a nozzle with a small diameter entrance and gently flaring to at least 4-times that diameter at its maximum cross section. Beyond this point its diameter rapidly decreases to its exit, where an adjustable (variable area) "choking" valve would be located. Partially superheated high velocity steam flowing into the diffuser would slow down as the diameter increases along its length. At constant enthalpy, the absolute temperature will rise to a maximum ratio of 1.29:1 and the pressure to a maximum of ratio of 1.8:1 at the point of maximum cross-sectional area. The interior of the diffuser may need to lined with ceramic (silicon-nitride or silicon-carbide) to manage material thermal tolerance levels, if superheated steam entered. Alternatively, the steam leaving the diffuser "choke" valve may be used to "cool" the hottest parts of the diffuser exterior (of a non-ceramic lined diffuser), before flowing to the turbo-charger. A third method would be to flow partially (low-temperature) superheated steam through the diffuser system, then fully superheat the steam to maximum temperature after leaving the "choke" valve and before entering the turbine.

The "choking" valve would reduce the highest temperature leaving the diffuser to its original (1.29 x 0.85 = 1.09) and pressure to 97% of original (1.8 x 0.54 = 0.972). For a 945-psia turbine being fed from a diffuser and an adjustable "choke" valve, a boiler pressure of 1000-psia to 1200-psia would be needed. A variable area throttle control valve regulating steam mass flow rate entering the diffuser, would be at the diffuser entrance or upstream of it. The adjustable/variable-area downstream "choke" valve would be mechanically linked to the adjustable upstream throttle and prevent the throttle from "choking", by providing the necessary downstream "back-pressure". Resetting the throttle valve's cross-sectional area re-adjusts the mass flow rate of steam passing through it. The "choke" valve cross-sectional area resets the steam pressure level flowing downstream to the turbine, gradually reducing steam pressure as steam mass flow rate is reduced. In some applications, a compound diffuser may be needed (one downstream of the other) to enable a very wide range of downstream steam pressures and related mass flow rates.

A wide range of steam pressures and mass flow rates would be available downstream of the diffuser, due to the combined operation of the diffuser's inlet throttle and outlet "choke" valves. The two would be mechanically linked and be directly or indirectly controlled (computer control) from one throttle/power-level control lever in the locomotive driving cab, in both single and compound diffuser operation. In the latter case, a single diffuser would operate for higher pressures and steam flow rates whereas compound diffuser operation would be be used for lower steam pressures and corresponding flow rates.

At lower steam pressures, higher velocity steam leaves the "choke" valve and may need to be re-superheated to further raise its specific volume and flow velocity. This will enable superheated steam to enter into the turbine stator at near sonic speed, over a wide range of steam pressures and related steam mass flow rates. Under such conditions, the CFR5 turbine will be able to operate more efficiently over a wide range of power output. It may be able to operate to as low 25% of maximum output, while maintaining a high level of thermal efficiency. Closed-cycle Brayton systems operating on variable air pressure have been able to deliver high efficiency to as low as 25% of maximum output. A turbo-charged LP turbine may also operate in a compound expansion cycle with the lower pressure boiler and diffuser system, enhancing part-load efficiency.

Lower-Pressure Turbine option:

The 595-psia model AFA6 turbine from KK&K develops 5,000-Kw (6,700-Hp) at 11,400-RPM, uses 450-degree C (850-degree F). It may operate from a high-pressure (1100-psia) water-tube boiler, connected to an accumulator tank and downstream "choke" valves (to regulate steam pressure and mass flow rate), followed steam re-superheating to operate at optimal efficiency over a wider power range. Alternatively it may also source steam from a 700-psia boiler, regulating steam pressure and mass flow rate by using a ceramic-lined diffuser, one equipped with upstream variable throttle valve linked to a downstream variable "choke" valve, followed by steam re-superheating ahead of the turbine. The AFA6 will need to be transversely mounted in the locomotive carbody, driving electrical gear, to reduce the risk of turbine damage due to rough railway operations. To enable efficient operation at very low power levels (below 30% of the AFA6's maximum) over prolonged periods, the 1600-Kw (2100-Hp) KK&K model CFA4 may be used as an auxiliary turbine, as it operates at the same steam pressure as the AFA6. Other low-power auxiliary engines may include a Quasiturbine or a Rand Cam.

Conclusion:

The part-load thermal efficiency of the classical era steam turbine locomotives by re-superheating the throttled steam, between the throttle valve and the turbine inlet. This would have raised the enthalpy levels as well as the specific volume (inverse of density) of the lower pressure throttled steam, causing the flow velocity to accelerate into the stator at the turbine inlet. A "choked" stator would have been maintained at the turbine inlet, over a wider range of superheated steam pressures and mass flow rates, resulting in higher thermal efficiency level over a wider high-end power range (60% to 100% power output). The methods described above are intended to enable higher part-load efficiency levels over an even wider power range (20% to 100% power output). A wide variety of alternate combinations of nozzle, throttle valve and "choke" valve layouts and designs can allow variable pressure downstream flow rates, from a constant upstream steam boiler pressure. In most cases, the downstream steam flow would have to be re-superheated prior to entry into the turbine stator. The research needed to improve the efficiency of the classical steam turbine locomotives, only appeared long after they had all been retired from service.

As an alternative to turbo-charging a LP turbine in a compound expansion system, the main power turbine itself may be turbo-charged during lower-level part-load power operations. Both the CFR5 and AFA6 turbines have the same exhaust pressure range, with the higher pressure occurring at part-load operation. Turbo-charging the main turbine, with turbo-charger compression occurring between the"choke" valves and the superheater, would raise steam inlet temperature and raise the stagnation pressure by accelerating the steam flow velocity downstream of the "choke" valve. A (superheated) steam inlet velocity near sonic speed at the turbine stator will maximise turbine efficiency over a wide range of steam densities, steam pressures, steam mass flow rates and at part-load power output levels.

Heat supply to the boiler would utilise gas producer combustion system (GPCS) technology. Water re-circulation systems would use multiple expansion valves to convert the exhaust saturated steam to hot water (higher-pressure saturated steam), which will be cooled in a series of radiators located on the locomotive power generating car, the water car as well as on the fuel-carrying car (2-tenders). Condensing exhaust steam for re-use would be a major challenge on future steam locomotive designs. The range of solid fuels would include clean coal technologies as well as biomass and agricultural pellet fuels, including the renewable fuel type used at the UK's Thetford Power Station. New research in metallic polymers may even enable long-distance fireless steam locomotive operation. Modern computer control technology would enable one-person locomotive operation, as well as optimise energy efficiency.

Harry Valentine, Transportation Researcher. harrycv@hotmail.com


Click here to return to the modern steam locomotive developments page.


Rob Dickinson

Email: webmaster@internationalsteam.co.uk