EVOLUTION OF AN INDUSTRY
Since the turn of the twentieth century, the steam turbine has evolved from an experimental device to the major source of electrical generation. Practical steam turbine inventions coincided with the development of direct-current electric dynamos first used to power arc-lighting street systems. In the United States, the first central station to provide electrical lighting service was Thomas Edison's Pearl Street Station in New York City in 1882. Powered by 72 kW of steam engines, it served 1,284 16-candlepower dc lamps. This installation demonstrated the feasibility of central station electricity. Initially, Edison's system needed a large number of scattered power plants because it utilized direct current, and dc transmission was uneconomical over large distances. In 1885, George Westinghouse's Union Switch & Signal Co. acquired rights to manufacture and sell a European-design transformer, and the company then developed alternating-current distribution capability to utilize its transformers, which made longer-distance transmission of electricity practical. The Westinghouse Electric Co. was formed to exploit this device. By 1900 there were numerous dc and a few ac generating stations in the United States, all with reciprocating steam engines or hydraulic turbines as prime movers. However, the ac technology quickly became a primary factor in stimulating the development of power generation.
Although they were reliable, the early steam engines were huge, heavy devices that were not very efficient. Nearly all the companies in the electric equipment business seized the opportunity to develop the steam turbine as an alternative. In 1895, Westinghouse acquired rights to manufacture reaction turbines invented and patented in 1884 by the English inventor Charles Algernon Parsons. Allis-Chalmers also acquired rights to manufacture under Parsons' patents, so early machines of these two manufacturers were quite similar. In 1887, the General Electric Co. (founded by Edison) entered into an agreement with Charles Curtis to exploit his steam turbine patent.
The Curtis and the Parsons turbine designs are based on different fundamental principles of fluid flow. The Curtis turbine has an impulse design, where the steam expands through nozzles so that it reaches a high velocity. The high-velocity, low-pressure steam jet then impacts the blades of a spinning wheel. In a reaction turbine such as the Parsons design, the steam expands as it passes through both the fixed nozzles and the rotating blades. High pressure stages are impulse blades. The high-pressure drops quickly through these stages, thus reducing the stress on the high pressure turbine casing. The many subsequent stages may be either impulse or reaction designs.
STEAM TURBINE CYCLES
The basic function of a steam turbine is to efficiently convert the stored energy of high-pressure, high-temperature steam to useful work. This is accomplished by a controlled expansion of the steam through stages consisting of stationary nozzle vanes and rotating blades (also called buckets by one major manufacturer). The size and shape of the nozzle vanes and rotating blades are such as to properly control the pressure distribution and steam velocities throughout the turbine flow path. Blading improvements have increased turbine cycle efficiency by reducing profile losses, end-wall losses, secondary flow losses, and leakage losses. Use of tapered twisted designs for longer blades reduces losses on the innermost and outmost portions of the blades.
A complete turbine generator unit could consist of several turbine elements connected in tandem on a single shaft to drive a generator. To extract as much energy from the steam as possible, as it decreases in temperature and pressure in its passage through the machine, the typical arrangement could include a high-pressure (HP), an intermediate-pressure (IP), and one or more low-pressure (LP) elements, as illustrated in Figure 1.
The HP, IP, and LP turbines may be either single-flow or double-flow designs, depending upon the volume of steam utilized. In a single-flow turbine, the total volume of steam enters at one end and exhausts at the other end. The double flow is designed so the steam enters at the center and divides. Half flows in one direction, and half in the other direction into exhausts at each end of the turbine.
The basic steam cycle for a steam turbine installation is called a Rankine cycle (named after Scottish engineer and physicist William John Macquorn Rankine). This cycle consists of a compression of liquid water, heating and evaporation in the heat source (a steam boiler or nuclear reactor), expansion of the steam in the prime mover (a steam turbine), and condensation of the exhaust steam into a condenser. There is a continuous expansion of the steam, with no internal heat transfer and only one stage of heat addition. By increasing the pressure and/or temperature and decreasing the heat rejected by lowering exhaust temperature, cycle efficiency can be improved.
Weir patented a regenerative feedwater heating cycle in 1876. The regenerative Rankine cycle eliminates all or part of the external heating of the water to its boiling point. In this cycle, a small amount of expanded steam is extracted from a series of pressure zones during the expansion process of the cycle to heat water in a multiplicity of heat exchangers to a higher temperature. Theoretical and practical regenerative cycles reduce both the heat added to the cycle and the heat rejected from the cycle.
Reheat involves steam-to-steam heat exchange using steam at boiler discharge conditions. In the reheat cycle, after partially expanding through the turbine, steam returns to the reheater section of the boiler, where more heat is added. After leaving the reheater, the steam completes its expansion in the turbine. The number of reheats that are practical from a cycle efficiency and cost consideration is two.
EVOLUTION OF THE STEAM TURBINE
The first central station steam turbine in the United States was built for the Hartford Electric Light Co. in 1902. Steam conditions for this 2,000-kW unit and similar units were approximately 1.2 MPa (180 psig) and 180°C (350°F). The evolution of steam turbine power generation in the United States is summarized in Figure 2. Plotted against time are the maximum inlet steam pressure and temperature, along with plant thermal efficiency, and maximum shaft output in megawatts. The steady increase in steam turbine inlet pressure and temperature achieved an increase in plant thermal performance. From 1910 to 1920 steam turbine generators were manufactured in the 30- to 70- MW range. By 1945, the median unit sold in the United States was still only 100 MW. By 1967 the median unit had increased to 700 MW, with a peak of 1,300 MW for several fossil-fueled units placed in service in the 1970s. (A 1,300-MW unit can generate enough electricity to supply the residential needs of more than 4 million people.) During the first fifty years of the twentieth century, inlet steam pressure and temperature increased at an average rate per year of 0.3 MPa (43 psi) and 7°C (13°F), respectively. Until the early 1920s, throttle pressures were 1.4–2.1 MPa (200–300 psi), and throttle temperatures did not exceed 600°F (315°C). Above 450°F (230°C), cast steel replaced cast iron for turbine casings, valves, and so on.
Figure 3 shows the thermal performance evolution of the steam cycle as a function of material development and cycle improvements, starting in 1915. By the early 1920s, regenerative feedheating was well established. Reheat cycles came into use in the mid-1920s. At the throttle temperature of 370°C (700°F) that was current when the pioneer 4.1- and 8.2-MPa (600- and 1,200-psi) units went into service, reheat was essential to avoid excessive moisture in the final turbine stages. As temperatures rose above 430°C (800°F) molybdenum proved effective. Using carbon-moly steels, designers pushed temperatures beyond 480°C (900°F) by the late 1930s. As a result of rising throttle temperatures, reheat fell out of use. By the late 1940s, reheat was reintroduced to improve plant efficiency, and second reheats appeared by the early 1950s.
Over the years, exhaust area was a major limitation on size. The earliest answer was the double-flow single-casing machine. Cross-compounding, introduced in the 1950s, represented a big step forward. Now the speed of the LP unit could be reduced. Thus, the last-stage diameter could be greater, yielding more exhaust annulus area.
Continued advances in metallurgy allowed inlet steam conditions to be increased, as illustrated in Figure 3. In 1959 this progress culminated with the Eddystone 1 unit of the Philadelphia Electric Co. and Combustion Engineering. With initial steam conditions of 34 MPa and 650°C (5,000 psi and 1,200°F) and two reheats of 1,050°F (570°C), Eddystone 1 had the highest steam conditions and efficiency (40%) of any electric plant in the world. The generating capacity (325 MW) was equal to the largest commercially available unit at the time. Eddystone 1 has operated for many years with throttle steam conditions of approximately 4,700 psi and 1,130°F (32 MPa, 610°C), and achieved average annual heat rates comparable to today's best units.
Figure 3, initially published in 1954, summarizes the evolution of the steam cycle from the year 1915 through the steam conditions of Eddystone 1 (average station heat rates were used) and a projection of where the industry might be by the year 1980. The relationship of operating steam pressures and temperatures to available materials in this figure indicates how the increase in pressure and temperature is dependent upon metallurgical development. The magnitudes of heat rate gains resulting from the application of various kinds of steam cycles also are shown.
Since the early 1960s, advanced steam conditions have not been pursued. In the 1960s and early 1970s there was little motivation to continue lowering heat rates of fossil-fired plants due to the expected increase in nuclear power generation for base-load application and the availability of relatively inexpensive fossil fuel. Therefore the metallurgical development required to provide material "X" for advanced steam conditions was never undertaken.
Raising inlet pressure and temperature increases the cycle's available energy and thus the ideal efficiency. However, pressure increases reduce the blade heights of the initial stages and decrease ideal efficiency, offsetting some of the ideal improvement, unless unit rating is increased commensurately. Based on potential heat rate improvement, there is no reason to raise the turbine steam conditions above 48.2 MPa, 760°/760°/593°C (7,000 psi, 1,400°/1,400°/1,100°F).
NUCLEAR POWER APPLICATIONS
The first steam turbine generator for nuclear power application was placed in service at Duquesne Light Co.'s Shippingport Station in 1957. Initially rated at 60 MW, it had a maximum turbine capability rating of 100 MW. This was the first of a series of 1,800-rpm units, which were developed from a base design and operating experience with fossil-fuel machines dating back to the 1930s. Inlet steam conditions at maximum load were 3.8 MPa (545 psi), with a 600,000 kg (1.3 million-lb.) per-hour flow. This single-case machine had a 1-m (40-in.)-long last-row blade. Second-generation nuclear turbines introduced reheat at the crossover zones for improved thermal performance. Nuclear turbines, ranging in size from 400 to 1,350 MW, have used multiple LP exhausts.
Since moisture is a major concern for turbines designed for nuclear operation, a number of erosion-control techniques were used in the LP turbine. For example, adequate axial spacing between the stationary and the rotating blades minimizes blade erosion. Moisture is removed at all extraction points in the moisture region.
Nuclear turbines designed for use with a boiling-water reactor will be radioactive. Radioactivity could build up in the turbine because of the accumulation of corrosion products. A fuel rod rupture could result in highly radioactive materials entering the turbine. Therefore internal wire-drawing-type leakage paths, ordinarily unimportant in steam turbine design, must be eliminated as much as possible. Where it is impossible to eliminate, the surfaces forming the leakage paths should be faced with erosion-resistant materials deposited by welding.
FUTURE ROLE FOR STEAM TURBINE POWER GENERATION
In 2000, the power generation cycle of choice is the combined cycle, which integrates gas and steam cycles. The fuel of choice for new combined-cycle power generation is generally natural gas. Steam turbines designed to support the large gas turbine land-based power generation industry are in the range of 25 to 160 MW. Steam conditions are in the range of 12.4 MPa (1,800 psi) and 16.6 MPa (2,400 psi) and 538°C (1,000°F) and 565°C (1,050°F).
Large natural gas-fired combined cycles can reach a cycle efficiency of 58 percent, higher than a typical steam turbine power plant efficiency of 36 to 38 percent. During the 1990s, experience with optimizing advanced supercritical steam turbine cycles led to thermal cycle efficiencies in the 40 to 48 percent range for units rated at 400 to 900 MW that are existing or planned to be located in Japan, Denmark, Germany, and China.
In Denmark, two seawater-cooled 400 MW units, operating at a steam inlet pressure of 28.5 MPa (4,135 psi), have an efficiency of about 48 percent. Chubu Electric, in Japan, has been operating two 700 MW units since 1989 with inlet steam conditions of 31 MPa (4,500 psi). The efficiency gains of these two units is about 5 percent more than that of previous conventional plants of comparable size at 45 percent.
With an estimated 400 years of coal available for future power generation, coal-powered steam turbines are expected to continue to dominate global electricity fuel markets.
In the United States, coal had a 57 percent share of the electric power fuel market in 2000, up from 46 percent in 1970. This amounts to 430,000 MW generated by steam turbines that are fueled with coal. When considering other sources of generating steam for electric power—such as nuclear reactors, gas- or oil-fired broilers, and waste heat from gas turbines—steam turbines now comprise more than 600,000 MW of capacity, or approximately 75 percent of all generating capacity in the United States.
Since 1900, manufacturers have made many step changes in the basic design of steam turbines. New technology and materials have been developed to support the industry's elevation of steam conditions, optimization of thermal cycles and unit capacity. Steam turbines will continue to be the principal prime mover for electricity generation well into the twenty-first century.
Ronald L. Bannister
Artusa, F. A. (1967). "Turbine and Cycles for Nuclear Power Plant Applications." Proceedings, American Power Conference 29:280–294.
Bannister, R. L.; Silvestri, G. J., Jr.; Hizume, A.; and Fujikawa, T. (1987). "High-Temperature, Supercritical Steam Turbines." Mechanical Engineering109(2):60–65.
Bennett, S. B.; and Bannister, R. L.; Silvestri, G. J., Jr.; and Parkes, J. B. (1981). "Current Day Practice in the Design and Operation of Fossil-Fired Steam Power Plants in the U.S.A." Combustion52(7):17–24.
Bennett, S. B., and Bannister, R. L. (1981). "Pulverized Coal Power Plants: The Next Logical Step." Mechanical Engineering103(12):18–24.
Church, E. F., Jr. (1950). Steam Turbines. New York: McGraw-Hill.
Keller, E. E., and Hodgkinson, F. (1936). "Development by the Westinghouse Machine Company." Mechanical Engineering58(11):683–696.
Miller, E. H. (1984). "Historical Perspectives and Potential for Advanced Steam Cycles." Electric Forum10(1):14–21.
Mochel, N. L. (1952). "Man, Metals, and Power." Proceedings, American Society Testing Materials 52:14–33.
Robinson, E. L. (1937). "Development by the General Electric Company." Mechanical Engineering 59(4): 239–256.
Silvestri, G. J., Jr.; Bannister, R. L.; Fujikawa, T.; and Hizume, A. (1992). "Optimization of Advanced Steam Conditions Power Plants." Transactions of the ASME, Journal of Engineering for Gas Turbine and Power 114:612–620.
Smith, D. (1999). "Ultra-Supercritical CHP: Getting More Competitive." Modern Power Systems19(1):21–30.
Somerscales, E. F. C. (1992). "The Vertical Curtis Steam Turbine." Newcomen Society of Engineering Technology Transactions 63:1–52.
Zink, J. (1996). "Steam Turbines Power and Industry." Power Engineering100(8):24–30.