by Vaclav Smil
Blades in Laval’s turbine are moved solely by impulse: there is no drop of pressure as the steam is passing the moving parts, and its velocity relative to the moving surfaces does not change except for the inevitable friction. Charles Parsons (figure 2.12) chose a very different approach in designing his successful steam turbine. As he explained in his Rede lecture,
It seemed to me that moderate surface velocities and speeds of rotation were essential if the turbine motor was to receive general acceptance as a prime mover. I therefore decided to split up the fall in pressure of the steam into small fractional expansion over a large number of turbines in series, so that the velocity of the steam nowhere should be great… I was also anxious to avoid the well-known cutting action on metal of steam at high velocity. (Parsons 1911:2)
Consequently, the steam moved in an almost non-expansive manner through each one of many individual stages (there can be as many as 200), akin to water in hydraulic turbines whose high efficiency Parsons aspired to match. Starting with a small machine, he could not avoid high speeds: he filed the key British patent (G.B. Patent 6,735) on April 23, 1884 (the U.S. application for rotary motor was filed on November 14, and U.S. Patent 328,710 was granted on October 1885) and proceeded immediately to build his first compound turbine. This machine (now preserved in the lobby of the Parsons’s Building at the Trinity College in Dublin) was rated at just 7.5 kW, produced DC at 100 V and 75 A, ran at 18,000 rpm, and had efficiency of a mere 1.6%. Many improvements followed during the decades before WWI (Parsons 1936).
FIGURE 2.12. There is no portrait of Charles Algernon Parsons dating from the early 1880s when he invented his steam turbine. This portrait (oil on canvas) was painted by William Orpen in 1922 and is reproduced here courtesy of the Tyne and Wear Museums, Newcastle upon Tyne.
Improvements concentrated above all on the design of blades and overcoming challenges of building a machine composed of thousands of parts that had to perform faultlessly while moving at high speeds in a high-pressure and high-temperature environment and do so within extremely narrow tolerances as clearances over the tips of moving blades are less than 0.5 mm (see figure 1.5). The first turbine had straight blading, for both rotating and guide wheels, which was later substituted by curved blades with thickened backs. Experiments showed that for the best efficiency the velocity of moving blades relative to the guide blades should be between 50% and 75% of the velocity of the steam passing through them.
Parsons also had to design a dynamo that could withstand high speeds and resist the great centrifugal force, and solve problems associated with lubrication and controls of the machine. The first small single-phase turbo-alternators for public electricity generation were ordered in 1888, and the first two of the four 75-kW, 4,800-rpm machines began operating at Forth Banks station in Newcastle upon Tyne in January 1890. Their consumption was 25 kg of steam per kilowatt-hour at pressure of 0.4 MPa (or conversion efficiency of about just over 5%), and they supplied AC at 1 kV and 80 Hz. In 1889 Parsons lost the patent right to his parallel flow turbine when he left Clarke, Chapman & Co., established C. A. Parsons & Co. and turned to designs of radial flow machines that proved to be less efficient.
A 100-kW, 0.69-MPa turbo-alternator that his new company supplied to the Cambridge Electric Lighting in 1891 was Parsons’s first condensing turbine. Parsons’s small turbines of the 1880s were noncondensing, exhausting steam against atmospheric pressure and hence inherently less efficient. The condensing unit was also the first one to work with superheated steam (its temperature raised above that corresponding to saturation at the actual pressure), albeit only by 10°C. In 1894 Parsons recovered his original patents (by paying merely £1,500 to his former employers) and proceeded to design not only larger stationary units but also turbines for marine propulsion.
FIGURE 2.13. An excellent engraving of Parsons’s first 1-MW turbogenerator (steam turbine and alternator) installed in 1900 at Elberfeld plant in Germany. Reproduced from Scientific American, April 27, 1901.
A rapid increase of capabilities then led to the world’s first 1-MW units: two of them, with the first tandem turbine-alternator arrangement, were built by C. A. Parsons & Co. in 1899 for the Elberfeld station in Germany to generate single-phase current at 4 kV and 50 Hz (figure 2.13). The first 2-MW (6 kV, three-phase AC at 40 Hz) turbine was installed at the Neptune Bank station in 1903, and Parsons designs reached a new mark with a 5-MW turbine connected in 1907 in Newcastle upon Tyne (Parsons 1911). That turbine worked with steam superheated by nearly 50°C at 1.4 MPa, and it needed only 6 kg of steam per kilowatt-hour, converting coal to electricity with about 22% efficiency. Parsons largest pre-WWI machine was a 25-MW turbo-alternator installed in 1912 at the Fisk Street station of the Commonwealth Edison Co. in Chicago. These large units converted about 75% of the steam’s energy to the rotation of the shaft.
American development of steam turbines did not lag far behind the British advances (MacLaren 1943; Bannister and Silvestri 1989). George Westinghouse bought the rights to Parsons’s machines in 1895, built its first turbine for Nichols Chemical Co. in 1897, and by 1900 delivered a i.5-MW machine to the Hartford Electric Light Co. In 1897 General Electric made an agreement with Charles Curtis (1860–1953), who patented a new turbine concept a year earlier. This turbine could be seen as a hybrid of Laval’s and Parsons’s designs: it is a pure impulse turbine, but the kinetic energy of steam is not extracted by a single ring but by a multipulse action at several (commonly two or three but up to seven) stages (Ewing 1911). Curtis machines eventually reached capacities of up to 9 MW, and they were installed at many smaller plants throughout the United States during the first two decades of the 20th century.
General Electric’s first turbine, a 500-kW unit installed in Schenectady, New York, in 1901, had a horizontal shaft, but its second machine of the same size was vertical, and the company soon began offering vertical Curtis turbines with capacities of up to 5 MW. Before GE abandoned the design in 1913, it shipped more than 20 MW of vertical-axis machines. The third American company that advanced the early development of steam turbine was Allis-Chalmers, originally also a licensee of Parsons’s machines. By 1910 the company was producing turbines with capacities of up to 4.5 MW. And during the first decade of the 20th century, Swiss Brown Boveri Corporation became the leading producer of turbines on the European continent (Rinderknecht 1966).
Consequently, it took less than two decades after the first commercial installation to establish steam turbines as machines that were in every respect superior to steam engines. Their two most desirable features were the possibility of very large range of sizes and unprecedented generation efficiencies. In just 20 years, the size of the largest machines rose from 100 kW to 25 MW, a 250-fold increase, and by 1914 it was clear that it would be possible to build eventually sets of hundreds of megawatts. Performance trials of steam turbines were conducted from the very beginning of their design, and Parsons’s (1911) figures show rapid gains of thermal efficiency. When his original numbers, expressed in terms of saturated or superheated steam per kilowatt-hour, are recalculated as percentages, they rise from about 2% for his 1884 model turbine to about 11% for the first low-superheat 100-kW machine in 1892 and to nearly 22% for the 5-MW, higher superheat design in 1907.
The efficiency for a 20-MW turbine installed in 1913 was just over 25% (Dalby 1920), which means that there was an order of magnitude gain in efficiency in three decades, clearly a very steep learning curve. In contrast, British marine engine trials of the best triple- and quadruple-expansion steam engines conducted between 1890 and 1904 showed maximum thermal efficiencies of 11–17% (Dalby 1920). This means that it made no engineering and economic sense to install steam engines in any large electricity-generating plant after 1904, if not after 1902. And so it was not surprising that America’s largest new station completed in 1902, the New York Edison plant fronting the East River, still had 16 large Westinghouse-Corliss steam engines (Anonymous 1902; see the frontispiece to this chapter). But it was sur
prising when in 1905 London’s County Council tramway power station in Greenwich began installing what Dickinson (1939:152) called a “megatherium of the engine world.”
The “megatherium” was the first 3.5-MW vertical-horizontal compound steam engine, housed in a building of “cathedral-like proportions.” This installation, so memorable for its poor engineering judgment, also provided the best contrast with inherently compact size and very low mass/power ratio of steam turbines. Greenwich machines were almost exactly as high as they were wide, about 14.5 m, and Parsons’s company used to distribute a side-by-side drawing of these giants and of its turbogenerator of the same power (3.5 MW), whose width was just 3.35 m and height 4.45 m. Even Parsons’s pioneering 100-kW turbine built in 1891 weighed only 40 g/W, or less than 20% than the best comparable steam engines. That ratio fell by 1914 below 10 g/W, and it was just over 1 W/g for the largest machines built after the mid-1960s (Hossli 1969).
Naturally, these declining power/mass ratios, especially pronounced with larger units (Hunter and Bryant 1991), translated into great savings in metal consumption and in much lower manufacturing costs per unit of installed power. Moreover, giant steam engines required additional construction costs for the enormous buildings needed to house them. Direct delivery of rotation power rather than awkward arrangements for converting the reciprocating motion was the most obvious mechanical advantage that was also largely responsible for lack of vibration that was often a problem with enormous piston engines.
The 20th century had witnessed a remarkable rise of turbine specifications very much along the lines that were set by Parsons during the first two decades of turbine development (Termuehlen 2001). Technical advances have been rather dramatic. Steam pressures rose by an order of magnitude by 1940, steam temperatures nearly tripled, and the highest ratings reached 1 GW by the late 1960s and eventually leveled off at 1.5 GW (figure 2.14). Three-phase alternators are now the world’s largest continuous energy converters (their top ratings are an order of magnitude larger than those of the record-size gas turbines). Large modern turbines rotate at up to 3,600 rpm and work under pressures of 14–34 MPa and temperatures up to just above 600°C. The highest conversion efficiencies of modern fossil-fueled plants are now up to 40–43% (compared to just 4% in 1890) and can be raised further by resorting to combined generation cycles.
FIGURE 2.14. Long-term trends in the performance of U.S. steam turbogenerators show higher steam temperature and pressure, improved average efficiency, and rising unit power.
And there has been also some remarkable continuity in the manufacture of large steam turbogenerators. Both Westinghouse and General Electric remained their major designers and producers throughout the 20th century (Cox 1979). But unlike GE, Westinghouse has been dismembered: in 1998 Siemens bought Westinghouse Power Generation, and the following year BNFL bought Westinghouse Electric Co. Allis-Chalmers Division making the machines was eventually bought by Siemens, and Brown Boveri joined with the Swedish ASEA. But, directly or indirectly, the earliest producers of these machines have remained, a century later, their major suppliers. Steam turbines and three-phase alternators produced by these companies dominate the world’s electricity generation, with some of the world’s largest units installed in nuclear stations.
In the year 2000, about three-quarters of the world’s electricity was generated by steam turbines, and while the future growth of generating capacities in Western countries will be relatively slow, modernizing countries in Asia, particularly China and India, and in Latin America have been recently installing more than 30 GW of new turbine capacity every year (EIA 2002). Unmet electricity needs in these countries are such that high expansion rates would have to be sustained for several decades before the supply were to approach the levels prevailing in affluent countries. Consequently, even should we allow for relatively rapid advances in photovoltaics and wind-driven generation, there is little doubt that Parsons’s machines will remain the leading producers of the world’s electricity at least throughout the first half of the 21st century.
Before leaving the story of steam turbines, just a few paragraphs on their marine applications: superiority of Parsons’s turbines for marine propulsion was demonstrated concurrently with increasingly more powerful stationary installations. Parsons’s 1894 prospectus seeking the necessary capital for the new Marine Steam Turbine Co. claimed, quite correctly, that the new system will revolutionize the present method of using steam as a motive power. Experimental vessel Turbinia was 30 m long, displaced 40 t, and was driven by a single radial flow turbine capable of 715 kW (Osler 1981). But its first trials were disappointing as the propeller slippage was almost 50% and the speeds were far below the expectation.
Cavitation, formation of air bubbles around the propeller that was spinning too fast, was recognized as the key problem, and it was remedied by installing three parallel shafts driving the total of nine propellers. This configuration made the Turbinia the fastest vessel in the world, reaching 34.5 knots, 4 knots above the fastest British destroyers powered by compound steam engines. Public display of Turbinia’s speed at a grand Naval Review at Spithead on June 26, 1897, convinced the Royal Navy to order its first turbine driven destroyer in 1898. Six years later, 26 naval ships, including the massive Dreadnought, had Parsons direct-drive turbines, as did soon four of the world’s largest passenger ships—Mauretania, Lusitania, Olympic, and Titanic, every one a great icon of the Golden Age of trans-Atlantic shipping of the early 20th century.
Diesel engines or gas turbines now power most freight and passenger ships, but more than a century after Turbinia’s triumph, steam turbines now capable of as much as 40 MW can still be found on vessels ranging from the largest aircraft carriers to the most modern tankers carrying liquefied natural gas. Vessels of the Nimitz class, the latest series of the U.S. nuclear carriers, have two pressurized water reactors that power four geared steam turbines. Large numbers of steam turbines also drive centrifugal pumps and compressors, including those employed during the Haber-Bosch synthesis of ammonia (Smil 2001). And, finally, commercial success of steam turbines led to the search for practical gas turbines whose remarkable rise will be detailed in this book’s companion volume.
Transformers
I cannot think of another component of the electric system, indeed, of another device, whose ubiquitous service is so essential for continuous functioning of modern civilization, yet that would be so absent from public consciousness as the transformer. Coltman (1988) is correct when he ascribes this lack of recognition to the combination of transformer’s key outward attributes: it does not move, it is almost completely silent, and it is usually hidden, be it underground, inside buildings, behind screens, or boxed in: plates protruding from large transformer boxes are external radiators used for cooling the device. Yet without these ingenious devices, we would be stuck in the early Edisonian age of electricity generation and distribution where the prevailing limitations on the distance over which electricity can be transmitted would have required the engineers to sprinkle power stations with a rather high density throughout the urban areas, and would have forced smaller places to rely on isolated generating systems.
This prospect clearly worried the early proponents of electricity. William Siemens (1882:70) was concerned that “the extension of a district beyond the quarter of a square mile limit would necessitate an establishment of unwieldy dimensions.” This would mean not only a large increase in the total cost of electric conductors per unit area but also that a “great public inconvenience would arise in consequence of the number and dimensions of the electric conductors, which could no longer be accommodated in narrow channels placed below the kerb stones, but would necessitate the construction of costly subways—veritable cava electrica."
Electricity is easiest to generate and most convenient to use at low voltages, but it is best transmitted over long distances with the least possible losses at high voltages because the power loss in transmission varies as the square of the current transmitt
ed through the wires. Consequently, switching to higher voltages of AC reduced the dimensions of conductors, but such high voltages could not have been produced for the transmission and then reduced for the final use in households and offices without transformers. These devices do what their name implies: they convert one electric current into another, by either reducing or increasing the voltage of the input flow. What is most welcome from an engineering point of view is that transformers do it with virtually no loss of energy, and that the conversion works effectively across an enormous range of voltages (Coltman 1988; Calvert 2001).
Basic equations quantify the advantage of a system with transformers. The rate of transmitted electricity is equal to the product of its current and voltage (W = AV), and because voltage equals current multiplied by resistance (Ohm’s law, V = AΩ), power is the product of A2Ω. Consequently, to transmit the same amount of power with 100 times higher voltage will result in cutting the current by 99% and reducing the resistance losses also by 99%: the combination of low current and high voltage is always the best choice for long-distance transmission. Transformers provide the necessary interface to step up the generated current for transmission and then to step it down for local distribution and actual use. At that point, many gadgets may need further voltage reduction and current conversion; for example, 120 V AC has to be transformed into 16 V DC to run the IBM PC on which this book was written.