TECHNOLOGICAL LEVEL AND PROGRESS
IN SOVIET POWER GENERATION

Aggregate data for major inputs and outputs in the United States and the Soviet Union for the electric utility sector (shown in Table 3-1) indicate high comparative productivity and, by inference, a comparatively high technical level for the USSR in this branch. The Soviet industry has higher utilization of installed capacity and a lower heat rate. Soviet labor productivity is only 29 percent of the U.S. level, but, considering that in coal mining Soviet labor productivity is only 7 percent of U.S. level, this must be considered a comparatively good showing. If we use the 0.5–0.7 range for the ruble/dollar ratio for electric power generating capacity suggested earlier, to put a dollar cost on the resources embodied in the Soviet capital stock, the USSR uses a stock 57–60 percent as large as ours to produce an output half as big. That seems a relatively good achievement and is all the more impressive when it is realized that the USSR has a higher ratio of transmission line to generating capacity than the United States does, and a higher share of its generating capacity in the very capital-intensive hydroelectric branch. Two of these indicators—the heat rate and utilization of capacity—merit more detailed discussion.

TABLE 3-1. Comparative Indicators for U.S. and Soviet Electric Power Utilities, 1975

SOURCES AND NOTES: Soviet data from standard Soviet statistical sources. U.S. data from Edison Electric Institute, Statistical Yearbook, FPC, Statistics of Publicly Owned Utilities in the U.S., Statistical Abstract of the U.S.

^aThese Soviet output figures differ from those reported in standard Soviet sources gross of station use. Correction to the net generation concept used in the United States is based on Nekrasov and Pervukhin, 1977, p. 18.

^bThe Soviet figure on employment refers to promyshlenno-proizvodstvennyi personal’, that is, it excludes employees in ancillary activities such as construction. The larger figure for the United States is the sum of employment in privately owned utilities (Edison Electric Institute) and publicly owned utilities. For the latter, the U.S. Statistical Abstract gives a figure for municipal utilities which I have adjusted upward for federal utilities in proportion to generation. The 503 thousand total includes 101 thousand construction account workers in private utilities, and these have been removed in the smaller figure. On the other hand, the Soviet figure probably contains a large number of personnel serving the by-product heating function characteristic of many Soviet power stations.

^cTo convert from BTU to the Soviet measure of tons of standard fuel, one quad (one quadrillion BTUs) = 36 × 10⁶ tons of standard fuel.

The Heat Rate

Comparisons of the heat rate over time with the U.S. and Western Europe are shown in Figure 3-1. In the early postwar period the USSR reported a heat rate for utility stations 20 percent above the rate for the United States. But the rate has come down much more rapidly in the USSR than in the United States and the USSR now claims an average rate appreciably below the United States. The USSR has also caught up with the heat rate achieved in Western European countries.

FIGURE 3-1. Heat Rate in Thermal Electric Power Generation, Utility Sector, United States, USSR, and Western Europe (gram of standard fuel per KWH)

SOURCES: U.S. data from Edison Institute, Historical Statistics of the Electric Utility Industry, and Statistical Yearbook. Soviet data from standard Soviet statistical handbooks. Western Europe here means the six original nations of the Common Market and is taken from EEC, Energy Statistics Yearbook. Later expansion to nine members raised the group average slightly, but the trend remained the same.

The Soviet heat rate is figured on a slightly different basis than in the United States. The U.S. method uses the “higher heating value” for fuel, including the latent heat of water vapor in the combustion products, while the standard fuel in which the Russians give heat rates is defined in terms of “lower heating value” (FPC, Steam Electric Plant Construction Cost and Annual Production Expenses; and Chernukhin and Flakserman, 1975, p. 290). Adjustment of the Soviet rate for 1975 to the American standard would raise it to 362 instead of 340, but would still leave it below the U.S. rate (based on data in Ravich, 1977, p. 27). But even with this correction, the Soviet heat rate still seems suspect to me. Certain differences give the Russians an advantage in fuel economy. First, there is a fairly heavy emphasis on heat and power combines (see below). Secondly, the share of supercritical units is probably slightly higher in total Soviet capacity than in U.S. capacity because of the difference in growth dynamics.

On the other hand, the low quality of the fuel burned in Soviet plants should raise fuel expenditure appreciably. The average heat content of coal burned in Soviet electric power plants is less than 4 gigacalories per ton (Elektricheskie stantsii, 1974:4, p. 13), compared to a U.S. average over 6 gigacalories per ton. Soviet authorities themselves say that this kind of fuel places a heavy constraint on the efficiency of combustion. One source cites a conversion from Cheliabinsk lignite to gas that raised boiler efficiency from 88 to 94 percent, and one from anthracite to gas that raised efficiency from 91 to 95 percent. (Ravich, 1974, p. 234.) It is difficult for me to believe that Soviet fuel performance for all condensing stations (reported as 365 grams of standard fuel per KWH in 1975— Elektricheskie stantsii, 1977:1, p. 82— or 389 grams in higher heating value) is that close to the U.S. average for all stations, reported as 374 grams/KWH in 1975.

In the USSR there must be a strong temptation to report good fuel performance and therefore a possible bias in the reporting system. According to Pruzner (1969), the heat rate is not part of the official plan of either the system or the plant but is figured as an auxiliary indicator. But this rate has to be important at some level and the Gosplan’s planning manual in its section on planning the power sector states that “besides the general economic indicators, among the most important indicators of effectiveness, the improvement of which must be envisaged in the plan, are the heat rate and the expenditure of fuel per gigacalorie of heat . . .” (Gosplan SSSR, 1974, p. 86). One hint that Soviet power plants may underestimate fuel consumption is found in a study advocating pricing energy coal according to heat value (Ivasenko, 1977). The author cites average heat contents reported separately by mines and by the receiving power stations for 95 lots of coal totalling 104 thousand tons. In 79 cases the power plants reported lower calorific values; in 21 cases they were more than 800 KKal/Kg lower, in another 27 cases 300–800 KKal/Kg lower (page 68). The mine may also have exaggerated heat content, but it seems to me its motive for doing so is less strong than the power plant’s for understating it.

Utilization of Installed Capacity

The Soviet figure for average hours of utilization of installed capacity has consistently been above that for the United States. (It was 5,377 hours for the USSR versus 4,332 hours for the United States in 1960, with the difference gradually increasing in recent years to 5,257 hours and 3,906 hours, respectively, in 1975.) A higher rate of utilization can reflect such aspects of technological progress as system integration, reduction in time lost through breakdowns and repair, and improvements in scheduling and locating capacity additions in relation to demand growth. In this case, however, most of it is caused by differences in load conditions, and by a Soviet willingness to do without reserves at peak demand.

The difference in average annual hours of utilization can be accounted for by two factors. First, the USSR has little reserve capacity. The ratio of peak demand to installed capacity in the power system of the Center region in 1970 was an incredible 0.98, and, though smaller elsewhere, it is still high overall. This is probably a weakness rather than a technological or planning accomplishment, since it often leads to cutting consumers off at peak load periods, and not maintaining frequency and voltage standards. Secondly, the USSR enjoys flatter load curves. The load factors for various Soviet systems, (figured on an annual basis by comparing annual output with peak load times the number of hours in the year) tend to be 0.65–0.69, whereas in the United States they are more likely to be 0.57–0.65. An average for the contiguous United States would be about 0.62 and for the European USSR about 0.68. This factor alone accounts for most of the difference in average hours of utilization. Correction for both these factors would make Soviet utilization of installed capacity inferior to U.S. performance.

The higher performance of Soviet power generation as measured by input productivity is mostly explainable by success in developing reasonably large and efficient generating equipment. Significant economies come from increased size of turbogenerator units and plants and from improving steam conditions. Moreover these achievements reflect numerous dimensions of technological sophistication and input quality such as steel quality, construction techniques, automation of station control, and so on. In the remainder of this chapter, therefore, discussion will focus on several aspects of Soviet experience in creating the basic equipment for fossil fired turbogeneration of electricity. This experience is also useful in revealing distinctive features and specific strengths and weaknesses in the Soviet approach to R and D.

CHARACTERISTICS OF SOVIET
POWER GENERATING EQUIPMENT

The changing structure of fossil-fired generating equipment by size of units, by steam parameters, and by size of station is shown in Tables 3-2, 3-3, and 3-4. Consider first the size of individual units.

TABLE 3-2. Distribution of Steam Turbogenerator Units by Size of Unit, Regional or Utility Stations, End of Year, Percent of Total Steam Turbine Capacity

SOURCES:1950–1960: Vilenskii, 1969, pp. 147–151, except that distribution of 150 MW and above for 1955 and 1960 is based on Pavlenko and Nekrasov, 1972, p. 102. For 1965, 1970, and 1975 plan the source is Pavlenko and Nekrasov, 1972, pp. 108, 110. 1974 is from Elektricheskie stantsii, 1975:1.

^aincludes for 1975 Plan the 250 MW heat and power turbine.

^bincludes a couple of 160 and 170 MW units.

^c24.5 is the total for these three classes.

The Soviet distribution by unit size has generally been somewhat inferior to those of the United States and Western Europe, as the USSR has lagged consistently behind the Western countries in introducing each successive generation of larger generating units. As of 1974 (the last year for which the actual Soviet structure seems to have been disclosed), the comparison with Western Europe (specifically the nine Common Market countries) shows the Russians with still over half their capacity in the less-than-100 MW category, compared to only 29.3 percent for the Western European group, and with only 0.8 percent in the 500 MW-and-up class, compared to 20.1 percent in Western Europe (EEC, Energy Statistics Yearbook). One reason for the relatively high share of equipment with capacities less than 100 MW is the heavy use of heat and power equipment in the USSR, a peculiarity of the Soviet structure that has compensating advantages. It seems impossible to make the same detailed comparison with the United States, but it is possible to estimate the share of equipment in the 500 MW- and-up and the 30oMW-and-up classes by cumulating information on new additions in FPC, Steam Electric Plant Construction Costs and Annual Production Expenses. This shows that at the end of 1974, the United States had about 49 percent of its fossil-fired steam capacity in units with capacities of 300 MW and over (compared to 24.8 percent in the USSR) and 27.5 percent in the 500 MW-and-over class (compared to 0.8 percent in the USSR).

TABLE 3-3. Steam Parameters of Thermal Power Generating Units, End of Year, Condensing and Heat and Power Stations

SOURCES:1958: Energetik, 1966:8, p. 2. Other years: Pavlenko and Nekrasov, 1972, pp. 94–95, 108, 110.

^aincludes some units up to 76 atmospheres, but most are in units of 35 atmospheres or less.

^bincludes 60–120.

TABLE 3-4. Distribution of Installed Capacity by Capacity of Station, End of Year, Percent of Total

*On the reasonable assumption that all stations above 100 MW are utility stations.

SOURCES:1950–1967: Vilenskii, 1969, pp. 130–31, 243, except 1,000 MW and above for 1965, from Elektricheskie stantsii, 1975:4. 1970: Elektricheskie stantsii, 1975:4. 1975: Energetik, 1976:6, inside front cover.

The share of units with higher steam parameters has grown significantly, as shown in Table 3-3. In particular, the share of units using steam at supercritical conditions rose in the decade from 1965 to 1975 from only 4 percent to about a quarter of the total. (I imagine that the planned 1975 distribution was approximately achieved.) This is probably one of the major factors in the good showing on heat rate performance compared to the United States, where the share of supercritical units in total capacity in 1974 was only about 20 percent.*

Parallel with the growth in unit size there has occurred over the same period a dramatic growth in station size, as shown in Table 3-4. (Note that this table refers to all utility stations, including hydro and nuclear.) In 1955 the largest station had a capacity of 510 MW, and in 1960 there was but one station with a capacity in excess of 1,000 MW. In the sixties a new strategy of building large stations was adopted, and in 1975 there were 43 stations with capacities exceeding 1,000 MW. The maximum station capacity at that date was 3,000 MW, but, with introduction of the 500 and 800 MW units, stations as large as 5,000 MW are expected to be built in the future. The growth in the larger stations comes as a result of both the building of new stations and the enlargement of older ones. For example, the Krivoi Rog regional station, by adding 300 MW units. has grown from 1,800 MW at the beginning of 1970, to 2,400 MW by the beginning of 1971, 2,700 MW by the end of 1972, and 3,000 MW at the beginning of 1974.

Distributions fully comparable to Table 3-4 are not available for Western Europe and the United States, but it appears that the Soviet distribution is distinguished both by a large legacy of very small plants, and by a tendency to build very large plants. In 1965, for instance, when the U.S. private utility sector contained no stations as small as 10 MW, the USSR had 17.4 percent of its capacity in that category (FPC, Statistics of Privately Owned Electric Utilities). Publicly owned utilities in the United States probably had smaller plants on the average than privately owned utilities, but their inclusion would probably not affect this particular comparison. At the same time the USSR had 18 percent of its capacity in stations with capacities of 1 GW and over, whereas in the United States only 12 percent of the capacity of utility stations was in units this large. The trend to large stations continued apace in the USSR in the decade following 1965, though the comparative standing in the share of 1 GW and over was about the same in 1975, when the United States had 31 percent of its capacity in such plants compared to 48.1 percent in the USSR. In the distribution for the United States, plants with capacities of 2 GW and up are not even distinguished, but it is possible to identify them fairly easily in the FPC plant data (FPC, Steam Electric Plant Construction Costs and Annual Production Expenses, and Hydroelectric Plant Construction Costs and Annual Production Expenses). They accounted for 10.3 percent in the United States and 27.8 percent in the USSR in 1975. In comparison with Western Europe also, the USSR has a heavy emphasis on large plants—in the nine Common Market countries at the end of 1975, stations with capacities 1 GW and up accounted for 41.1 percent and plants of 2 GW and up for 17.4 percent of all thermal plants (EEC, Energy Statistics Yearbook— this tabulation excluded hydro plants, but their inclusion could not change the picture much).

This background makes obvious the importance of improved equipment as the main technical line for raising the technological level of power generation. I shall accordingly devote the rest of this chapter to examining the development histories of some of the main models of generating equipment. The growth in station size is less obviously a matter of technological sophistication as such, and I will have much less to say about it.

BASE-LOAD BLOCKS FOR
CONDENSING STATIONS

The principal steps in the history of development of equipment for fossil-fired condensing plants can be briefly summarized. In the prewar period the basic units in condensing stations operated at 35 atmospheres, but after the second World War a shift was made to units at 90 atmospheres and 500°C as the mainstay in new additions. Units with 150 and 200 MW capacity with steam at 130 atmospheres and 565°C were first introduced at the end of the fifties. The first K-150–130 was installed in 1958, and the first K-200–130 in 1960 (Neporozhnyi, 1970, p. 203). These were also the first standard units to embody block design—i.e., units in which the boiler, turbine, generator, and transformer are combined in an integral unit. Block design decreases the capital expenditure per KW of capacity by decreasing the cost of components and the volume of building necessary to house the equipment. The 150 and 200 MW units constituted the main additions during the sixties—some 14.1 units were installed during the decade, and a significant number of them are still being installed today.

During the years these units were being developed the producers were experimenting with still higher steam parameters. In 1952, a 150 MW unit at 170 atmospheres and 52o°C was installed in the Cherepovets station. The Leningrad Metal Factory (LMZ) produced a 50 MW unit at 200 atmospheres and 4oo°C to add onto an existing turbine at the Cheliabinsk station, and a unit at 300 atmospheres and 65o°C with reheat, designed and produced by the Khar’kov Turbine Factory (KhTZ), was added onto an existing turbine at the Kashira plant. All used austenitic steel, but it was decided that this was too expensive and that future designs would use only perlitic steel. At the end of the fifties a decision was being made as to what the next step should be to produce a standard unit at supercritical parameters. One opinion was that it would be possible by 1965 to produce a 600 MW unit, at 580° and 240 atmospheres, using perlitic steel (Elektricheskie stantsii, 1958:2, pp. 2–6 and 1959:7, p. 2). But for some reason it was decided that this was too bold a step, and further tests also led the designers to conclude that given the quality of available perlitic steel, the design temperature should be 565°C rather than 58o°C. One source suggests that the ambitious designs were favored as the result of a kind of “engineering prejudice” favoring a lower heat rate as an end in itself, and that the decision for a less ambitious design was a victory for those who in the second half of the fifties helped strengthen economic calculation and sophistication in the power industry (Kulikov, 1964, pp. 122–124). The upshot was a decision for a 300 MW unit at supercritical steam conditions of 240 atmospheres and 565°C for both initial and reheat temperatures. The history of this model is instructive in assessing the effectiveness of R and D in this field, and worth describing in detail. LMZ and KhTZ each produced a prototype (what the Russians call a golovnoi obrazets) in 1960–61 (Vestnik mashinostroeniia, 1967:10, pp. 18–22). These were installed by the end of 1963, at the Cherepovets station, and the Pridniepr station (Pavlenko and Nekrasov, 1972, p. 100). As background, consider that the United States had installed the first 500 MW unit at supercritical temperature and pressure in 1960, and in 1965 the first 1,000 MW unit (Electrical World, March 4, 1968, p. 34; and CIA, 1965, p. 6). This became the basic unit for new condensing stations. Three more of the 300 MW units were installed in the USSR in 1964 and five in 1965. Annual numbers built up to a peak of 15 or 16 in the early seventies and then gradually fell off. Some will still be installed in the 10th Five Year Plan.

The most striking aspect of introducing the 300 MW model is how long it took to “master” these units once they had been installed. Typically only after four or five years would a unit operate anywhere near its design load or its design fuel rate. In 1965, when 12 were in operation, their average availability was only 41.2 percent, and the heat rate was 416 grams/KWH versus the design rate of about 320 grams. Problems were multiplied when the fuel presented any unusual difficulties. The Ermakov plant burning Ekibastuz coal had the following indicators:

(Elektricheskie stantsii, 1975:3)

Thus, it was only in the sixth year of operation that it achieved tolerably satisfactory operation. There were frequent breakdowns among the 300 MW units, 60 percent of which were due to damage to the heating surfaces of the boilers (Energetik, 1975:2, p. 4). But reliability problems arose with all major components, including turbines, condensers, feed pumps, and hot parts. Somewhere during the process it was found that steel quality was inadequate for operation at 565°C, leading to excessive breakdowns, and a decision was made to operate all equipment at 545°C both for initial steam and for reheat. Since Soviet commentators do not like to talk about this, there is not much information on when and how this decision was reached, but it is stated unambiguously to be the case (Leonkov, 1974, p. 74).

One interpretation for slow mastery and poor performance of the 300 MW units is that “serial production of the 300 MW units was begun before adequate experience had been gained by operating the prototype. Some of the first 300 MW machines installed may never operate satisfactorily . . .” (CIA, 1965, p. 6). As time went on, experience with getting these sets to operate did improve—it is said that those installed in the early seventies could be brought somewhere close to their design indicators within a few months rather than within a few years.

Another difficulty with the protoypes and early series units was failure to design and deliver suitable auxiliary equipment along with the main units. In the early stages it was necessary in installing these blocks to use makeshift equipment developed for other purposes for pumping, draft, deaeration, feed water heating, and other tasks (Elektricheskie stantsii, 1968:8, pp. 17–18). This is a traditional problem, as the following description of experience with the earlier 160–200 MW units shows:

The principal boiler and turbine factories do not provide for complete delivery of auxiliary equipment. . . . For example, while new types of boilers with steam productivity of 500–640 t/hr already are being manufactured by factories not even the drawings have been produced for the auxiliary equipment—feed pumps, coal pulverizers, draft equipment, ash removal equipment. . . . Incomplete deliveries take place also with respect to turbines. Currently the first turbines of 150–200 MW capacity are already being manufactured, while the necessary condenser pumps, vaporizers and various types of fittings for them are lacking. [Elektricheskie stantsii, 1958:5]

This is more than just a failure of coordination in deliveries, and reflects rather a failure to coordinate and phase correctly the various parts of the development program. Indeed, the history of the 300 MW units suggests that there is something even more fundamentally wrong. The original conception was never realized in a perfected complex of equipment. The average heat rate for the 300 MW units in 1975 was still 341 grams/KWH rather than the design figure of 320 grams. The lowest rate I have seen quoted for any single unit is about 334 grams/ KWH for mazut-fired stations and 355 grams/KWH for coal-fired ones (Energetik, 1973:3, p. 4). This is not much better than what is achieved in the 200 MW units at 130 atmospheres. Moreover, the supposed advantage of these larger units—savings in capital costs—has apparently not been realized. Whereas it was expected that the 300 MW unit would be cheaper per unit of capacity than the 200 MW units, it is in fact more expensive (Karnaev, 1972, p. 105). This 300 MW unit has worked so badly that the 160 MW units and 200 MW units have continued to be produced in significant numbers in the seventies (Pavlenko and Nekrasov, 1972, p. 102), though it was being said in 1968 that in the next couple of years production of the 160 MW unit would be completely stopped (Elektricheskie stantsii, 1968:8, pp. 17–18).

Experience with the next step upward to 500 MW and 800 MW block units has been roughly similar. The specifications (tekhnicheskoe zadanie, the documents guiding the design of a new unit) for these were approved by 1962, and it was planned that mastering ( vnedrenie, which I take to mean testing after installation) would begin in 1965 (Elektricheskie stantsii, 1962:8, p. 3). There was some slippage and the first 800 MW unit in a two-shaft version did not actually begin operation until 1968 (Pavlenko and Nekrasov, 1972, p. 100). But this model “did not justify itself in operation” and no further such units were produced (Energetik, 1975:1, pp. 5–6). I do not know all that was wrong with this prototype, but one of the main difficulties was with the boiler; it had been designed to work on coal, but this apparently was impossible, and in the end the boiler was totally redesigned to work on gas and residual fuel oil (mazut), then finally on mazut alone (Energetik, 1975:3, p. 7; and Elektricheskie stantsii, 1975:1). Six years after installation, in 1974, it actually began to work decently, producing 5.1 BKWH, but its heat rate was still 353 grams/KWH (Energetik, 1975:3, p. 7). The second 800 MW unit (the one-shaft version) was “accepted for operation” in December 1971 and then worked badly for the next two years. Only sometime in 1973 did it get to full power (Energetik, 1974:1, pp. 5–6). In 1974, it achieved more or less normal operation, producing 4.25 billion KWH (61 percent utilization of capacity) at a heat rate of 366 grams/KWH. I have not seen what is claimed as the design heat rate for these units, but since it was supposed to be somewhat below the design rate for the 300 MW units (320 grams/KWH), 366 grams is far above the design rate.

The next two 800 MW units were commissioned sometime during 1975, at the Uglegorsk and Zaporozh’e stations. They seem to have been mastered more rapidly. In 1976, the coefficient of use of capacity was only 49.2 percent, and the heat rate very high (Energetik, 1976:9, p. 2). Another source, however, says they have carried a load at nominal capacity and have had periods of operation at a heat rate of 322— 328 grams/KWH (Energetik, 1977:4, p. 6).

The first 500 MW prototype was installed at Nazarovo. Originally, vnedrenie was to begin in 1965, though in fact the station was not commissioned until 1968.* A second 500 MW unit was installed at the Troitsk station and commissioned 24 June 1974. According to Pavlenko and Nekrasov (1972, p. 233), it was supposed to have been commissioned in 1973. The Nazarovo and Troitsk models were alternative designs intended to constitute a kind of design competition. The one at Nazarovo apparently never worked at all, and so it is impossible to find much detailed discussion about what the problems were. The other, burning Ekibastuz coal, was made to work after many adjustments and modifications, and an improved second prototype was installed in the same plant. One source says that with this second unit the problem of burning Ekibastuz coal in a 500 MW block has been mastered.

One puzzle about these larger units is that small reductions in heat rate seem to have been bought at very high costs in extra investment. It is said that in the 800 MW turbines the cost per KW is double that for the 300 MW units (Planovoe khoziaistvo, 1976:10, p. 11). Of course, there may be some offsetting gains in other elements of the whole complex.

It is intended that during the 10th Five Year Plan the new 500 and 800 MW units will be produced and installed on a considerable scale in fossil-fired condensing stations as the standard models (Planovoe khoziaistvo, 1976:3, p. 63). They are supposed to be supplemented with a newly developed 1200 MW unit, 300 MW units, and a new 500 MW semipeaking unit, to be discussed later (Nekrasov and Pervukhin, 1977, p. 222). But if things go as they usually have in the past, it is likely that all of the newer models will be delayed so that the 300 MW units will play a larger continuing role than intended.

This brief description seems to support several conclusions. The Soviet Union is still perhaps a decade behind the Western countries in its ability to design and produce large scale conventional steam-based power sets. Each step upward is one in which the experience of the Western countries has already shown what is feasible, but even to follow along behind has strained Soviet R and D capabilities to the limit. There seems to be a poor progression through successive R and D stages, with a tendency to move ahead fairly early to a prototype that is probably inadequately supported by previous experience, design, and material studies. The Soviet system seems to follow a philosophy of getting the new designs into production and then patching them up as production, installation, and operation proceed.

HEAT AND POWER COMBINES

One of the distinctive features of Soviet energy policy is heavy reliance on heat and power combines (teploelektrotsentraly or TETs), which in 1975 generated almost exactly one-third of all Soviet fossil-fired power output. Such combines offer the possibility of large fuel savings through capture of what would otherwise be rejected heat. But the intensity of the Soviet commitment to this idea has the smell of an engineering prejudice, and it deserves careful examination. As an indication of the commitment to this principle, it is universally accepted that future nuclear plants should be designed as heat and power combines. That means they would have to be located in heavily populated areas. Also, since nuclear plants are going to be very large (the Leningrad plant will have a capacity of 2 gigawatts) the heat consuming area and hence the average transport distance for steam would have to be relatively large. It seems possible this is one of those cases in which a principle is established at the strategic level, without taking into account the operating realities at the lower levels, which lead in operation to results much less favorable than envisaged by the planners. The Soviet power industry does obtain significant fuel savings from teplofikatsiia (as this practice is called), but its potential has been eroded by improvements in the equipment available for condensing stations, changing fuel prices, and mistaken decisions in the design of the dual purpose equipment.

The statistical basis for evaluating the fuel savings from TETs is tricky, since there are some heat and power units in what are basically condensing stations, and since in both TETs and condensing stations some heat is supplied to the heating network in the form of steam and hot water produced in separate units. Nevertheless, the situation is approximately as follows: in 1970, the pure TETs had a heat rate of 325 grams of standard fuel per KWH, whereas the average heat rate for condensing thermal stations (KES) was 389 grams (Elektricheskie stantsii, 1977:1, p. 82). Since the TETs produced about 250 BKWH of electric power in 1970, the saving would be about 16 million tons of standard fuel. This is a little misleading, however, since the heat rate cited for KES includes many old stations. The rate for the more modern equipment (i.e., all units using equipment at 130 atmospheres and up) was 368 grams/KWH, and, measured against this standard, the saving would be only about 11 million tons of standard fuel. The saving has grown over time, both because TETs output increases and because the heat rate of the combines has fallen relative to condensing stations. By 1975, the saving would have grown to about 25 MT of standard fuel (ibid.). These savings are large in absolute amount, but look rather small when measured against the total fuel consumption by all stations of 472.4 MT of standard fuel in 1974 and in view of the high share of TETs in all output. Soviet authors often cite much bigger figures for the fuel savings from TETs on the argument that, if this heat were supplied by individual units, the latter would be burdened with very low thermal efficiencies. But those gains are equally available from centralizing heat production alone, and should probably not be counted as gains from teplofikatsiia as a technological principle.

The explanation for this surprisingly low saving is that the TETs work a great deal of the time as condensing stations. In addition most are small and have unfavorable steam parameters. At the end of 1970, there was no cogeneration equipment with supercritical parameters, and units at 130 atmospheres and 565°C accounted for only 15 percent of all TETs’ capacity (Pavlenko and Nekrasov, 1972, p. 94). By 1975, a few supercritical extraction turbines were in operation. Because of the relatively small size of these units, nonfuel costs (such as labor, amortization, and if it were counted, interest) are also relatively high for electric power produced in TETs, and the fuel saving is bought at a considerable cost in other resources. An authoritative work on power station costs says that investment per unit of capacity is 60 percent higher for TETs than for KES (Avrukh, 1977, p. 23). Another relevant consideration is that heat and power turbines get less intensive utilization as power producers than do condensing units—average annual use is 5,300 hours for TETs vs. 5,432 hours for all thermal stations in 1970 (Elektricheskie stantsii, 1974:11, p. 7).

The most important reason the stations have not lived up to their promise for fuel saving is that the heating potential their designers planned for has been underutilized. Most Soviet heat and power turbines are equipped with condensers (about 90 percent of all heat and power turbines by capacity), and when there is no heat load they work as pure condensing turbines. One source says, “As is well known, the heat rate for power in a large number of TETs and in TETs in general is significantly above the projected rate, which is explained basically by the underutilization of their heat capacity” (Elektricheskie stantsii, 1977:8, p. 17). The fraction of operating time during which heat and power turbines worked under a heat load has generally been low, though it has improved over time, as shown in the following tabulation.*

Another indication of dependence on condensing regimes is that average annual hours for utilization of heat capacity are about a thousand hours less than for electrical capacity (Elektricheskie stantsii, 1977:8, p. 17).

Some of the explanation for poor utilization of heating capacity is variation in heat demand, but this is far from the whole story as indicated by the fact that even today peak heat loads are far below heating capacities. In 1975, the aggregate heat capacity of all TETs was 272,000 gigacalories per hour, but the aggregate of peak loads was only about 200,000 gigacalories per hour. The ratio of capacity to load was higher for Minenergo (1.42) than for other stations (125), a difference probably explainable by the problem of coordinating capacity and load when heat is supplied on a utility basis ( Elektricheskie stantsii, 1977:8, p. 16). One especially bad example cited is a TET with a heat capacity of 848 gigacalories per hour for which the maximum load was only 289 gigacalories; another has no heat network at all connected with it, nor is it likely to have one soon (Energetik, 1976:4, P. 37).

A second type of heat and power turbine operates under back pressure from the heat load and must shut down or operate at reduced electrical output when the heat load is absent or reduced below the design figure. These units, too, have had low utilization. Since the heat load curve does not correspond with the power load curve, their electrical capacity is thus often not available to deal with peak needs.

It now appears that a serious mistake made in designing the stations was to provide them with extraction turbines adequate to meet peak heat loads. Even in cases where a station’s heat capacity has been matched to its peak heat loads, it will work most of the time at a heat load well below its maximum capacity, and must send a lot of steam to the condensers. If such stations had been equipped with additional boilers and hot water heaters for meeting peak heat needs, the heat extraction combines could work a much larger fraction of the time at their design heat loads and thus save more in fuel. This weakness has been recognized for a long time (see Elektricheskie stantsii, 1958:2, p. 14), but little has been done to remedy it ( Elektricheskie stantsii, 1977:8, p. 16). It is not clear to me whether these criticisms and the recommendations that such peaking equipment be added are really optimizing or whether they are only an effort to perform better by the wrong criterion—i.e., reduction of heat rates for power.

It is interesting to ask why the heat loads are so much below those projected by the designers. The main explanation offered is that the information regarding industrial and housing heat demands on which station projects were based have been inflated. When the customer enterprises were finished and in operation, their demands were behind schedule and were below the original requests (Energetik, 1976:4, p. 37). Such inflation of requests, as a kind of insurance under a physical rationing economy, is prevalent for current inputs, and it is not surprising that it should also emerge in this longer-horizon setting.

Apparently it also happens that customer enterprises whose needs were included in making demand forecasts have a preference for, and end up with, their own heat plants. The minister of the electric industry, P. S. Neporozhnyi, claims that there is a further anomaly in that the bigger the industrial user, the more likely he is to be able to get permission to equip his plant with its own boiler plant. Minenergo is thus deprived of the opportunity to build larger TETs and use the larger units that would then be more economical (Neporozhnyi, 1972, pp. 161–162). He suggests that permission to build such in-plant boiler installations should require permission from Minenergo, but one can certainly understand why the Ministry of the Chemical Industry, say, would not want to put itself in Minenergo’s hands on so crucial a matter. There seems to be another structural defect in that heating networks are planned, financed, and constructed under separate procedures, so that even when a TET and its customers are ready, there may be no heating network to connect them (Energetik, 1976:2, p. 38). Pipe is typically a “deficit” item in Soviet investment projects, and this is an important reason for delays in completing heating networks (Neporozhnyi, 1972, p. 160).

From following the discussion of policy on TETs over the years, I have the impression that until the late sixties the policy of teplofikatsiia was implemented without a great deal of careful optimizing. M. A. Vilenskii says that the question of the effectiveness of TETs is controversial and such echoes as we hear from these arguments seem quite unsophisticated. P. S. Neporozhnyi, the Minister, once expressed the opinion that, with growth in the efficiency of condensing stations, TETs had lost their attractiveness. Subsequently, however, Neporozhnyi seems to have been converted and is said to have taken the position that electricity should replace hot water for space heating and that process steam for industry should be produced in electric boilers! Both positions sound quite eccentric. In his explication of plans for 1976–1980, Neporozhnyi says a big effort will be made to develop teplofikatsiia, including atomic TETs; but he then goes on to recommend the use of atomic reactors as pure heating plants. It seems an easily grasped idea that, in any situation where heat demand is concentrated enough to justify heating from an atomic heating station, it would be still more economical to combine the heating function with power generation. Neporozhnyi came to his position as Minister via the construction side of the business rather than from a background as an electrical engineer, and the presence of a technological amateur at the top may create conditions for strategic technological biases.

Beginning with the policymaking for the 7th Five Year Plan, the teplofikatsiia decision seems to have been approached with a great deal more subtlety. Zolotar’ev and Shteingauz make the sensible points that the savings from combined production depend on how cheap fuel is, and that the decline in the cost of fuel with the rising share of oil and gas reduces the effectiveness of teplofikatsiia. They link teplofikatsiia to the issue of hydrostations, saying that the cheapening of fuel has made both less attractive and that designers must figure out ways to save on capital costs to justify combines (Zolotar’ev and Shteingauz, 1960, p. 139). The same kind of analysis is well developed in Levental’ and Melent’ev (1961); they lay out the kind of considerations that should be taken into account in deciding on the role of TETs in the system. In particular, they mention as mistakes that must be corrected: delayed construction of the heat network, inadequate variety in the types of turbines and boilers for such stations, poor design of some of the extraction turbines, the fact that many TETs are too small and could be replaced with large ones, and the doctrine that condensing-type equipment was preferable to the back-pressure type.

By the time of the discussions for the 9th and 10 Five Year Plans, as reflected in the Pavlenko-Nekrasov and Nekrasov-Pervukhin books, from which I have cited so often, the policymakers seem to have a very clear idea of all the considerations that should guide the design of the equipment and plants and their interrelationship with systems. One example is the analysis in one of these sources of the choice of the ratio of electric power capacity to heating capacity. The newest unit, the 250 MW extraction turbine operating at supercritical parameters, has a higher ratio of electrical power output to heat output than did earlier, smaller equipment. Using the new rather than the older units to serve a given heat load can thus be thought of as a way of adding to the electrical capacity of the network. But it is a capital-expensive way to do so, compared to conventional condensing stations. Calculations show that this is justified only in regions where fuel costs are high enough that the fuel savings from the higher efficiency of supercritical temperatures justifies the extra capital investment (Nekrasov and Pervukhin, 1977, pp. 93–94).

PEAKING EQUIPMENT

The problem of equipment for meeting peak electrical loads is an instructive example of a failure to do the R and D needed to create the equipment for a recognized need. It is my hypothesis that the explanation in this case is a kind of “criterion bias” in the thinking of electric power R and D authorities or, what comes to approximately the same thing, an excessively narrowly focussed priority system in R and D operations. My hypothesis is that the R and D people have long had an excessive interest in reducing the heat rate, and when they came up against the peaking problem, where economy requires accepting a higher fuel rate in order to save on capital investment, they could not give the task the attention it merited in an overall system-optimizing context. Several Soviet sources say this more or less explicitly—one example is in Teplotekhnika (1971:3).

For a long time Soviet power industry planners have recognized a need for equipment to handle peak and “semipeak” loads (the latter are called cycling or intermediate loads in the United States). The diurnal fluctuation in load on a power station or system looks something like that shown in Figure 3-2a. There is an economic advantage in having in the system some units that operate continuously, fully loaded, and others to operate for perhaps only half the day, handling the increment in load that arises during daytime hours. Still more specialized equipment is needed each day for only a few hours at the time of the very highest demand. The rationale for specialized equipment comes from two factors.

First, the base-load units may not be very flexible in their load carrying capacity—a large block may not be able to shed any significant part of its load and still maintain the various combustion and heat transfer processes its operation involves.* The lower limit for 160–200 MW coal-fired blocks is given in one source as about 65–70 percent of capacity (ENIN, 1968, p. 41). Moreover, most Soviet base-load units require very long periods for starting up and shutting down. It is said that the 300 MW units (which, as will be remembered, constitute about a fourth of Soviet fossil-fired capacity) require eight hours for start-up and approximately the same for shutdown and that they cannot be operated at less than about 80 percent of their capacity when fired with solid fuel. Obviously, it is not possible to start up and shut down such units each day so that they can operate for only 8–10 hours within the 24 hour period.

The second advantage of having special equipment for peaking operations is that its design can be optimized for the smaller number of hours it will operate each year. It is possible to greatly reduce capital cost per KW of capacity by accepting lower thermal efficiency, and, given the relatively small number of hours peaking equipment operates, this penalty may be more than made up by the saving in the opportunity cost of investment tied up in generating plant.

FIGURE 3-2. Measures of Daily Load Variation in Soviet Power Systems

a. European grid, a weekday in December, 1975

1. Hydrostations

2. TET s and old condensing equipment

3. Nuclear and integrated blocks

SOURCE: Nekrasov and Pervukhin, 1977, p. 171.

Though the Soviet Union for a long time had a relatively flat load curve compared to most other countries, and hence a relatively high number of hours of utilization of capacity, the situation has now changed a great deal. One measure is given in Figure 3-2b, showing the steady fall in the ratio of the minimum load to the maximum load for the major power networks of the country. (I believe the Siberian and Far Eastern networks may be an exception and so are omitted.) As another indication, the average hours of utilization for all Minenergo stations was 5,944 hours in 1950, but had dropped to 5,257 in 1975

b. Change in Ratio of minimum to maximum load, 1966–1975

As another gross indication of the need for more flexibility in peaking capacity, Soviet power networks have very high ratios of peak load to capacity. In 1970, the ratio was 0.81 for the European system as a whole and within that an incredible o.98 for the Center system. Analogous figures for the U.S. are more like 0.77–0.80 (Pavlenko and Nekrasov, 1972, p. 198; and FPC, Electric Power Statistics).

On the demand side, the difference between the daily peak and low has increased in recent years because of a shift in the composition of sectoral demand, and a shortened work week has increased variation during the week. On the supply side, the declining share of hydroelectric capacity in total capacity (used in the past to meet daily peaking needs) has hurt. As one offset, the growth of regional systems and interties between them has made it possible to meet peak needs in part by transfers. One disadvantage of that solution, however, has been that in the geography of the Soviet energy sector the direction of power transfers has run counter to interregional fuel flows! (Teplotekhnika, 1971:3, p. 10.) The peaking problem is especially great in the European USSR where water shortages place an additional constraint on the hydro answer, and where there are few big continuous users like aluminum plants to raise the share of base-load demand. Such users have been located in Siberia to take advantage of cheap labor (Styrikovich, in Energetika i transport, 1973:2, p. 6). The problem will become even more serious as the share of nuclear power rises, especially in the European part of the USSR. These nuclear plants are intended for base-load operation. Technically, they can handle some load variation; but one type of reactor has been chosen specifically to accumulate plutonium to stock the first generation of breeder reactors, and variable operation greatly reduces its plutonium-producing potential.

The lack of adequate capacity to meet peak loads leads to cutting off customers at high demand periods, and voltage drops below the established standards. The problem has been discussed for many years, but the task of developing peaking equipment has been put off. In typical Soviet fashion, such development work as was done on peaking equipment was concentrated for a long time on a single solution—gas turbine equipment. Some development work was done on pumped storage, but Soviet power officials were pessimistic about the economic advantage of that approach compared to gas turbines. Only recently has any attention been given to design modifications of large boilers and steam turbine units that would make them easier to start up and shut down. Another possible solution not yet far advanced is combined gas-steam turbine cycles. The history of each of these efforts is revealing about the Soviet innovation system; let us consider them in turn.

Gas Turbines

The development of gas turbines has been slow, with many technical setbacks. Work started in the USSR to create this technology in the late fifties, at about the same time as in the United States. But where as the United States quickly mastered and diffused gas turbine facilities for peaking, the USSR has not been successful in creating this technology.

The program for gas turbines, when it was originally set up, envisaged development of the following series of units (Sivakov, 1968) :

Gas turbines were first used for electric power generation in connection with underground gasification of coal in the Moscow basin, using the 12 MW unit. One of these was installed in the Shatskaia station near Tula and brought to full power in 1958 (Energomashinostroenie Leningrada v 1959—1965 gg, p. 14). Two such units were built, but “this equipment received no further development,” because of the problem of cleaning the gas, which turned out to be too expensive (Energeticheskoe mashinostroenie, 1917–1967, 1967, p. 55). The 12 MW model was also used in a station at Nebit-Dag, where four units were put into operation in 1964–1968 (Teploenergetika, 1970:11, p. 31; Energeticheskoe mashinostroenie, 1917–1967, 1967, p. 56; Energetik, 1975:9, pp. 16–17). Not much has been said about these 12 MW turbines in later discussion, but it is clear they were not designed as peaking units at all; they just represented electric power applications of the units that the Nevskii machine-building plant was producing for compressor stations. Their justification at Nebit-Dag was a plentiful supply of oil-well gas; another advantage in that desert environment was that their requirements for cooling water were small.

The literature on the 7th Five Year Plan (1959–1965) says that two larger experimental units—one 25 MW and one 50 MW—are being developed (Abramov, 1959, pp. 35–36). LMZ was to produce the 25 MW unit of the original plan and the Khar’kov turbogenerator plant, a 50 MW unit (Zhimerin, 1960, p. 181), probably as a modification of the originally assigned 35 MW unit. These were intended for power plants in Kiev and Khar’kov respectively with experimental operation planned to begin in 1962 (Novikov, 1962, p. 231). Both these projects encountered serious troubles. Both units were supposedly installed in 1963 (Teploenergetika, 1970:11, p. 3), but as late as 1968 a special decree of the Council of Ministers regarding the electric power industry complained that both were behind schedule and still not ready for operation. (The decree is given in Resheniia Partii i pravitel’stva po khoziaistvennym voprosam, vol. 6, Moscow, 1968, pp. 643–655.) By 1970 the 25 MW version had operated for 14,000 hours, and the 50 MW unit was in experimental operation (Teploenergetika, 1970:11, p. 3). Apparently the 50 MW unit has not been a success, and nothing more is being said about it, though the Khar’kov plant is still working on gas turbines and has produced a 35 MW unit for use in a steam-gas combined-cycle unit (Pavlenko and Nekrasov, 1972, p. 239).

The 25 MW size was sufficiently successful to encourage further use. LMZ was reported to be producing several 25 MW units to be used in the Iakutsk GRES (Neporozhnyi, 1970). The first of four units was started up in 1970, and a second was installed and tested soon after (Energomashinostroenie, 1970:4, p. 5). But the 25 MW units are probably best considered as an experiment. It is quite possible that they are not expected to play a peaking role but are being used to take advantage of low cost gas for a small station. LMZ is also supposed to be producing a 30 MW unit that can be used either as peaking or base-load equipment, which sounds like a dubious notion, since design needs to be optimized according to quite different criteria for the two purposes.

The next step and the main focus of current effort is the 100 MW unit assigned to LMZ for production and intended for use in the Krasnodar GRES. In the 1968 decree mentioned above, the Council of Ministers directed that this 100 MW unit be finished and started up no later than 1970. A 1969 source reports it as already produced and in the process of being installed (Vestnik mashinostroeniia, 1969: 12, p. 3). But there have clearly been great difficulties with getting it into operation, and in 1975 it was reported as still in the process of being mastered (Energetik, 1975:2, p. 5). The 9th Five Year Plan envisaged that six such units would be commissioned (Baibakov, 1972, p. 101), but so far as I can tell no additional units were in fact commissioned in the 9th Five Year Plan. It is said that LMZ has significantly improved the 100 MW turbine on the basis of the experience with the Krasnodar plant, and is to produce ten such turbines in 1976–1980 (Nekrasov and Pervukhin, 1977, p. 201).

For completeness I should add that there is a miscellany of smaller gas turbine units in use in electric power stations. On 1 January 1968, there were 30 gas turbine units in operation for electric power generation with a capacity of 300 MW (P. S. Neporozhnyi, Elektrifikatsiia SSSR, Moscow, 1970, pp. 2, 5). That means an average of 10 MW each, and most seem to be intended for standby emergency use. There is also a ship-base generating station with a total capacity of 22 MW using gas turbines created by the shipbuilding industry (Ekonomicheskaia Gazeta, 1976:2). Here the rationale is mobility rather than peaking use.

In addition to the slow progress in developing and mastering the experimental gas turbine units, it seems that the models so far produced are not really suitable for the intended purpose. The 100 MW unit was the first gas turbine unit supposedly designed for peaking use (Energeticheskoe mashinostroenie, 1917–1967, Moscow, 1967, p. 56). According to Academician M. A. Styrikovich, however, even this unit is not really suitable for peaking, since it is quite elaborate and expensive in design (to achieve fuel economy) and has a high capital cost per KW of capacity. Moreover, it requires 35 minutes from a cold start to full load, which Styrikovich contrasts with a U.S. 150 MW unit that can do so in half the time (Energetika i transport, 1973:2, p. 7). The Styrikovich claim may be overoptimistic—two of these units installed in a power plant in Hungary require 40–45 minutes to be brought to full load (Energetik, 1976:6, p. 29).

The Hungarians report a heat rate of 500–514 grams of standard fuel per KWH, and the Krasnodar unit is said to have experienced a rate of 500 grams/KWH (Karol’, 1975, p. 30). Since the U.S. rate for gas turbine plants is about 545 grams/KWH (FPC, Gas Turbine Electric Plant Construction Costs and Annual Production Expenses), one wonders whether the designers may have gone too far in trying to save fuel, though it could be that they are only responding to a higher relative cost for fuel. A failure to optimize for peaking purposes would be understandable since LMZ has specialized in producing turbines for compressor stations, which involve continuous use. Also, all these gas turbine units have been combined with standard generators, rather than with special generators optimized for the low number of hours of utilization (Sivakov, 1968, p. 39).

The contrast of this history with U.S. experience is striking. The first U.S. gas turbine plant with a capacity of 300 MW was installed in 1958. The FPC first began reporting on gas turbine units in U.S. utility plants in 1963, when there was a total of 600 MW of installed capacity, and by the end of 1975 total installed capacity was 43,533 MW. And by that date this included single units of 300 MW capacity. If the plan for producing and installing 10 LMZ–100–750 sets in the 10th Five Year Plan is actually achieved, Soviet capacity at the end of 1980 would not exceed about 1,500 MW. In view of the long delays compared to targets for individual programs (five to ten years) and the long lag behind the diffusion of this technology in the United States, my conclusion is that the Soviet gas turbine program has thus far been a failure.

I suggest several factors to explain these failures in the gas turbine program: the whole Minenergo structure has not put a high priority on peaking problems; the plants and designers producing turbines are much more concerned with turbines for compressor stations and other continuous uses than with turbines for electric power generation; there seems to be a specific technological problem with metallurgy and with reliability that is also evident in the gas turbine program for compressor stations.

Pumped Water Storage

An alternative for meeting daily peaks is pumped storage, in which a system’s unused capacity in off hours is used to lift water into an artificial reservoir from which it will flow back to generate power in the peak periods. Investment costs for construction and equipment can be reduced by using a reversible unit as a motor-pump combination in one part of the cycle and as a turbine-generator unit in the other. The research organization Gidroproekt imeni Zhuka began studies on pumped storage stations in 1959, and work was begun in 1963 on the first station, near Kiev, which came into operation in the early seventies (ENIN, 1963, p. 257; and Karol’, 1975, p. 3). One of the main functions of this project was to test the new reversible units for this kind of work. The station has a 200 MW capacity in six units, of which three are reversible, with pumping power of 42.4 MW each (Gidroproekt, 1972, p. 182). How successfully this station has operated is unclear—a recent survey of the Ukrainian power industry mentions it but gives no details (Shvets, 1970, p. 112).

Soviet analyses of the costs of pumped storage for a long time concluded that pumped storage was not competitive as a peaking approach. One source asserts that its capital cost per KW is higher than for other types of peaking units and is “approximately on the same level as base-load condensing stations” (Sivakov, 1972). The approach used is to figure a permissible cost for pumped storage equipment that would not raise costs in the system above that which would result from using standard condensing steam stations instead (i.e., a kind of shadow price); Sivakov says that the projected cost of the pumped storage capacity would be somewhat above this. Nevertheless the need is urgent, and a useful and more recent comprehensive summary of Soviet analyses of pumped storage (Karol’, 1975) suggests a more favorable evaluation of its competitive position. The 10th Five Year Plan guidelines direct Minenergo to accelerate the construction of pumped storage stations (Ekonomicheskaia Gazeta, 1975:51, p. 6). A 1,200 MW station is now being built on the Kunia River near Zagorsk (Energetik, 1974:4). For future pumped storage it is apparently intended to use more powerful equipment—LMZ is to produce the golovnoi 200 MW reversible unit during 1976–1980 for the Zagorsk facility (Nekrasov and Pervukhin, 1977, pp. 200–201).

It is very important to find some solution for this peaking problem, especially in the European USSR where a large number of nuclear stations are being built and where there is little unused hydroelectric potential. One source indicates that pumped storage is expected to be used primarily to complement base-load nuclear plants (Teplotekhnika, 1971:3, p. 5).

“Semipeak” Equipment

In addition to sharp morning and evening peaks in the daily demand pattern there is a daytime/nighttime difference that calls for equipment to be operated for half the day or more, but to shut down for the other half. As the Russians have shifted to 300, 500, and 800 MW block units, this need has been ignored and the large blocks must be kept in operation even at low loads just because they are too difficult to start up and shut down within the daily cycle. The research organizations originally recommended that a simplified medium-size unit with steam at 160 atmospheres and 520–540°C would be the best bet (Teploenergetika, 1971:3, p. 10); but the actual decision has been for a 500 MW unit with steam at 130 atmospheres and at 510°C for both initial and reheat stages, burning either gas or residual fuel oil. It would be started and shut down about 300 times per year and would work about 3,000–3,500 hours (Energetik, 1974:8, p. 4—the detailed rationalization of the specifications is explained in an article in Teploenergetika, 1975:5, pp. 11–16). The new unit is calculated to have a capital cost of 95 rubles/KW and a fuel expenditure of 370–375 grams/KWH. The guidelines for the 10th Five Year Plan reemphasize the importance of such equipment, directing that the electric power industry “accelerate the mastering of highly flexible 500 MW generating blocks” (Ekonomicheskaia Gazeta, 1975:51, p. 6).

LMZ has been working on the prototype since the end of the 9th Five Year Plan, and it is to be produced in the 10th (Nekrasov and Pervukhin, 1977, pp. 198–199). No clues as to current progress have been found, but it is not likely that this new equipment will be ready soon. According to one author: “work on equipment for semipeaking loads is in an embryonic stage. . . . The work on creating equipment adapted to work in the semipeak part of the load curve must be forced, since the need for such capacity is already 5 million KW” (Energetik, 1975:2, p. 5).

Combined Cycles

Finally, it is hoped to combine gas and steam turbines in equipment which will both have peaking potential and offer big fuel economies.

The R and D effort was focussed at first on a cycle in which gas was burned under high pressure in a steam generator, with the combustion products passed through a gas turbine and feed water heating equipment while the steam was used to run a traditional steam turbine. The first such unit of 16,000 KW capacity was commissioned in 1965 at a Leningrad station; it consisted of a 4,000 KW gas turbine and a 12,000 KW steam turbine (Leningradskaia promyshlennost’ za 50 let, 1967, p. 362). It took a long time to make this unit work. As late as the decree of January 1968 on the power industry, the Central Committee and the Council of Ministers complain that this experiment was not on schedule. A couple of other experimental units were also installed in the sixties. On the basis of these experiments it was decided to produce a 200 MW unit that would include a 30–40 MW gas turbine, to be installed in the Nevinnomysskaia power station in 1971. There was some slippage, but tests with natural gas began in 1972. This unit was intended for use with either fuel oil or gas and tests began in 1974 to develop the fuel oil modification (Teploenergetika, 1975:6, pp. 27–28). There were reliability problems, but the equipment operated for a significant amount of time and was considered a success. The tests revealed a fuel rate only about equal to supercritical steam units, but the combined-cycle equipment saves on capital costs and could be used for semipeaking operations.

Apparently the original intention of using gas in these units was not really practical in Soviet circumstances in most areas, because gas could not be assured reliably year round, or would require big investment costs for delivery. Hence, a decision was made to design the units to operate on fuel oil as well. Ironically, about the time the tests on the Nevinnomysskaia prototype were completed (mid-seventies), it became clear that oil was a deficit fuel, and it became a policy goal to avoid use of fuel oil in new power plants. Hence, it is not intended to make extensive use of these units in the 10th Five Year Plan. Some possible application is seen in Tiumen’ oblast’ using surplus oil well gas and in Central Asia on the basis of local gas fields. For these applications, what will apparently be a new round of development work will go ahead on a 250 MW version (Nekrasov and Pervukhin, 1977, pp. 108–109).

The combined-cycle idea can be embodied in an alternative form that can burn mostly coal. In this version, gas or fuel oil is burned to run a gas turbine and the gases are exhausted from the turbine to a traditional boiler burning coal. These gases contain both heat and oxygen from the excess air introduced to maintain the gas temperature at the turbine inlet at no more than 750°. This form is desirable in the new circumstances, since it offers the prospect of getting a unit for semipeaking operations at no fuel penalty compared to the large condensing units. This direction, too, will require starting a development program anew.

With the benefit of hindsight, one is tempted to say that the combined-cycle program was bungled in the sense that the R and D effort was pursued on the basis of a very mistaken forecast about objectives. On the other hand, it may be too much to expect that Soviet R and D planners should have seen in the mid-sixties the shift in the fuel balance that would occur ten years later, thus saving themselves the trouble of creating a technology that could not economically be used. Still, this case clearly shows once again that the long range vision sometimes hypothesized as guiding Soviet fuel policy and energy R and D policy (and offered as a contrast to our own) may also stumble over mistaken forecasts.

CONCLUSIONS

To conclude this chapter, it may be helpful to draw together some intermediate level generalizations and questions about Soviet R and D practices revealed in the cases we have described. They all involve the creation of a very elaborate item of equipment—the 300 MW generating unit, for example, includes a boiler, turbine, and generator as main items plus a great deal of auxiliary equipment—which must in addition be fitted into a larger system. Moreover, major elements will be produced in different plants—the boiler in one organization, the turbine in another, and the generator in yet another. There is thus an extraordinary integration task to be performed, both at the original design stage and at the test and adaptation stage. I have an impression that in approaching this problem, the Soviet system goes rather directly to the creation of what is called a golovnoi obrazets at a fairly early stage and relies on this to settle many issues that might have been settled earlier if more effective design work were done. This golovnoi obrazets might be called a prototype, though it seems to be preceded and accompanied by too little development work even to assure that it will work when installed. It must be supplemented with makeshift auxiliary equipment; significant components of that prototype (such as the boiler or a whole stage of the turbine) may be replaced in an effort to make it work once installed. The history of all these examples shows a long period of naladka— i.e., adjusting, adapting, fixing, so that the item will function. The same pattern will be observed in cases of nuclear power R and D and magnetohydrodynamic power generation. One would think a great part of these delays and this expense could be avoided by more careful testing of components, more careful scheduling of the whole process in advance.

But this golovnoi obrazets is not a prototype in the sense of being just a test or test-bed—it is a full fledged, expensive piece of equipment intended to be used commercially. This is what happens in any country to some extent, but I suspect that U.S. firms would do a great deal more preliminary work and would then expect the new piece of equipment to work essentially as the designers intended.

Another aspect of the Soviet approach is that a new development effort is often limited to a part of the system, relying on standard associated equipment or components to round out the system. Thus, the peaking units supposedly now being developed will use standard generators rather than new designs optimized for peaking. The 250 MW heat and power unit was consciously designed to make it possible to use as auxiliary equipment that already developed in connection with the 300 MW condensing blocks.

The telescoping of successive phases seems to continue through later stages, as well. For example, the Soviets started producing the 300 MW block in series before it had been perfected to the point where it could be delivered, installed, integrated, and made to work with the kind of predictability one would expect for a “developed” technology, embodied in equipment produced in series. Sometimes the problems never get ironed out. I assume that the notion that all the models from the 150 MW units on would operate at a temperature of 565° must involve significant wastes when the original designs continued to be produced, only to be universally operated at lower temperatures.

It is interesting to think about the costs and benefits of this strategy. It is my hypothesis that the USSR could produce a usable, acceptable, model more quickly, have the models they produce in series work better, and so on, if they did more testing and more careful design work first. On the other hand, it may be that plunging ahead will be an effective way to identify problems, focus effort on crucial issues, and accelerate the transition to commercialization. One way to think about it is to say that this rapid push to commercialize is advantageous in the sense that it shows what the problems are and reveals what adaptations need to be made. But the feedback cycle may be too slow to take advantage of this supposed virtue. Thus, the Soviets consistently found that because the coal delivered to power plants had less than the expected heat content, the boiler outputs were often insufficient to fully utilize the capacity of the 300 MW turbogenerator units. So far as I can tell, however, this new intelligence had no effect on the design of the equipment; at least the impact is delayed for a long time. It is interesting to find one Soviet commentator expressing the idea that there are big costs to the Soviet way and that it is wasteful:

The technical base of experimental facilities often does not permit the proper development of designs for machines. . . . Thus, because of the lack of test equipment for developing the golovnoi obrazets of the 300 MW turbine, at the Khar’kov turbine plant and the generator at the Elektrotiazhmash plant, the losses to the national economy exceeded by several times the cost of construction and equipping test facilities. [ANSSSR, 1969, p. 166]

In other words, the Soviet approach is perhaps less a well-thought out R and D philosophy than merely an expedient, faute de mieux.

The power industry people are aware of the problem and are trying to change it as fast as they can. Nekrasov and Pervukhin describe big expansions of test facilities at most of these plants, and the Soviet economist Efimov, noting the adverse effect that a poor experimental base has on Soviet R and D generally, notes Elektrosila and the Khar’-kov turbine plant as exceptions where adequate facilities ensure testing along the way in the test process with desirable results (Planovoe khoziaistvo, 1974:11, p. 18).

But I think the problem has additional dimensions. I am not sure that the developers see experimental operation and debugging as a way to help them improve, perfect, and optimize the technology. They may think they have done their job when they have delivered the golovnoi obrazets and may resent rather than welcome any feedback that tells them what changes must be made. They may not even be much involved in the process. In its 1968 decree on power industry problems, the Central Committee and the Council of Ministers made a special point of saying that responsibility for mastering the golovnye obraztsy of new equipment is not the task of Minenergo alone. Some responsibility must also be assumed by the ministries, plants, and design organizations that produced the equipment (p. 648). I feel that one of the reasons Minenergo has done relatively so well among the various energy sectors in achieving technological progress is that it has the organizational structure and the bureaucratic power to act as its own general contractor, not just for investment, but for the associated development processes as well. This has enabled it to achieve a tolerable level of integration. I suspect that one of the things crucial to its success is that it can get most of the elements of the core technology— the turbines and the generators—out of a relatively small number of very large plants.

____________

*It is quite possible that this is understated, and if some of the highest cost figures associated with obviously more capital-intensive uses such as nuclear are used—say up to 140 R/KW—the ruble/dollar ratio would rise to 0.7.

*There were 123 such units in operation in the United States at the end of of 1974. It is not worth the effort to reconstruct their total capacity from the plant data in FPC, Steam Electric Plant Construction Costs and Annual Production Expenses, but the average unit capacity cannot have been far from 600 MW, which would make their share in all steam units 20.1 percent.

*This is the unit which, according to Hedrick Smith in The Russians, was commissioned in 1968, amid stories in the press about the surge of power into the Siberian grid, when in fact the generator was not yet even on the site! Five years later the unit was still not in operation.

*Based on Elektricheskie stantsii, 1974:11, p. 6; Vilenskii, 1963, pp. 83–84: Levental’ and Melent’ev, 1961, p. 17: Gorshkov (ed.), 1967, p. 51; Elektricheskie stantsii, 1977:1, p. 82.

*And in any case low loads on this equipment raise the unit cost of power because of fixed fuel costs in operating it.