GAS PIPELINES

Technical Level of Soviet Gas Transport

The technological inputs that govern the productivity of a gas pipeline system are numerous and varied, but two elements in the system are of central importance—the pipe itself (with its qualities of diameter, strength, integrity over time and under adverse conditions) and the compressors that maintain pressure in the line. There is a strong implication in the aggregate statistics of the Soviet gas pipeline network that its productivity, and hence presumably its technological level, is low. The transport work performed by the Soviet gas pipeline system and various elements determining the capacity of the system are shown in Tables 7-1 and 7-2. A comparison with similar data for a couple of years for the U.S. system, shown in Table 7-3, suggests that the work performed by the Soviet system is surprisingly small in relation to various capacity-determining dimensions.

TABLE 7-1. Gas Transported on Mingaz Lines

Shipments, Total: 1959–66–Gal’perin, 1968, p. 21: 1967–73–Khaskin, 1975, p. 15; 1974—by working back from net deliveries; 1975–76—Ekonomika gazovoi promyshlennosti, 1977:11, p. 24.

Shipments, Net Deliveries: 1959–66—Gal’perin, 1968, p. 21; 1967–73—Khaskin, 1975, p. 15; 1974—Ekonomika gazovoi promyshlennosti, 1977:5, p. 7; 1975–76—Ekonomika gazovoi promyshlennosti, 1977:11, p. 24.

Shipments, Losses and Own Use: by subtraction, except 1974, which is interpolated. There is clearly something fishy about the 1969 figure.

Average Distance Shipped: 1959—implied by data for work and shipments; 1960–73—Khaskin, 1975, p. 45; 1974—Gazovaia promyshlennost’, 1975:5, p. 8; 1975–76— Ekonomika gazovoi promyshlennosti, 1977:11, p. 24. These figures differ from others often quoted, which are figured by dividing work done by net deliveries rather than total shipments.

Transport Performed: 1959—Gal’perin, 1968, p. 31; 1969–73—Khaskin, 1975, p. 45; 1974—by multiplication from columns 1 and 4; 1975–76—Ekonomika gazovoi promyshlennosti, 1977:11, p. 24.

A pipeline system is made up of many segments of different sizes linked in various ways. To simplify our understanding of how much gas it can transport, however, we can think about it in the following way: in relation to the amount of gas that moves through the system (throughput), think of an average short segment, the most relevant characteristic of which is its cross section. As Table 7–3 shows, in 1970 the average cross section of Soviet lines was 1.79 times that of U.S. lines and, in 1975, 2.35 times as great, whereas Soviet throughput in these years was only 0.42 and 0.62, respectively, as great as U.S. throughput. Since for a given pressure, throughput should vary in proportion to cross section, the implication is that Soviet gas pipelines transport much less gas than they could be expected to.

TABLE 7-2. Capacity Indicators for Mingaz Lines

Length of Lines, With Branches: All data refer to end of year; 1959–66—Gal’perin, 1968, p. 7; 1967–70—TsSU, Transport i sviaz’; since 1970, data on the Mingaz share have not been published, but I adjust the figures given in TsSU, Narodnoe khoziaistvo SSSR for all pipelines by an estimated 3,000 km for local pipelines, based on their length in previous years; 1975, 1976 —Ekonomika gazovoi promyshlennosti, 1977:11, p. 24; 1980 plan—Gazovaia promyshlennost’, 1977:11, p. 5.

Length of Lines, Without Branches: 1959–66—Gal’perin, 1968, p. 7.

Average Diameter, With Branches: 1959–66—Gal’perin, 1968, p. 33; 1970, 1975— Semenova, 1977, p. 10.

Average Diameter, Without Branches: 1960—Gal’perin, 1968, p. 33; 1961, 1965, 1970–73—Khaskin, 1975, p. 59; 1962–64, 1966–69, 1974—Furman, 1978, p. 72; 1975–76 —Ekonomika gazovoi promyshlennosti, 1977:11, p. 26.

Compressor Capacity: (One kw = 1.36 HP.) 1959–60—Gal’perin, 1968, p. 34; 1961–73—Khaskin, 1975, p. 17; 1975–76—Ekonomika gazovoi promyshlennosti, 1977:11, p. 26. So far as I can tell, this is compressor stations on transmission lines, and I assume it is all on Mingaz lines. The figure for 1980 plan is from Ekonomika gazovoi promyshlennosti, 1977:11, p. 26.

TABLE 7-3. Comparison of U.S. and Soviet Gas Pipeline Systems

USSR: From Tables 1 and 2—refers to Mingaz lines and is based on gas delivered rather than shipped, as is true also for U.S. statistics.

U.S.: These data refer to the interstate network, with some small definitional departures. Compressor horsepower and length of system from FPC, Statistics of Interstate Natural Gas Pipeline Companies, 1970. Average diameter for 1970 is given in FPC, National Gas Survey, vol. III, 1973, pp. 126 and 23: and comparisons of the distributions by diameter for 1970 and 1975 in FPC, Statistics of Interstate Natural Gas Pipelines, suggest the structure did not change noticeably by 1975. Data on deliveries in FPC, Statistics of Interstate Natural Gas Pipelines is not quite suitable for our use, since it apparently includes a great deal of double counting. But a special FPC study, National Gas Flow Patterns, 1975 (February 1977, p. 1), gives estimated throughput for 1975 on A & B interstate lines as the amount shown in the table, and we have estimated throughput in 1970 as the same ratio to gas sales by A & B interstate companies. There appears to be no standard statistical series on the average distance gas is transported on U.S. pipelines, but according to what I found in The Economics of Soviet Oil and Gas (p. 153), the average distance was about 500 miles in the sixties and is probably about the same now.

There is a second dimension in the transport job in addition to moving the gas through short segments. The pressure drop due to friction as the gas moves through the line must be counteracted if throughput on an extended section is to be the same as through short sections, and this is the task of compressor stations. And the amount of this work will increase, the greater the average distance the gas must be moved. At one point the low throughput on Soviet lines was explainable by inadequate compressor capacity, but the data in Table 7-3 suggest that this is not the explanation now. As the table shows, by 1975 the Soviet network had more compressor horsepower per volume of line than did the U.S. network and about as much horsepower per cubic meter-kilometer of transport work. Both these measures reflect the distance dimension that compressors are dealing with. They thus allow for the fact that the Soviet gas must be transported on the average farther than the U.S. gas, and we are left with the fact that the considerably larger average diameter of the Soviet pipe ought to make it possible to move much larger amounts of gas through the Soviet system than through the U.S. system.

There are a number of possible explanations for this low productivity. First, the Soviet system probably uses somewhat lower pressures than do U.S. pipelines; most Soviet lines have been designed to operate at maximum pressures of 55 atmospheres (about 800 pounds per square inch). The first Soviet line designed to operate at 75 atmospheres (about 1,100 pounds per square inch) was commissioned only in 1972 (Ekonomika gazovoi promyshlennosti, 1977:5, p. 23). According to FPC, Natural Gas Survey (vol. I, 1973, p. 33), pressures in the range 700–1,000 pounds per square inch are common on U.S. pipelines, and some operate at pressures exceeding 1,000 pounds per square inch. This difference would mean that design capacities on Soviet lines may be somewhat smaller in cross section than U.S. lines. A second explanation is that even in relation to design capacities, the rate of utilization of Soviet gas pipelines is very low. One source gives the average percentage utilization of design capacity on the lines of the Ministry of the Gas Industry as follows:

[Ekonomika gazovoi promyshlennosti, 1977:5, p. 22]

The low utilization of rated capacity is the result of several factors. (1) There may be lengthy service interruptions because of breaks and leaks in the line pipe. (2) The rated capacity of the line may be seriously impaired because lines are badly fouled with condensate, water, and solids as a result of the failure to prepare gas adequately before shipment (Gazovaia promyshlennost’, 1978:10, p. 33). (3) The installed compressor capacity may not be fully available for work. Soviet compressor units apparently experience frequent breakdowns, and spend excessive time out of commission for repair or in reserve status. One author says that in the United States the planned availability of compressors is about 99 percent and actual availability 95–97 percent (Gazovaia promyshlennost’, 1975:9, p. 59). By contrast, the time budget for compressor capacity on Soviet lines in 1973 was as follows (in percent):

Khaskin, 1975, p. 51

(4) There may be inadequate storage at the delivery end of the line to permit reasonably uniform utilization of capacity over the year. (5) It is quite possible that Soviet compressors are not efficient in terms of compression work performed per unit of capacity.

This analysis cannot be translated directly into a conclusion that the quality of Soviet pipe and compressors imposes a low technological level on the Soviet gas pipeline system. Through the latest year shown in the tables (1976), there was very little foreign compressor equipment in operation in the system, but the system has for a long time embodied extensive inputs in the form of imported large-diameter pipe. It does seem, however, that for the system the Soviet gas transport industry has built, it does not get the transport output it should. There is no doubt a large number of contributing factors, but there is a strong presumption that a low technological level both in terms of pipe quality and reliability of compressor stations is important in explaining the low productivity of the pipeline system.

Importance of Imported Pipe for the Gas Transport System.

A very large share of the increment in large-diameter pipelines since 1960 (defining large diameter as 1020mm and above) seems to have been accounted for by imported pipe, as suggested by the data in Table 7-4.

In 1961–1975, about 17 million tons of pipe were laid in oil and gas lines of 1020mm and above. Total imports of large-diameter welded pipe during the same period were 10.084 MT, or 58 percent. Moreover, dependence on imported pipe has increased over the period. In 1971–1975 total pipe investment in large-diameter lines was about 9.648 MT, of which 6.324 MT, or 65.5 percent, was imported. Those data refer to both oil and gas pipelines, but it seems likely that reliance on imported pipe is even greater for gas lines alone, since the quality demands for pipe used in gas lines are higher than for that used in oil lines.

TABLE 7-4. Gas and Oil Pipeline, 1020mm and Above

Figures in brackets are planned figures.

Network Length, Gas: 1960–66, 1969—Gal’perin, 1968, p. 33; 1967–68—Gazovaia promyshlennost’, 1970:4; 1970–75—Orudzhev, 1976, pp. 45–46; 1976—Ekonomika gazovoi promyshlennosti, 1977:11, p. 24; 1980 plan—oil and Gas Journal, May 29, 1978, p. 99.

Network Length, Oil: 1965–67—Rubinov, 1977, p. 28; 1970, 1974—Dubinskii, 1977, p. 5; 1971—Truboprovodnyi transport, vol. 6, 1976, p. 13; 1975—Semenova, 1977, p. 10.

Total Increment in Year: Most Soviet pipe 1020mm and above has a wall thickness of 11–12mm, according to Spivakovskii (1967, pp. 68–69). The weight of such pipe is given by Friman (1976, p. 21) as follows: 1020mm—299 tons/km; 1220mm —358 tons/km; 1420mm—446 tons/km. We also know that for all gas pipelines laid in 1971–75, only 60 percent of which was 1020mm and above, the average metal investment was 334 tons/km. I assume that 350 tons/km cannot be far off for all pipeline 1020mm and above.

Imports of Large-diameter Welded Pipe: Soviet foreign trade handbooks. Unfortunately, beginning in 1976, large-diameter pipe was no longer shown separately. I have not found a source that could settle definitely what is included in the foreign trade commodity class “welded pipe of large diameter," but it seems very likely that imports were almost exclusively pipe of 1020mm and above.

TABLE 7-5. Estimate of Domestic Output of Pipe, 1020mm and Above

All Pipe Output: This is a standard handbook series.

Share 1020mm and Above, Percent: This percentage is arrived at as the product of information on the shares of “large-diameter welded pipe” in all steel pipe output, and of the share of the class “1020mm and above” in all large-diameter welded pipe.

The former has changed little over this whole period—it was 22 percent in 1969, 23.5 percent in 1965, 24 percent in the early seventies (Poliak, 1965, p. 60; Belan, 1962, p. 187; Spivakovskii, 1975 pp. 59, 248, 279). The latter has risen rapidly from 8.78 percent in 1960 to 51.1 percent in 1965, 57 percent in 1970, and 61.2 percent in the early seventies (Spivakovskii, 1967, pp. 68–69; Spivakovskii, 1975, pp. 59, 248, 279).

I have interpolated between the given years and extrapolated to 1975 in each series and multiplied the results to get the series shown in column 2.

A possible source of error in these estimates is inconsistency in the definition of what constitutes “large diameter welded pipe”—in early sources it is described as 478mm and up, in later sources as 529mm and up, and in another case it is described as including 426–1640mm pipe (Tartakovskii, 1978, p. 5).

Share of 1020mm and Above: product of first two columns.

There is a problem in reconciling these data with estimates of domestic production of pipe 1020mm and above, shown in Table 7-5. That table suggests domestic production of about 21 million tons of pipe 1020mm and above in 1961–1975, which together with imports gives a total supply of about 31 MT, whereas we figured the amount used in oil and gas lines as only 17.746 MT. There is a similar discrepancy for 1971–1975, with a total supply from domestic production and imports of 17.099 MT versus 9.648 MT accounted for by newly commissioned lines. Even if we assume an average lag of as much as three years between production and import on the one hand and commissioning of finished line on the other, there is a serious discrepancy between 12.623 MT supplied in 1968–1972, and the 9.648 MT in lines commissioned in 1971–1975. The gap might be the result of underestimating the average weight of pipe, a considerable share of sizes less than 1020mm in the import figure, or an overestimate of domestic production. Overall, however, I conclude that the ratio of imports to pipe in new lines figured above probably overestimates somewhat the Soviet dependence on foreign pipe.

Today the Soviet steel industry can itself produce and is producing 1020, 1220, and 1420mm pipe (Sotsialisticheskaia Industriia, 23 December 1975). One might therefore contend that the USSR imports pipe, not because of technological incompetence, but in order to evade production capacity bottlenecks or to avoid high costs at the margin from expanding domestic output. I would take the position that even these considerations reflect some technological weakness, especially in view of the fact that imports have continued over a couple of decades. In the years when pipe was imported, the Soviet authorities found it physically impossible, or at least excessively costly to organize the production of this pipe domestically, and that is more likely to represent a technical weakness than a temporary production bottleneck. Moreover, this is a technological import in the sense that Soviet pipe is of lower quality than can be obtained abroad, in yield strength, in wall thickness, and in general quality. So far as I can tell, Soviet pipe is inadequate for lines operating at more than 55 atmospheres. One author states rather diffidently that “we have experience with transporting gas at 75 atmospheres pressure in pipe of domestic manufacture (Tartakovskii, 1978, p. 6), but I believe that the newer lines operating at 75 atmospheres have been built essentially with imported pipe. The Russians produce both spiral-welded and straight-welded pipe, the latter from two strips so that there are two separate welds, which is said to add to the problem of inconsistency in quality (Stroitel’stvo truboprovodov, 1977:10, p. 31). So my case is that, if they had not been able to import pipe, they could not have created even the capacity the system presently has, because they could not have produced these sizes and strengths themselves.

Domestic Compressor Development versus Importation.

The USSR has had difficulty developing and producing compressor equipment domestically. A short historical sketch of the development efforts for turbine-powered compressor units shows frequent slippages, long delays in mastering the production of large-size units, and a frequent shift of tactics in response to technological failures.

Soviet pipeline designers have shown a general preference for gasturbine-powered centrifugal compressors as the basic equipment for large transmission lines. The rationale for this preference has been lower cost, autonomy from external electrical power supply, and the example of trends in other countries. As Table 7-6 shows, by 1976 the share of gas turbine prime movers in all compressor power on Soviet gas pipelines was 71 percent. Though this is a much higher share than in the United States, where it is more like 50 percent, it still represents a failure to produce the amount of this equipment considered desirable by Soviet gas industry planners. Electrically powered centrifugal-and piston-type compressors have been substituted to some extent to compensate for the failures and delays encountered in getting the desired gas turbine-powered equipment. As one author says in referring to the mid-sixties, “the high share of electric powered compressors resulted to a large degree because of the slow mastery of production of gas turbines” (Gal’perin, 1968, p. 36). The failure to produce the desired mix continued into the 8th Five Year Plan (1966–1970); it was hoped that gas turbines would account for 77 percent of capacity by the end of 1970, while in fact their share was only 57 percent (Gerasimov, 1969, p. 6).

The original Soviet gas turbine-powered compressor for gas pipelines, first produced in 1958, was based on the GT-700–4 turbine (capacity, 4 MW). That turbine, produced by the Nevskii machine building plant, was originally developed for driving power generators. The turbine was later adapted and upgraded to create more powerful gas pipeline compressor units with capacities of 4.25 MW, 4.4 MW, and 6 MW (Gerasimov, 1969, pp. 44, 139–40). This family of models became and remained the workhorses of the Soviet gas pipeline system right up to the mid-seventies.

Well before the seventies, however, an acute need for larger units was recognized. Between 1965 and the end of 1975, pipe with a diameter of 1020mm and above accounted for 55 percent of all gas pipeline added, and for these large diameter lines units much larger than 6 MW are desirable. Indeed a development program for 10 MW, 16 MW, and 25 MW compressor units had been instituted already in the sixties. But the program has progressed rather slowly, and its goals have been modified at various stages. The 10 MW unit was successfully put into production and has been widely used. During the 9th Five Year Plan (1971–1975) a large number of 10 MW units were installed, and by the end of 1974 they accounted for about 26 percent of all capacity, which implies about 174 units out of the 1,918 shown for that year in Table 7-6 (Truboprovodnyi transport, Vol. 7, p. 45). By the end of 1974, a prototype of the 16 MW model had been produced and accepted and recommended for production by an interdepartmental commission (Gazovaia promyshlennost’, 1975:1, p. 32). The assertion in the source that it would be used in compressor stations in 1975 was probably premature, though we know that three of them had been commissioned on the West Siberian lines by the end of 1976 (Bogopol’skaia, 1979, P. 5). There were also six of the GK.-25I models installed on lines by the end of 1976 (ibid.). The program was amended in 1971 to include a 40 MW unit, but I have seen neither evidence nor claims that significant progress has been made on this model.

In 1974 it was decided to sidestep some of these obstacles encountered in developing and using the original family of models by moving to units based on aircraft engines (Gazovaia promyshlennost’, M, 1975, p. 68). One of the major attractions of the aviation-type units is their light weight, compactness, and block construction, which greatly simplifies and cheapens the construction of compressor stations. The Soviet approach has been to use a turboprop engine hooked directly to the compressor; this has made it possible to use the most powerful Soviet aircraft engine, the NK-12-ST, originally used to power the AN-22, the Bear bomber, and the TU-114. This, by the way, represents a case of reverse technology flow from military to civilian uses, a phenomenon that is often thought to be negligible in the Soviet economy. In its aircraft applications this engine was rated at 8.948 MW and then 11.033 MW in a redesigned version, but for compressor use it has been possible to get only 6.3 MW from it ( Janes’ All the World’s Aircraft, 1977, 1978). This approach has also made it possible to sidestep the development of a new lightweight industrial gas turbine, a necessary element in the Western models, and a step that is difficult for the Soviet Union. This choice has, however, hindered achievement of the goal of easy installation and repair, since careful alignment of the turboprop engine with the compressor is not easy to achieve. Western models use a pure jet only as a combustion chamber so that there is no mechanical-drive connection between the jet engine and the turbine-compressor module. Also, the Soviet machine seems very cumbersome—its weight is 45 tons, whereas a 24.4 MW model produced by United Technologies weighs only 35 tons ( Gazovaia promyshlennost’, 1978:8, p. 8; 1978:11, p. 40).

TABLE 7-6. Compressor Equipment on Soviet Gas Pipelines (end of year, except for size of incremental units)

Number of Stations: 1959–64—Kortunov, 1967, p. 101; 1965–75—Orudzhev, 1976, p. 47; 1976—Ekonomika gazovoi promyshlennosti, 1977: 11. p. 26.

Aggregate Capacity Installed: 1959–76—Same sources as for Number of Stations; 1977–79—Gazovaia promyshlennost’, 1979:5, p. 5.

Turbine-Powered Centrifugal Compressors: 1959–63—Gal’perin, 1968, p. 36; 1964–74—Orudzhev, 1976, p. 47; 1975—Gazovaia promyshlennost’, 1977:4, p. 10, 1976—Furman, 1978, pp. 57, 72. Absolute amounts calculated from percentages given in sources cited.

Average Size of Unit: 1959—Gal’perin, 1968, p. 35; 1960–73—Khaskin, 1975, p. 17; 1974—Planovoe khoziaistvo, 1975:1, p. 22; 1975— Gazovaia promyshlennost’, 1975:12, p. 3; 1976—Furman, 1978, pp. 57, 72.

Implied Number of Units: Calculated.

Size of Incremental Units: Calculated.

The prototype was accepted by Mingaz in 1972, and several such units were tested in the Kupianskaia compressor station on the line from Shebelinka to Ostrogozhsk. On the basis of this experience it was decided to produce 46 units to be used in 12 compressor stations on the Orenburg-Kuibyshev-Center line, and the Nizhniaia-Tura-Center line. Some of these stations were reportedly commissioned in 1974 (Gazovaia promyshlennost’, 1975:5, p. 7). The series model of this aviation-type unit (the GPA-Ts-6.3) has been in production since 1974, and a statement in late 1977 reports that about 150 units had been produced, with a total of 453.6 MW (or about half the capacity that 150 units would aggregate) installed in compressor stations (Gazovaia promyshlennost’, 1977:11, p. 38). By the end of 1977, there were stations with a capacity of 605 MW in operation ( Gazovaia promyshlennost’, 1978:8, p. 10).

The prototype used an engine retired from aviation service, and the series model apparently uses large amounts of parts from retired aircraft engines (Gazovaia promyshlennost’, 1978:8). The early versions of this domestically produced aviation model did not work well, but by 1978 it was being described as a great achievement. Apparently, however, the planners do not see this as a solution to the need for large amounts of compressor capacity, since the USSR has now ordered a significant number of aviation-type units abroad, units which have a capacity double that of the Soviet model (see below).

Another expedient announced as being tried is pipeline compressors powered by marine gas turbines with capacities of 10 MW and 16 MW (Orudzhev, 1976, pp. 65–67). Orudzhev has subsequently said that early experiments with these engines have been successful, and that more powerful models of these machines should be produced for the northern lines (Gazovaia promyshlennost’, 1978:4, p. 10). That sounds as if heavy use of these units is unlikely to come for several years.

Current development objectives include redesigning and repackaging units with capacities already mastered (i.e., the 4–6 MW models) into modular blocks. This program also has fallen far behind schedule. According to an early story seven such models were to be mastered in 1973–1974, but mastery of only three was actually begun (Gazovaia promyshlennost’, 1974:9, p. 12). A prototype (golovnoi obrazets) of the GTN-6 (produced by the Urals plant and the first model in this modular “third generation”) was supposed to complete test trials and be approved by an interdepartmental commission in 1975. The goal was to produce it in 1975 to power stations on the Punga-Vuktyl-Ukhta lines (Gazovaia promyshlennost’, 1975:11; 1975:2; 1975:7), and it is reported that series production did begin in 1975 (Gazovaia promyshlennost’, 1977:11, p. 43). The prototypes (golovnye obraztsy) of the GTN-10, GTN-16, and GTN-25 compressors were retargeted for production by Mintiazhmash by 1976 ( Gazovaia promyshlennost', 1975:6, p. 6); by the end of 1976 a prototype of one of these had been produced—the GTN-16—which it was hoped might be in production by 1979 (Gazovaia promyshlennost’, 1977:11, p. 43).

The major models using an electric motor as prime mover (produced originally at the Nevskii plant but then at Khabarovsk) have capacities of 4 MW or 4.5 MW. The most common piston models have about 1,000 KW capacity (Gerasimov, 1969, pp. 134–137). Development of a 3.4 MW piston model was begun in the sixties, though in 1969 it was still not mastered (Gerasimov, 1969, p. 11). The largest piston unit attempted is one with 5,000 HP (or 3.7 MW) capacity, but it is not clear that it has ever emerged from the R and D process. In any case it is intended for storage reservoirs and head station use rather than for transmission use.

The best indication that none of these efforts to produce large-capacity models has been very successful are the data in Table 7-6 showing the average size of the incremental units in the stock. As the table shows, the average size of all compressor units in service on transmission lines has remained relatively small over the whole period, and even for compressors newly installed the average has moved up only very slowly.

The models that have been produced apparently require complicated and expensive installation work. In typical Soviet fashion they are not shipped as units and must be assembled on the construction site. This is a second source of delay in getting compressor capacity commissioned, and the plans for adding compressor stations have generally been badly underfulfilled. In the 9th Five Year Plan (1971–1975) the mileage targets for new pipeline were basically met, but the targets for compressor stations were met only in part, as shown by the following figures:

[Ekonomika gazovoi promyshlennosti, 1977:4, p. 9]

Domestically produced compressors also seem to have very unsatisfactory service lives. The GPA-Ts-6.3 is being touted as a great success for having achieved in 1978 an average time to breakdown of 1,970 hours (Gazovaia promyshlennost’, 1978:8, p. 19). This in spite of the fact that the guarantee period for this model was originally set at 4,000 hours between capital repairs and, from 1977, was raised to 8,000 hours (Gazovaia promyshlennost’, 1978:8, pp. 11, 26). American firms are advertising in the Soviet journals models, individual units of which have performed 25 to 40 thousand hours of continuous service (Gazovaia promyshlennost’, 1978:11, p. 41 and back cover).

The obstacles to meeting the need for compressors by domestic development seem to be the usual combination of purely technical difficulties (as in metallurgy), organizational and managerial weaknesses, and poor R and D management. Since the experience with aviation engines and with gas turbines for power generation has been similar, it is probably justifiable to conclude that gas turbine technology is simply a tough technology for the USSR to master. In any case it seems abundantly clear that the machinery plants just cannot produce the promised models on the requisite timetable, and it is easy to understand the Soviet decision to turn to import of foreign compressors on a large scale.

The import of gas line compressor equipment is difficult to follow in the official trade statistics, but it may be reconstructed from other sources. In 1973, the USSR made its first deal for importing large gas turbine-powered compressors and, by the end of 1976, had ordered about 3,000 MW of such units (Gas Turbine World, July, 1976). The main deals in this import program were as follows:

(1) In 1973, 37 Solar units of about 2.65 MW capacity were each ordered (and apparently delivered in 1974, to judge from the bulge in the trade statistics of that year), for an implied aggregate capacity of 98 MW. They were installed in the Ukraine, though I do not know on what line (Gazovaia promyshlennost’, 1975:3, p. 4; Oil and Gas Journal, 9 July 1973).

(2) For the Bratstvo line, 63 GE gas turbine units were ordered. These apparently had capacities of 14,600 HP (i.e., 10.9 MW) each (Moscow Narodny Bank, Press Bulletin, 8 August 1976, p. 1; and 15 December 1976, p. 3) and were produced by several companies. This lot of compressors would have a total capacity of 686.7 MW.

(3) For the Soiuz line from Orenburg to eastern Europe, 158 GE gas turbine units were bought, the production of which was divided among several companies (Wall Street Journal, 4 June 1976; and Moscow Narodny Bank, Press Bulletin, 15 December 1976). There will be 22 stations on this line. I gather that those are similar to the units ordered earlier for the Bratstvo line—i.e., with capacity of about 10.9 MW each and implied aggregate capacity of 1,722.2 MW. There must have been some other contracts as well, since it is said that “over 240” of these GE units (compared to the 223 total mentioned here and in the paragraph above) have been bought (Moscow Narodny Bank, Press Bulletin, 24 August 1977).

(4) The USSR ordered 42 units using the Rolls Royce Avon engine in combination with a lightweight industrial turbine for the Chelyabinsk line (Moscow Narodny Bank, Press Bulletin, 15 December 1976). The average capacity of these is apparently about 13.5 MW (Gas Turbine International, July/Aug. 1977, p. 29).

Overall, I read this record as clearly demonstrating an inability to develop the gas turbine compressors, or at least an inability to produce satisfactory ones on time, and in the end a decision to resort to foreign firms for compressor technology.

Benefits from Importing Pipe and Compressors

It would be gratifying to be able to conclude this section by quantifying the net benefits generated for the Soviet economy, by pipe and compressor imports, in somewhat the same way Philip Hanson calculated the benefits from importing fertilizer plants. There is too little information to permit doing this seriously, but we can find some approximate data that give a notion of how beneficial these imports have been. One might ask what the capital cost would have been to provide the same gas transport capacities with purely domestic equipment rather than with imported pipe and compressor units. Planning handbooks give varied figures for construction costs of compressor stations with domestic equipment, as for example the following:

Using GKT-6—191 rubles/KW (Khaskin, 1975, p. 69)

Using GTK-10—180 rubles/KW (Khaskin, 1975, p. 69)

Using GTK-10—146 rubles/KW (Semenova, 1977, p. 77)

The Semenova book suggests that only about half the figure quoted represents the equipment itself. But of course the Western equipment saves much of the expensive construction and installation cost that is needed for the Soviet equipment.

Another calculation shows the following: total investment in gastransport in the 9^th Five Year Plan was 6.260 BR (Ekonomika gazovoi promyshlennosti, 1977:5, p. 5). According to Semenova (1977, pp. 70–71), compressor stations account for about 20 percent of the cost of large pipelines, which implies an investment in compressor stations of about 1.252 BR; if this is divided by the increment in compressor station capacity over the same period, average cost is 259 rubles per KW. The 3,074 MW of foreign compressor equipment listed earlier cost the Soviet Union about $1,199 million (same sources as quoted for the import deals). If we assume that the investment cost using domestic equipment would be somewhere between 150 rubles and 200 rubles per KW, the dollar/ruble ratio for this equipment is between 2.6 and 2.

For pipe, the imports of 10.048 million tons of pipe in 1961–1975 cost 2.325 billion foreign trade rubles. Over that period, the foreign trade prices have been converted to foreign trade rubles at varying exchange rates but mostly at about 1.33 dollars per ruble, which implies a 3.093 billion dollar expenditure of foreign exchange. The average wholesale price of the better qualities of domestic pipe, 1020mm and above, as of 1976 varied from 182 to 258 rubles per ton (Semenova, 1977, p. 84), and if we use a price of 200 rubles per ton, the cost of the pipe if domestically produced would have been 2.010 billion rubles, for a dollar/ruble ratio of just over 1.5. There do not seem to be any good studies of the domestic cost to earn a dollar’s worth of foreign exchange, but Tremi has estimated that in the sixties the ratio of foreign exchange ruble valuations to domestic valuations on all exports was about 1.21, when the official rate of exchange was 1.11 dollars per foreign-exchange ruble, implying that a ruble’s worth of exports earned on the average 1.34 dollars worth of foreign exchange (1972, p. 163). On the basis of these figures it would be difficult to argue a huge resource saving from the imports.

There is, however, a second, independent gain in the form of accelerating the creation of gas transport capacity. It seems likely that, left to its own devices, the Soviet economy would not have been able to provide the transport capacity to get the gas to market without considerable delay. Suppose that there had been no imported equipment and that, as a result, the delivery trajectory realized would have lagged behind the actual by one year. To make it simple, assume that domestic outlays are the same for the two variants. That is, assume that the time profile of domestic resource outlays, using domestically produced pipe and compressors would be no different from that required to follow the import alternative—all that is different is the time profile of when they get this capacity in operation. This seems a minimal estimate of how great a delay would have been imposed by a lack of access to foreign equipment. The difference figured as of the mid-seventies would then be about 20 billion cubic meters of gas lost forever. That would be worth about 750 million dollars at the export price of the mid-seventies and would justify considerable expenditures on foreign equipment.

SUBMERSIBLE ELECTRIC PUMPS

A third notable instance in which the USSR has resorted heavily to imports of technology is that of electric submersible pumps. Such pumps are used primarily to enhance the flow from a well from which there is no longer sufficient reservoir energy to force oil at a satisfactory rate. Compared to other forms of mechanized lift (gas lift or pumps worked with pumping jacks), electric submersibles have higher outputs, do not require elaborate installations for above ground equipment, and should require less maintenance. The Russians have been using electric submersibles to some extent since the mid-fifties and, as old fields have become depleted, have come to resort to them more and more, especially in the Volga-Ural region. In West Siberia they are seen as an important aid in the effort to intensify production from fields, even at an early stage of development.

The first submersibles the Russians used—in the Ishimbai field in 1941–1943—were imported probably under Lend Lease (Amiian, 1962, p. 83). Until the 1970s, however, they subsequently depended on domestically produced pumps. Domestic production was deficient in several respects.

A relatively small selection of models was produced, and those that were produced tended to be of low capacity. Some 42 models are listed in Murav’ev (1971, pp. 190–91), including some up to 86 HP, but apparently only some of those listed were ever actually produced. In the Volga-Ural region on 1 January 1975, 81 percent of the pumps on hand had rated capacities of 100 m³/day or less (Galkin, 1966, pp. 114–115—these rated capacities are stated for work on water, and actual capacities for oil-water emulsions are much less). A. P. Krylov states that, at that time, there were only 24 models available (Neftianoe khoziaistvo, 1973:1, p. 23). Over the years there is a great deal of discussion to the effect that industry is “mastering” the more powerful models, that an experimental lot of some model has been sent for testing, and so on, with subsequent statements indicating that these plans never were carried out or were much delayed. For example, V. D. Shashin, Minister of the Oil Industry, was saying in 1968 that it was essential to get more powerful pumps that could be used in 6-inch casing, but only in 1973 were such pumps tested and series production begun (Neftianoe khoziaistvo, 1968:6, p. 5; and 1974:3, p. 7). The design characteristics of submersible pumps have to be adapted to the specific and highly varied conditions of individual wells, and the limited selection meant poor utilization of the capacity of those the Russians did produce. The pumps finally bought from the United States were rated at 165–215 HP (Oil and Gas Journal, 29 January 1973) and had capacities of 800–1,000 cubic meters per day (Muravlenko, 1977, p. 83); apparently it was only in 1975 or 1976 that domestically produced prototypes of similar capacities had reached the test stage.

Moreover, the Soviet equipment has been low in quality and reliability, so that periods of operation between breakdowns and overall service lives are very short. There is a lot of contradictory evidence on the actual between-repair periods. What purports to be a statement for the USSR as a whole puts the between-repair period for 1975 at 225 days (Neftianoe khoziaistvo, 1974:7, p. 29), though many other accounts suggest much shorter periods. U.S. pumps are said to be designed to operate for 500 days continuously (A. A. Meyerhoff). Quality deficiencies are made worse by the common Soviet weaknesses of limited selection and unavailability of repair parts, so that, in the field, producers are often forced to perform makeshift repairs and content themselves with holding a very large share of the available pumps idle as a reserve. One source complains about the high incidence of pumps which, on being started up again after repair and reinstallation, fail immediately (Neftianoe khoziaistvo, 1975:6, p. 33).

As the Volga-Ural fields became more heavily invaded with water, it became necessary to lift very large volumes of fluid to maintain oil output, and the need for high-capacity pumps became acute. The Soviet oil-field equipment industry has long had programs for developing pumps with capacities of 1,000 cubic meters per day, but it was not until 1975 that domestic R and D had produced some models of this capacity for testing (Neftianoe khoziaistvo, 1975:3, p. 24). Given the urgency of maintaining output from declining fields, the oil industry could not wait, and in the 1970s it was decided to acquire such pumps from U.S. companies. During 1972–1977 the USSR imported about 1,020 of these pumps, and there are additional orders outstanding for approximately 300 more to be delivered in 1978–1981.* The aggregate lifting capacity of the 1,020 pumps already delivered is about 3.5 million barrels per day, for an average of 546 cubic meters per day, though some of these pumps have capacities up to 1,000 cubic meters per day.

It seems likely that the availability of the U.S. pumps has been crucial in achieving output goals during the seventies, though there are many uncertainties in estimating how much of a boost these imported pumps have given to output. Meyerhoff says that, because the Soviet clients did not give the American suppliers sufficient freedom in adapting these pumps to the specific locations and because they have not used the U.S. companies’ help in maintaining them, the pumps failed to achieve anything like their potential productivity. Also, it is reported that many of these pumps have already been withdrawn from service. Nevertheless, we can probably attribute millions of tons of extra output to this import decision, along the lines of the following calculations.

TABLE 7-7. Electric Submersible Pumps in Soviet Oil Production

Number of Wells Equipped with Submersibles: 1959, 1972—Umanskii, 1974, p. 101; 1960—Umanskii, 1962, p. 127; 1964, 1974 —Ekonomika neftianoi promyshlennosti, 1975:7, p. 7; 1967—Neftianik, 1968:9, p. 6; 1969—Neftianoe khoziaistvo, 1972:1, p. 53; 1975—Ekonomika neftianoi promyshlennosti, 1977:11, p. 38.

Output from Those Wells: 1960, 1964, 1972-Campbell, 1976, pp. 27–29, 102; 1967 —Neftianik, 1968:9, p. 6; 1974—Neftianoe khoziaistvo, 1975:6, p. 33.

It seems clear that the Soviet electric submersibles have a much lower productivity than the American pumps. The first two columns in Table 7-7 summarize Soviet statements as to the number of wells equipped with such pumps and the output produced from those wells. Dividing output each year by the number of wells (pumps) working at the end of the year and assuming a specific gravity of 0.8 for oil, gives an approximation of pump productivity in terms of oil shown in the third column (I have converted from annual to daily output by dividing by 365, though obviously these wells worked less than 365 days a year because of breakdowns). The oil/liquid ratio in the output of wells produced with submersibles was 39.8 percent in 1960 (Amiian, 1962, p. 89), making productivity in terms of liquid 78 cubic meters per pump per day in i960. The oil/liquid ratio has surely fallen significantly since then, and if it was as low as 25 percent in 1972, average productivity would be 264 cubic meters of liquid per day. In 1972 there were still no U.S. pumps, so we might take that figure as an indication of the productivity achieved with Soviet pumps.

If we assume that by 1975 as many as 600 of the American pumps were in operation, with an average productivity of 500 cubic meters of liquid per day, and that the other 7,985 pumps (Soviet) had an average productivity of 250 cubic meters per day the liquid raised in 1975 would have been:

Given that oil output from wells equipped with electric submersibles was about 165 MT (206 Mm³ at specific gravity of 0.8), this would make the oil/liquid ratio about 0.25. If we figure the contribution of the American pumps on the basis of the differential average productivity and assume no difference in the oil/liquid ratio, the availability of the American pumps gave the Russians an extra 13.5 MT of oil above what they could have produced without these pumps. That is probably an exaggeration, since the American pumps have probably been most heavily used in wells with the lowest oil/liquid ratios. But, however much we play with these figures, that extra output attributable to being able to use the American technology is very large. Moreover, it seems to have been an excellent bargain for the USSR. The Soviet foreign trade handbooks show an average price on oil exports to hard currency countries in 1975 as 60.4 foreign trade rubles/ton, f.o.b. the Soviet border. During 1975 the prices at which exports were sold to hard currency areas were converted to foreign trade rubles at 1.32 dollars per ruble, making the average price earned on oil exports 79.7 dollars per ton. At that price, 13.5 MT is worth 1.1 billion dollars. Considering that the pumps cost probably no more than 110–120 million dollars, they seem to have a very high benefit cost ratio. (An article in Oil and Gas Journal, 29 January 1973, says a contract for 100 of the pumps was worth 6 million dollars; another source quotes 100–120 million dollars for the total including spare parts.)

DECISION-MAKING CRITERIA
AND CALCULATIONS FOR
TECHNOLOGY IMPORTS

The next task is to explain how these technology import decisions are made. The logical way to think about this is to suppose that planners set up some calculation in which the costs and benefits of import versus domestic development would be assessed in the usual cost-benefit manner. There is an interesting discussion in EKO (1972:4) that illustrates what some of the variants in such a decision might look like. That discussion puts heavy emphasis on the long-range gains, difficult to quantify, from developing domestic technical competence. The big trade-off is cost savings by import versus building scientific and technical capabilities. I believe, however, that this is not a good representation of what really happens; the more carefully I examine these three examples the less it seems that the USSR has any strongly institutionalized way of making a careful choice between domestic R and D and foreign technology.

There is quite a bit of evidence as to how the gas industry planners think about the value of pipe of various technical characteristics. The general approach is to start with some value for gas delivered at the major consumption areas, the production cost of gas in West Siberia, and the cost of various other inputs, using these to impute a value for pipe, which can be thought of as the limit price for pipe, above which building the pipeline and shipping the gas is not economically justified. Within that limit, the cost of alternative designs for the pipeline is compared according to the criterion of minimizing privedennye zatraty (i.e., operating cost plus an opportunity cost charge for the capital committed). This method is well illustrated in Tartakovskii (1978). He uses the method in a quite sophisticated way to evaluate the effectiveness of employing alternative grades of steel that would permit changes in line pressure, wall thickness, diameter, and so on. It is interesting, however, that he never considers the alternative of imported pipe for its potential to improve the “economic effectiveness” of the pipeline. One of the most interesting features in the book, however, is a conclusion that, with gas worth 30.8 rubles per 1000 cubic meters in the consuming area, the limit price for pipe is 913 rubles per ton. Since the value of gas must be still higher as a replacement to free oil for export or as an export itself, the limit price for pipe intended to support exports must be far above that figure. Since we earlier figured the import cost of pipe at $308 per ton, it would seem that, if they are simply unable to produce the needed pipe domestically, they could afford an import price far above what they actually pay.

Another author does take this line of reasoning further to evaluate the desirability of importing pipe. He seems to be saying that the “metal balance’’ is strained (i.e., that the domestic supply of pipe available for gas transport is fixed) and that in this situation it is appropriate to start with the shadow price of gas and work backwards to impute a worth to imported pipe. If pipe can be purchased abroad at a price lower than that, the pipe should be imported (Ushakov, 1972, p. 110). The author seems to envisage using some shadow exchange rate for converting the cost of pipe in foreign exchange into domestic prices. There are many references in these discussions to such a rate under the name of the “import equivalent,” which is usually described as the ratio of domestic prices (costs) of importables to their price in foreign exchange. I don’t know how far the concept of the import equivalent has been officially legitimated and embodied in routine decision making, but it seems fairly common in the economic literature. Unfortunately, none of these discussions give much hint as to what the authors see as the actual import equivalent coefficient.

Another author has essentially the same idea, except that he interprets the problem with domestic provision of pipe as one of poor assortment and low quality. The first constrains optimal design, and the second means that optimal design, under the constraints of domestic pipe characteristics, will require lower pressures and thicker walls, than under those associated with imported pipe, and hence investments of a very large quantity of metal to do a given job. Use of the same tonnage of metal in the form of imported pipe will permit a larger delivery of gas, which can be evaluated at its shadow price (Smirnov, 1975, pp. 62–63).

One of the most elaborate and revealing of these discussions of the import decision, and one that introduces some new wrinkles, is an explanation by two economists from the All-Union Scientific Research Institute for the Construction of Pipelines of how they propose to choose the optimal technology import for compensation deals in gas (Vainshtein and Takhenko, 1975, pp. 142–180). The essence of their approach is to create a considerable number of alternatives, embodying variations in most of the conditions of the project. The variations that interest me here are those involving differences in the mix between inputs of Soviet equipment and foreign equipment. In each variant, imports and interest on the credit are paid off by exports, and the goal is to choose the variant for which the domestic gas availability stream compared to the domestic resource input stream is the most advantageous. In this approach, since the variants involve different degrees of dependence on foreign technology, the choice between technology transfer and domestic R and D falls out of the overall optimization.

The fascinating and ironic thing about each of these is that they settle the issue of technology import versus domestic R and D without ever confronting the choice explicitly. In the Ushakov and Smirnov examples the possibility of domestic R and D to improve pipe quality is simply ignored; in the last example, what is considered is not the cost of improving domestic technology but the cost disadvantage of using existing domestic technology.

I believe that the decision to import pipe has indeed been made in something like this fashion. The reason the pipe case does not look like an R and D issue is that it involves a technology that is firmly embodied in a tradeable product and in which the task of substituting domestic pipe for imports is more one of organization, calculation, and flexible adaptation than it is of drawing on the capacities of the R and D organization of the system. Solving the technological problem domestically is more a battle against the uncertainties of organization than against the uncertainties of the physical world. It does not seem to the policymakers that they can solve the problem of getting large-diameter pipe of the desired quality and assortment, by the time they need it, by expending R and D resources proper, so they do not think explicitly in terms of domestic R and D versus technology transfer.

We might say that they do see the trade-off as one of domestic innovation (by which I mean something broader than R and D proper) versus technology imports, but it becomes very difficult to get any quantitative expression of the cost of domestic innovation in this sense to weigh against the cost of technology imports. They are more likely to go with a basic judgement as to the feasibility of the domestic innovation alternative within a given time horizon.

The interpretation of the compressor case is probably similar. The USSR has had an ambitious and long-standing R and D program for developing this equipment. Compared to pipe, compressors present a more typical R and D problem, in that domestic supply will involve creating new models that require extensive R and D. Some of the best equipped, generally successful, competent Soviet R and D facilities and resources have been put into this area—i.e., the resources of the Nevskii plant, LMZ, the Ural turbomotor plant, and the associated research institutes and design bureaus. Probably because it was thought that this effort could match the achievement of foreign companies, the planners did not at first consider importing the equipment.

Soviet industry has simply been unable to meet the timetable for producing the large compressors needed on the large-diameter lines, and the planners have perforce turned to foreign suppliers. A Soviet analysis of the optimal design and equipment for the Cheliabinsk line, for which the British aircraft-derived units have been bought, originally assumed it would use Soviet compressors—specifically the GTN-16 or GTN-25. But progress on these was apparently too minimal for that intention to be carried out (Gazovaia promyshlennost’, 1977:1, p. 9). Furthermore, it seems that only late in the game were GE units decided on for the Orenburg line. An interview with the head of the prime contractor in 1975 indicated that 25 MW units would be used— presumably the still to be developed GTN-25 (Soviet News, 10 June 1975).*

It is revealing that in the area of gas turbine technology the USSR originally tried to organize technology transfer on a cooperative basis, in areas nearer the R and D frontier rather than as a straight import of technology embodied in equipment. The State Committee on Science and Technology (GKNT) has a scientific technical agreement with General Electric, signed early in 1973, in which gas turbine research is one of the high priority areas (New York Times, 13 January 1973). The Soviet planners thought they had the basic capabilities for development work on this kind of technology well in hand so that they could produce their own equipment. Any borrowing would involve ideas and basic research and would be paid for by a reciprocal exchange in kind rather than by conventional exports.

The submersible pumps, also, seem to fit fairly well into this basic interpretive framework. The USSR has produced these pumps for a long time and in an important sense had command of the technology and some experience with it. They failed, however, to upgrade them in the relevant dimensions fast enough to meet urgent needs. It is interesting that the imported technology is taken as a kind of standard for which the domestic R and D effort should aim. We might think of their decision-making process not as one of deciding whether a given result is best achieved by transfer or domestic effort, but of accepting technology transfer only in extremis and then gauging the amount of resources they put into R and D as whatever is required to match that result domestically and to free them, as time passes, from import dependence.

These examples confirm some well-established generalizations about Soviet R and D. The Soviet R and D establishment has excellent scientific capability, but it does not extend down through the engineering and innovative stages, so that what I assume are good enough basic research, concepts, and design ideas fail to be embodied in reliable, series-produced equipment. But planning goes on in a context of taking at face value the promise of the R and D performers that they will be able to handle the assigned task. Only later is it discovered that failure has occurred at the production end of the “science-production cycle.” At that point, the decision makers have a variety of choices: they can use second-best domestic technology (electrically driven centrifugal compressors, metal-intensive pipe, etc.); accept delays (not build compressor stations on schedule); accept the actual technological level achieved and design around it (install large reserve capacity to make up for lack of reliability in the equipment). But in the oil and gas ministries, at least, and in the era of detente it surely occurs to the higher levels fairly routinely that technology importation is another alternative. And it must happen more often now that, when the planners cost out this alternative, it is sufficiently attractive in relation to any realistic assessment of what their own R and D people will deliver that it could be a permanent solution. I imagine that the question always involves, as a separate issue, whether the required foreign exchange is available. It is not enough to demonstrate its economic advantage—that is only the first step. Surely one reason the oil and gas people have been able to persuade the decision makers to accept technology imports is the fact that these commodities themselves earn so much of the foreign exchange that is required; the easiest case of all must be when the project itself involves direct export earnings. Availability of foreign exchange becomes very important when one moves from an enclave project to something like the compressors imported for the Cheliabinsk lines, where the project is intended to serve domestic needs. In this case, the availability of the needed foreign exchange probably becomes an independent question, even if the project is demonstrated to be cost-effective by the formal calculation.

One can understand the reason for this; it is difficult to relinquish decisions about foreign exchange expenditures to local project makers or to put considerations like stimulus to domestic technical level into value terms. Even when economists suggest how the latter might be done, political decision makers prefer not to relinquish these decisions to economists or to whiz kids with models.

Another factor that undercuts any belief that the domestic R and D versus TT decision is based on the relative costs is that it is unrealistic to imagine that the USSR will ever really abandon an entrenched domestic R and D effort in any significant field. Even when technology is imported to meet a crisis, the domestic R and D program will continue but perhaps be retargeted to copying or bypassing the technology embodied in the imported innovations.

Some implications of this interpretation are worth spelling out. (1) Such technological borrowing can decrease the technological gap and save large amounts of resources used directly in production. It will not, however, save R and D resources for the USSR. (2) The biggest gains are likely to come from borrowing strongly embodied technology. I do not see that the USSR will get much out of the high-level scientific and technical agreements—in this kind of activity they are close to us in relative standing or can get what they need via other channels. In the area of fusion, for example, the USSR may be giving us as much as it is receiving. (3) The real problem in the Soviet system is in the organization and management of innovation. In this downstream end of the R and D spectrum we find feeling out, adapting, optimizing, with increasing knowledge and heavier commitments taken and a smaller range of choices and unknowns to play with. I do not see that they can do much in the way of borrowing in this range of the R and D spectrum, since the structure they work in just puts very stringent limitations on change. For instance, it was said that the transition to series-production on their aviation-type compressor was held up for a whole year because there was no testing stand for it. That is a goof so tied up in systemic cumbersomeness that no amount of outside help or advice could do much to correct it.

____________

*Another source says that the USSR has imported 1,200 of them (Meyerhoff).

*There is a similar example in the coal industry. It was originally intended to equip the Neriungri mine with domestic shovels (EKG-20), draglines (ESh-40/85 and ESh-65/85), and a domestic 180-ton truck (Ugol’, 1976: 2, p. 40). In the end, however, imported trucks and shovels have been used, no doubt because there was no prospect of getting the projected domestic models developed in time.