4

“May the Atom Be a Worker, Not a Soldier!”: A New History of Soviet Reactor Design Choices

When the international scientific community convened in Geneva in the summer of 1955 to attend the first United Nations Conference on the Peaceful Uses of Atomic Energy, the Soviet contributions created quite a stir. Soviet scientists announced what they had achieved the previous year—connecting a small nuclear reactor near Moscow to the public grid and thus creating the world's first nuclear power plant. With a control room that resembled a toyshop, the little graphite-water reactor near the Obninskoe train station would turn out to be the cradle of Soviet nuclear power technology and expertise. Although its 5 MW generating capacity was barely enough to power a locomotive, it established the Soviets as the first to make fission energy available for civilian use. The UN conference confirmed that Stalin's nuclear physicists and engineers were among the international leaders in this cutting-edge area of science and technology.

The Obninsk reactor would serve as an experimental base for all subsequent graphite-water power reactors and as a training site for generations of nuclear specialists. It also served as a public showcase, a demonstration of Soviet scientific and technical prowess that would jump-start yet another Cold War race, this one for peaceful nuclear applications.

The Obninsk plant was the first civilian outcome of the Soviet military atomic project. Though it was much smaller than military reactors, it was based on the same materials, and its history began even before the detonation of the first Soviet atomic bomb. In 1948, nuclear scientists and engineers put forth several design proposals for experimental reactors, and on May 16, 1949, the government issued a decree to start planning an experimental nuclear plant.¹ A year later, on July 29, 1950, Stalin signed another decree ordering construction of installation V-10 (ustanovka V-10) at Laboratory V in Obninsk. Originally, planners intended the plant to have a combined power output of 500 kW from three different types of compact reactors, one from each of three different nuclear research institutes.² In the end, the Obninsk plant had only the graphite-water type. Its selection was the first in what would become a prolonged and complex series of decisions about reactor design in the Soviet nuclear industry.

In this chapter, I introduce the reactor designs that planners considered for power plants in the Soviet Union during the 1950s and 1960s, the reactor types they chose to build experimentally, and the ones they ultimately chose to implement on an industrial scale. By reconstructing which research institutes promoted which design and who lobbied for which type, we can better understand why the Soviet government, despite tight resources, agreed to financially support a protracted, expensive research-and-development process that led to the construction of numerous prototypes. We can also trace why the Soviet Union chose two major reactor designs for their civilian nuclear power program and implemented them in parallel, rather than consolidating scarce resources behind one design.

In 1955, the architects of the Soviet nuclear power program chose an initial design to standardize and implement all across the country. They picked the VVER, whose initials stood for vodo-vodianoi energeticheskii reaktor or water-water power reactor. A light water cooled and moderated, pressurized water reactor, the VVER was similar but not identical in design to Western pressurized water reactors.³

Over a decade later, planners decided to also implement the RBMK, a graphite-moderated, light water cooled, boiling-water reactor. Since 1986, some have referred to the RBMK as the “Chernobyl-type” reactor, and many nuclear physicists and engineers both in the former Soviet Union and abroad have condemned it as inherently unsafe. As we will see, the RBMK was in fact a puzzling choice for standardization and widespread implementation, in part because well before Soviet planners decided to use it, nuclear power programs all over the world had already adopted pressurized water designs.

In the history of Soviet nuclear reactors that follows, I focus on power-generating designs and their immediate predecessors, and I devote the most space and attention to the pool of designs that economic planners and political decision makers actually considered for mass implementation. Because the choice of the RBMK raises numerous questions, I spend more time on that design and less on the VVER. I also analyze the series of contingencies (historical and otherwise) that affected Soviet reactor design choices. I argue that choosing certain reactor designs and abandoning others were not steps along a strictly logical path. Instead, the selections resulted from a long, twisted process that involved technological artifacts, many decision makers, trial and error, innumerable personal interventions, international cooperation and delimitation, rivalry between military and civilian interests, as well as economic considerations. The decision turned in part on the reputation of certain design institutes, on the clout of individual scientists, and on political patronage.

A Multidirectional Approach

The first Soviet nuclear reactors produced plutonium for nuclear weapons. As I have mentioned, Kurchatov ordered tests on graphite-water reactors in 1946 to determine their suitability for power production, and by 1948, scientists were also scrutinizing several other reactor designs for civilian applications.⁴ These reactors’ dual military/civilian features were not always perceived as a negative: it was, after all, the Cold War, and producing isotopes for nuclear weapons was at least as legitimate as supplying electricity and heat for public consumption.

Unlike other nations that settled on their reactor types during the 1950s, Soviet scientists kept experimenting for another decade with a variety of different reactor designs, apparently with the backing of the Soviet government. Deliberately or not, by repeatedly halting and modifying the nuclear power program, the Soviet government effectively supported an expensive research-and-development process that led to several operational power reactor prototypes. Once planners decided to restart the nuclear power program and, more importantly, committed to funding it in the 1960s, scientists presented them with the luxury of options—something most other nations did not have at the time.

Plans for the development of the civilian nuclear industry often closely reflected international trends rather than solid economic indicators. As a consequence, though the Soviet program was repeatedly amended and at times curtailed, it maintained an ambitious trend toward growth and expansion. After Obninsk came online, the next five-year plan (1956–1960) laid out optimistic intentions to create more and more powerful plants, but government officials repeatedly changed those plans and cut back the financial allocations for them. The planning elite continued to hotly contest the projected profitability of nuclear power and even the technical feasibility of large nuclear power plants, as we have seen in chapter one.

Ultimately, Soviet leaders streamlined their nuclear program considerably later than the leaders of many democratic market economies did—a somewhat paradoxical phenomenon in the “country of the plan.” ⁵ Simultaneously pursuing the development of multiple reactor types has sometimes been justified as a “multiple research options approach.” ⁶ At least in the 1950s, Soviet planners considered permanent stabilization of reactor designs not only not a priority, but outright undesirable.

In 1956, Kurchatov reported that the Soviet Union was constructing up to ten different reactor types.⁷ In addition, a massive supply industry, especially heavy industry and fuel manufacturing, grew alongside these reactors. Scientists argued that only by developing in many directions and experimenting with a variety of design options, coolants, and cooling schemes could they produce a rich and diverse scientific and technical knowledge base for the nuclear industry.⁸ In other words, what from today's perspective might look like an irrational, chaotic diversity of approaches made good sense at the time.

Furthermore, a multidirectional strategy had proven reliable during the military nuclear program, so economic planners in fact considered a diversified approach to civilian reactors preferable to choosing one design too early.⁹ Funding a series of competing designs minimized the risk of failure and increased the probability of overcoming technical problems. The idea of “duplication” (dublirovanie) can be traced back to Stalin, who, instead of trusting one person or group, tended to assign responsibilities for the same task to at least two independently operating teams. According to an apocryphal (but probable) story, Stalin had in fact duplicated the entire group set up to develop the atomic bomb.¹⁰

But the power of tradition notwithstanding, supporting researchers’ creativity and the development of multiple reactor designs to the stage of functioning prototypes put enormous strain on the country's financial, material, and labor resources. Such a strategy distributed scarce funding widely and risked insufficient support for the development of a few designs. It was also at least potentially more expensive. These massive resource allocations increasingly required justification.

The pursuit of this multidirectional model of technological development also meant that reactor designers enjoyed a significantly extended period of funded research and development for different reactor options, before having to choose a limited number of designs for mass implementation. At the time, it was still unclear which problems would prove most significant. Reactor designers competed fiercely, and negotiations over design selection reached relative stabilization only by the late 1960s, some ten years after the United States had settled on its commercial reactor models. The specific features of the Soviet state did not let tensions over the choice escalate into public controversy; the state's decisions had to be indisputably rational—or at least had to be perceived as such.¹¹

Obninsk proved the technical feasibility of nuclear power plants, but proving their economic viability remained a challenge. In a 1956 speech at Harwell in the United Kingdom, Kurchatov stressed the significance of civilian reactors’ operational safety, but within the Soviet Union, the ability of nuclear power plants to produce electricity at prices competitive with conventional power plants was of paramount concern.¹²

This balancing act between domestic priorities and the aspiration to impress international audiences also shaped the arguments Soviet reactor designers made for or against a specific reactor type. At different times they invoked similarities with and differences from Western reactor designs and stressed how a given design either fit with international trends and preferences in nuclear engineering or, by contrast, established the Soviet Union's national uniqueness.

From “Laboratory V” to the Institute of Physics and Power Engineering

Like many other nuclear institutes, including Kurchatov's Laboratory No. 2 in Moscow, Laboratory V executed both open contracts for the civilian nuclear industry and secret research and development for the weapons program.¹³ As I detailed in chapter three, Soviet nuclear research centers did not simply develop a reactor design and then turn it over to industry but instead took on the role of scientific director.

Laboratory V was created in December 1945 as part of the national nuclear crash program.¹⁴ Initially, a total of 40 scientists and 200 staff members ran three scientific divisions for theoretical physics, radiochemistry, and reactor materials. Almost immediately after the first successful nuclear weapons tests, Aleksandr Leipunskii, an accomplished nuclear physicist, was appointed scientific director of Laboratory V. Leipunskii's career is extraordinary by any standard: born in the Russian Empire, he studied under Ioffe in Leningrad and under Rutherford in Cambridge; in the 1930s, he convinced famous German physicists to join the scientific staff at the Ukrainian Physical-Technical Institute in Kiev, of which he was the director. But before Leipunskii started to develop Soviet nuclear weapons and later nuclear reactors, Stalin had him locked up for allegedly assisting “enemies of the people” and similar charges.¹⁵ Leipunskii survived, the Party renewed his membership in 1946, and his career soared: he not only directed the Institute of Physics at the Ukrainian Academy of Sciences, but also served as scientific consultant for Laboratory No. 2, as a member of the Scientific-Technical Council of the Council of Ministers’ PGU, and as dean of the Moscow Mechanical Institute (the later MIFI).¹⁶ Leipunskii came to Obninsk straight from the Semipalatinsk test site, and his institute was tasked with preparing a pilot nuclear power plant to test compact reactor types for submarine propulsion.¹⁷

In 1960, after Kurchatov's death, Laboratory V would be renamed the Institute of Physics and Power Engineering (Fiziko-energeticheskii institut, FEI). Over the years and under Leipunskii's capable leadership, it established itself as the leading research-and-development center for breeder reactors, submarine propulsion reactors, and reactors for space.

Obninsk

The three reactor types Obninsk was supposed to test in the mid-1950s included one from the Institute of Physical Problems in Moscow with a graphite moderator and helium coolant (“agregat Sh”) and one with a beryllium moderator (“agregat VT”) from Laboratory V. The third type, the only one actually built, was a graphite-water reactor (“agregat VM”) that Laboratory No. 2 and Dollezhal's design and construction institute NIIKhimmash developed jointly (figure 4.1).¹⁸

Figure 4.1 The world's first nuclear power plant in Obninsk started up in the summer of 1954. Generations of future nuclear specialists first learned their trade at the plant's small 5 MW graphite reactor. The picture shows the control room in 2009, seven years after the reactor had been shut down. The reactor control panel resembles a toyshop, compared to later, industrial-scale plants.

Source: Photograph by Ilya Varlamov, http://zyalt.livejournal.com. Reprinted with permission.

In mid-1951, Kurchatov became scientific director of the Obninsk project and Nikolai Dollezhal and his institute became chief design engineer. The construction bureau OKB Gidropress took responsibility for constructing the steam generator, and the State Specialized Design Institute GSPI (later called NIPIET) for constructing the plant.¹⁹ This long list of participants illustrates that building even a modestly sized reactor involved a small army of leading scientific, industrial, and managerial experts and their institutions. It also shows that the specific collaboration and division of labor first set up for Obninsk represented a fusion of the arrangements characteristic of the weapons program and those typical of the conventional power industries.

Minister Zaveniagin invoked the shortage of enriched uranium to justify why only the graphite-water reactor was actually built.²⁰ The wartime experiences of Soviet physicists may in fact have affected the choice of reactor design for Obninsk, as several researchers have suggested.²¹ But although enriched uranium was certainly a scarce resource, choosing already-proven design features allowed the Soviets to avoid losing time by experimenting with new reactors and to establish a lead in the new race for nuclear power.²²

The idea for the active zone of this graphite-water reactor came from Igor Kurchatov and Savelii Feinberg.²³ While Kurchatov was one of the country's most visible nuclear scientists—he spoke at Party congresses and appeared as an author in the pages of Pravda and Izvestiia—Feinberg was hardly known outside the nuclear physics community.²⁴ In later years, he would spearhead the development of the SM-2 at Melekess, a research reactor that started up in 1961, and would be among the brains behind the RBMK.²⁵

Beloiarsk

Laboratory V nuclear physicists who worked on the Obninsk plant subsequently designed two larger graphite reactors for one of the first commercial-scale nuclear power plants, the Beloiarsk plant near the city of Sverdlovsk (today Ekaterinburg).²⁶ But their designs remained prototypes, never standardized.

The Beloiarsk plant used graphite-water channel type reactors but included innovative design modifications aimed at improving the steam parameters and thus the efficiency and profitability of the plant. Nuclear superheating of steam generated indices comparable to conventional power plants.²⁷ The Beloiarsk reactors were modeled on the Obninsk reactor, but their electrical power was much greater: 100 MW and 200 MW respectively. The first reactor at Beloiarsk operated from 1964 to 1983, and the second from 1967 to 1990. Allegedly, designers soon recognized that the Beloiarsk design was a dead end: “Based on the experience of operating the two . . . units at the Beloiarsk site, it became clear that this direction was not promising in economic terms. At the same time, experience with military reactors had been accumulated, and these reactors looked promising, regarding their construction and their economic indicators.” ²⁸

But the development of the RBMK, a different type of graphite-water reactor based on military designs, began in 1964, the same year the first Beloiarsk reactor started operating.²⁹ Clearly industry officials decided to develop the RBMK design before the Beloiarsk reactors could have proven economically inefficient. The early prototypes at Beloiarsk struggled with a multitude of problems related to technical calculations, industrial capacity, manufacturing know-how and expertise, economic efficiency, safety and reliability of operation, and the availability of qualified cadres, to name but a few. An argument that their economic parameters disqualified them early on, then, reflects the tendency of many histories of technology to flatten and streamline a turbulent, volatile past and to transform the challenges posed by unruly technologies into a linear narrative of logical development.³⁰ Also, the military reactors’ design, which ultimately won out over the Beloiarsk design, had not yet been adjusted to produce electricity: the military reactors’ task was to produce plutonium. It was far from easy—and it took a long time—to learn to operate them in a way that used the surplus heat they produced as a by-product.³¹

Ultimately, the economic argument would, of course, become important. According to Vladimir Goncharov, a leading nuclear physicist at the Institute of Atomic Energy, operating the Beloiarsk reactor types revealed a series of technical flaws—for example, problems caused by the utilization of stainless steel in the active zone. (Later, zirconium would replace the steel.) These ongoing technical difficulties also resulted in unsatisfactory economic indicators. Nevertheless, the operation of the Beloiarsk reactors, perhaps precisely because of their design quirks and the insights that resulted from having to cope with challenging problems on a daily basis, helped nuclear specialists accumulate tremendous experience, which was later used to design reactors with 1000 MW power output.³²

Bilibino, Breeders, and Beyond

Scientists in Obninsk also developed other designs. In 1963, nuclear physicists started working on small reactors for a combined heat and electricity nuclear plant in the remote Chukotka region. The Bilibino Station (BATETs) was the first nuclear plant north of the Arctic Circle.³³ The parallel construction of its four reactors and the plant's quick start-up within three years exemplified the idea that nuclear reactors would eventually be built in a conveyer-belt fashion. Bilibino's identical heterogeneous graphite-water reactors, called EGP-6 for their six loops, featured an innovative scheme for natural coolant circulation through the reactor channels.³⁴ However, like the Beloiarsk reactors, the Bilibino design was never built at any other location.³⁵ Instead, nuclear power planners switched to a different design that the nuclear weapons complex had used extensively, that promised economies of scale more quickly, and that was supported by a different set of scientists and engineers.

Laboratory V also developed fast neutron reactors. This reactor type accomplished fission by relying on fast neutrons as opposed to slow, or moderated, neutrons. The cores of these reactors could be configured to produce more fissile material than was needed to start up—hence their nickname “breeders.” Breeders also needed significantly less enriched uranium than the VVERs or RBMKs to operate, and if properly configured they could produce customizable amounts of start-up fuel. Ultimately, the Soviets saw fast neutron reactors as the pillars of a closed fuel cycle based on a “plutonium economy,” and therefore as the logical next step in the development of nuclear power.³⁶

Theoretical research on breeder reactors began in 1947 and practical research in 1949, under Leipunskii's leadership.³⁷ From modest beginnings in 1955, when the first breeder reactor (BR-1) started up in Obninsk, Leipunskii's team developed multiple experimental prototypes.³⁸ But even before Chernobyl and despite predictions to the contrary, breeder reactors in the Soviet Union never reached the stage of mass diffusion. Explanations range from the high cost of fast reactors to the significant risk of fires to proliferation concerns.³⁹ This picture may have changed somewhat in the post-Chernobyl era, where costly equipment updates and safety retrofits have made traditional thermal reactors more expensive and breeders in turn more competitive. One could speculate that abandoning Laboratory V's designs also had to do with the growing rivalry between the Institute of Atomic Energy and Obninsk over the power to influence the development of a large-scale nuclear power industry.

Kurchatov's Laboratory No. 2

Laboratory No. 2, the Mosow-based research facility created for Kurchatov's group of nuclear weapons developers, was nominally under the Soviet Academy of Sciences. Its later incarnations were called the Laboratory for Measuring Instruments, the Institute of Atomic Energy, and the Kurchatov Institute.⁴⁰

The “Second Ivan” (EI-2)

In close collaboration with the Institute of Atomic Energy, the design institute that had built Soviet production reactors proposed a reactor that would both produce plutonium and generate electricity. For this purpose, the Scientific Research and Design Institute of Energy Technologies (NIKIET) modified a graphite-water reactor for dual use and promoted it as the “Siberian nuclear power plant.” ⁴¹ The EI-2 (“E” stands for power (energeticheskii), “I” for isotope, i.e., plutonium production) was a graphite-moderated, channel type, boiling water reactor, sometimes referred to as “Second Ivan” (the “First Ivan” being a production reactor of similar design at the same site).⁴²

Construction of the EI-2 commenced in 1954, and it started operation in 1958.⁴³ In the summer of 1958, during the Second United Nations Conference on Peaceful Uses of Atomic Energy in Geneva, Soviet media announced that the Siberian nuclear power plant had been launched successfully. Meanwhile, the commission on site was unable to get the reactor to start up. In the presence of high government officials, the reactor specialists agonized over how to accomplish the launch. A representative of the Institute of Atomic Energy apparently suggested replacing part of the nuclear fuel, an option rejected due to the scarcity of enriched uranium. Another member of the start-up commission, relying on his experience with the plant at Obninsk, ventured that the reactor's graphite had absorbed too much humidity during the lengthy start-up process and proposed a solution that involved “drying” the graphite. Powered by an outside source, engineers reversed the process of heat supply: the coolant (water) was heated and used to warm up and eventually “dry” the graphite moderator. After an extended period, the graphite returned to its design parameters, and the reactor start-up succeeded.⁴⁴

Not unlike the plutonium production facility at Hanford, Washington, these first Soviet plutonium reactors used water from the adjacent river to cool the core and afterward released the water back into the river.⁴⁵ They were called priamotochnye, which referred to the continuous flow of coolant. Today, they are recognized as a main source of radioactive contamination at their formerly secret military sites. In the late 1960s, Sredmash completed several more dual-use graphite-water reactors based on the EI-2 design to produce weapons-grade plutonium as well as heat and hot water for the nearby towns.⁴⁶ These dual-use reactors were located at sites long kept secret. The EI-2, and the subsequent ADE reactors, all dual-use designs with graphite moderator and water coolant, were immediate predecessors of the RBMK.⁴⁷

In early 1968, Slavskii and Neporozhnii approached the chairman of the Council of Ministers, Kosygin, with a proposal to use the heat generated as a by-product at the plutonium production reactors for district heating in the nearby city of Tomsk.⁴⁸ Although it was typically Minenergo that administered power and heating plants, in this case Sredmash was appointed the project manager, and the capital investments were allocated to its budget under “electrical power engineering.” Despite the fact that the project's cost more than doubled by 1970, Gosplan eventually increased the annual capital investments to make the project possible.⁴⁹

Starting with the “Second Ivan,” it became clear that dual-use reactors created all kinds of management problems: it proved tricky to run facilities that contained secret parts and processes related to military applications and were at the same time supposed to connect with the civilian power industry. Planners realized that it would be easier to separate nuclear power plants from military production reactors—the former could not only be optimized for electricity generation but also built at sites near large, energy-hungry urban centers. As part of the power industry, nuclear plants could connect to the country's Unified Power System, the gigantic, Union-wide grid of transmission lines that distributed electricity and heat from generating plants to industry and households, and that managed this distribution through an integrated dispatch system. But to connect to this system, a power plant site not only had to meet the siting considerations I mentioned earlier—sufficient water supply, suitable seismic conditions, and adequate space for an occupational health and safety zone—but also demonstrate sufficiently high energy demand, existing transmission lines, and easy access to both a developed industrial base and large numbers of construction workers.⁵⁰

The VVER

Simultaneously with the first Beloiarsk reactor, Soviet scientists from Laboratory No. 2 began to supervise the construction of a second industrial-scale reactor of a different design. In 1956, the government newspaper Izvestiia published an interview with Slavskii, then chairman of the Chief Administration for the Use of Atomic Energy but soon to become minister of Sredmash.⁵¹ Slavskii mentioned that Kurchatov and Aleksandrov were spearheading a project involving a reactor of elegantly simple construction, compact size, and efficient fuel use.⁵² In fact, two years earlier, Kurchatov had instructed the construction bureau OKB Gidropress to draw up plans, and later a technical design, for a pressurized water reactor with thermal power of 760 MW.⁵³ At first, planners discussed several potential sites, including a plant near Moscow that would provide electricity and heat.⁵⁴ Ultimately, they chose a site 500 km south of Moscow, at the river Don and near the city of Voronezh, and workers built a series of five unique reactors, each larger than the one before. But construction did not proceed smoothly. In a letter to Sredmash on June 6, 1957, Kurchatov and Aleksandrov lamented implementation delays and called for a return to previously approved schedules.⁵⁵

Kurchatov took great interest in the construction of the Novo-Voronezh nuclear power plant. He never hesitated to intervene on behalf of the plant and to use his political clout vis-à-vis the government and the Party to request that already-approved plans be fulfilled. He sometimes asked local Party committees to help the factories that were running late manufacture equipment for the nuclear industry. He often succeeded, but despite his efforts, the sixth five-year plan (1956–1960) was not a good one for nuclear power and remained unfulfilled.⁵⁶ Only in 1962, two years after his death, would construction at Novo-Voronezh resume full force and the nuclear industry finally acquire momentum.

In 1964, the first VVER, a 210 MW prototype, went operational at the Novo-Voronezh site; it operated until 1984.⁵⁷ Unit 2 (a 365 MW prototype) came online in 1969 and operated until 1990, and units 3 and 4 (440 MW each) started operation in 1971 and 1972, respectively. The VVER-440 became the first standardized Soviet pressurized water reactor and the reactor model for export. In parallel with the construction of units 3 and 4 at Novo-Voronezh, Soviet specialists started construction of two reactors in Finland. This joint Soviet-Finnish venture prompted the Soviet nuclear industry to develop nuclear power plant safety requirements that met international standards.⁵⁸ The VVERs at Loviisa started operation in 1977 and 1980, respectively.

In 1969, design work for unit 5 at Novo-Voronezh began. This would become the first VVER with a power increase to 1000 MW. Designers achieved this increase by improving the reactor's fuel burn-up while only marginally enlarging the reactor vessel's diameter. Still, several more design changes became necessary: designers intensified the coolant flow rate through the core and increased the length and surface of the fuel elements, which also made the core higher.⁵⁹ The vessel walls had to be thinner—and so be made from different materials—to accommodate the reactor core. In contrast to the early VVER models, which were vulnerable to gradual weakening (embrittlement), the later VVER models would feature a stainless steel cladding on the inside of the reactor pressure vessel. Unit 5 at Novo-Voronezh went critical in 1980, eleven years after construction had begun and two years after construction of these reactors for other sites had already started. Between 1984 and 1993, fourteen reactors of the VVER-1000 type went online.⁶⁰

Like the RBMK, the VVER design had its roots in military applications: developers created it for nuclear submarine propulsion.⁶¹ In contrast to the graphite-water design, which was too large and heavy for submarines, the pressurized water design allowed high uranium burn-ups (and thus efficiency) from a comparatively small core.⁶² The United States also used pressurized water reactors for their nuclear submarines, and just as in the Soviet Union, early use—together with reliability—ultimately gave this reactor type an advantage over other designs that had less or no operational track record.⁶³ The submarine reactor was smaller than the pressurized water reactors later used in power plants, and it had much higher fuel enrichment.⁶⁴ Pressurized water reactors were also used for the first civilian transport applications, starting with the nuclear icebreaker Lenin, which was launched in 1957 and started full operation in 1959.⁶⁵

The VVER design features two circuits or loops. Powerful pumps circulate pressurized water through the reactor core; the water serves simultaneously as moderator and coolant. The heat from the first loop is passed on to the second loop, and the fact that only the water in this second loop reaches the turbines once it has turned to steam diminishes radioactive emissions. VVER reactors also featured horizontal steam generators that provided more lag time under accident conditions than typical Western pressurized water designs, which relied on vertical steam generators.⁶⁶ The pressure vessel, by far the most complicated piece of equipment, was the Achilles heel of the entire VVER line: it was forged of solid shells without longitudinal welds, which meant it required extraordinary factory equipment and welders’ skills.⁶⁷

The thickness of the reactor vessel's walls and its diameter limit the maximum pressure coolant can attain inside the reactor.⁶⁸ The Izhora factory in Leningrad, the only factory in the country capable of guaranteeing the high quality standards required, manufactured the VVER vessels, which had to meet strength requirements and show good resistance to radiation.⁶⁹ The reactor vessel also had to fit the load standards of Soviet trains, so it could be transported from the factory to the nuclear power plant across bridges and through tunnels.⁷⁰ In other words, the Soviet transportation system's capacity for oversized cargo determined the maximum size of the reactor, which at least initially limited its power output.⁷¹

The VVER design was thermally more efficient than its competitors at the time and therefore—at least in theory—cost less to operate.⁷² All VVERs have to be shut down for refueling.⁷³ Typically, plant operators completed that process during the annual shutdown for maintenance and repair, but some saw this aspect of the VVER design as a disadvantage compared with the RBMK's online refueling option.⁷⁴ Finally, designers equipped all VVERs after the first-generation 440 with a steel and concrete containment structure.⁷⁵

Planners also expected to use the pressurized water design in nuclear district heating plants.⁷⁶ The five-year plan for 1956–1960 specified construction of such a combined electricity and heating plant in Khovrino, near Moscow. Planners later abandoned that site in favor of the Novo-Voronezh nuclear power plant, but in 1981 construction began on two nuclear district heating plants, one near Gorky (now Nizhnii Novgorod), the other near Voronezh. Other projected sites included Leningrad, Odessa, and the Kola Peninsula region. To be profitable, nuclear heating plants had to be located near densely populated areas, and for that reason, designers incorporated additional safety features in these reactors.⁷⁷

Public opposition following the Chernobyl disaster prevented both the Gorky and Voronezh plants from going operational.⁷⁸ Although Sidorenko, an early advocate of nuclear district heating plants, believed that the industry could convince the public of the plants’ safety and foresaw that the first such plant would operate by the year 2000, neither prediction has come true.⁷⁹

The RBMK

In 1965, Sredmash's first deputy minister, Aleksandr Churin, and academician Aleksandrov convened a meeting at the Bolshevik factory in Leningrad. The topic of the discussion was a proposal the Institute of Atomic Energy put forward to develop a 1000 MW graphite-water reactor that would help meet the plan target of electricity generation at nuclear power plants. This venerable steel factory, whose tradition reaches back to Czar Alexander II, had produced heavy machinery for Soviet industry, including tractors and tanks. Bolshevik's construction bureau was the top candidate for the task of chief design engineer for the B-190 reactor (as it was then referred to). But when Sredmash's Scientific-Technical Council received Bolshevik's first design draft a year later, the council found it technically unsound.⁸⁰

Time was of the essence. The government had issued a decree on September 29, 1966, ordering that the first two units at the Leningrad and Kursk nuclear power plants use these new reactors. After the disappointing evaluations received for the Bolshevik design, planners resorted to familiar players: Savelii Feinberg and Nikolai Dollezhal set out to revise Bolshevik's initial design proposal. Dollezhal's home institute, code-named Scientific Research Institute No. 8 (NII-8), already possessed significant experience with military and dual-use reactors. NII-8 engineers had designed and built the country's production reactors and the Second Ivan. But now they had to drastically improve the new reactor's physical and thermal efficiency so it would work exclusively for generating electrical power.⁸¹ Dollezhal's institute was appointed chief design engineer and the Kurchatov Institute of Atomic Energy was designated the scientific director.⁸² Dollezhal made the design and construction of what his institute called the RBMK-1000s their first priority.⁸³ NII-8 presented a new design, which included significant modifications, to Sredmash's first deputy minister Churin in February 1967. In June 1967, Churin approved these suggestions and the RBMK design was commissioned in 1968.⁸⁴ In other words, the government authorized the construction of four massive, 1000 MW nuclear power reactors before the reactor design was developed, let alone approved. This course of action makes sense only in the context of a planned economy and in light of an established tradition of duplicating critical research-and-development phases.⁸⁵

At this point, we need to review a few basic specifications of the RBMK, because they are integral to the rest of this story. The RBMK-1000 consists of a large graphite block structure (about seven meters high, and about twelve meters in diameter) that serves as the moderator for slowing down neutrons. The fuel, slightly enriched uranium dioxide, is encased in pellets and is assembled in the form of a rod.⁸⁶ Eighteen such fuel rods, each with a diameter of 13.6 mm, are cylindrically arranged to form 3.5-meter-long fuel assemblies. These fuel assemblies are suspended into the 1,661 vertical tubes mounted throughout the reactor core, and the channel head is sealed.⁸⁷ Of those tubes, 211 are reserved for control rods that can be moved up and down to regulate the reactor's reactivity. Each fuel channel, a fuel assembly inserted into a tube, is individually cooled by pressurized water that enters at the bottom of the tube and passes the fuel from below; the water boils by the time it reaches the upper part of the tube. A nitrogen-helium gas mixture that slowly circulates in the space between the graphite and the channels is used to control the integrity of the channels.⁸⁸ The coolant has a temperature of 290°C when it reaches the steam drum, where steam is separated from water and passes on to the turbine.⁸⁹ There is only one circuit, which increases the reactor's efficiency but allows slightly radioactive steam to reach the turbines.

The RBMK can be refueled online, using a remote-controlled refueling crane; consequently the reactor is almost constantly available, and operators control the level of fuel burn-up, and, correspondingly, plutonium production.⁹⁰ In contrast to the Second Ivan, the RBMK's fuel elements stayed in the core longer and were exchanged less frequently.⁹¹ At least initially, when plutonium was still scarce, the possibility of customizing the production of weapons-grade plutonium was seen as an attractive asset of this reactor type.⁹² The RBMK's designers were well aware of the potential for plutonium production—one indication being that the RBMK was never exported beyond the borders of the Soviet Union.⁹³

The RBMK can also be assembled largely on site, since many parts do not require sophisticated factory manufacturing. Once the designers had accumulated experience with the RBMK, they managed to expand the reactor's output without significantly modifying the size of the active zone: the two RBMKs built in Lithuania had a 1500 MW output with the same reactor core size as the standard 1000 MW RBMKs. The RBMK design was considered particularly safe because of its modular structure: accidents were likely to affect only individual channels, never all of them.⁹⁴ But the RBMK was also massive, and the channel design involved complicated internal plumbing.⁹⁵

Its disadvantages (which nuclear specialists at the time considered either irrelevant or controllable) include the relatively slow emergency shutdown systems, the lack of a containment structure (which would have doubled the cost and presented an engineering challenge due to the enormous size of this reactor), and most importantly, a positive void coefficient, a feature that was involved in the Chernobyl accident. “Void coefficient” refers to the interaction of coolant and moderator. A positive void coefficient means that when cooling water is lost in the RBMK (water converts to steam, which reduces its cooling capacities), the speed of the chain reaction in the reactor core increases instead of decreasing, because the graphite moderator still facilitates nuclear fission. (By contrast, the void coefficient in a VVER is negative: when water is lost, reactivity decreases, because water serves as both the coolant and the moderator.)⁹⁶

The first RBMK-1000 was built at Sosnovyi Bor, a dedicated Sredmash site on the Bay of Finland near the city of Leningrad (figure 4.2). On September 10, 1973, at 10:35 p.m., the physical launch of the reactor started with the insertion of the first fuel assembly.⁹⁷ Savelii Feinberg headed the start-up commission, which consisted of over sixty representatives from the Institute of Atomic Energy.⁹⁸ On September 12, at 6:35 p.m., the reactor reached criticality. After the festivities came the power start-up—that is, the connection of the nuclear reactor to the electrical part of the plant.⁹⁹ On November 15, 1973, the reactor was brought to a level of 150 MW, and the official start-up took place on the evening of December 21, 1973, the day before the “Day of the Power Engineer.” ¹⁰⁰ Both Aleksandrov and Dollezhal, representatives of the scientific director and the chief design engineer, respectively, were present for that event. After the reactor had been working at 160 MW power for three days, the commission approved the reactor for experimental industrial operation.¹⁰¹ Engineers had to compensate for less than optimal design decisions, especially with the fuel assemblies that kept wedging in their channels.¹⁰² Several of these fuel assemblies had to be sent back to Moscow for repeated testing. Finally, on November 1, 1974, the reactor reached its designated power level.¹⁰³

Figure 4.2 Construction and assembly of the Leningrad nuclear power plant, where the country's first RBMKs would start operating in the early 1970s. The first reactor started up seven years from groundbreaking; two years later, unit 2 came online, with units 3 and 4 starting up in 1979 and 1981, respectively. These relatively short construction periods (by contemporary standards) are all the more remarkable given difficult weather conditions, manufacturing delays, and rapid labor turnover. Nuclear construction sites all over the country would continue to face similar challenges. The image shows the assembly of the reactor pit at Leningrad's unit 4. The banner reads: “It is our socialist obligation to deliver the reactor pit by the October anniversary—we will fulfill it!” This promise is then crossed out and replaced by: “We fulfilled it!”

Source: Courtesy of the Publishing House “Master,” Moscow. Originally published in Elektroenergetika (ed. A. I. Vol'skii and A. B. Chubais, series Stroiteli Rossii: XX vek (Moscow: “Master,” 2003), 234–235). Reprinted with permission.

The second reactor at Sosnovyi Bor came online in 1975, and the first RBMK at the Kursk nuclear power plant a year later. Over the next ten years, a total of fourteen RBMK-1000 reactors (at sites near Leningrad, Smolensk, Kursk, and Chernobyl) went operational, in addition to one RBMK-1500 at the Ignalina nuclear power plant in Lithuania. Their capacity would eventually constitute half of all Soviet nuclear power plants, and in 1985, they produced 60 percent of the Soviet Union's nuclear electricity.¹⁰⁴ After Chernobyl, but before the collapse of the Soviet Union, a second 1500 MW RBMK in Lithuania as well as a third 1000 MW unit at the Smolensk site started up.

In contrast to the Beloiarsk reactors, which used steel in the core, RBMKs from the outset relied on zirconium—a material whose characteristics designers considered superior for use in the reactor core.¹⁰⁵ Overall, the RBMK underwent no fewer than seven distinct modifications. Former Minenergo deputy minister Gennadii Shasharin grouped them into three generations, which differ with regard to their emergency cooling and their accident localization (containment) systems.¹⁰⁶

The first generation of RBMKs comprises the first two units at Leningrad and the first two units at Kursk and Chernobyl, all of which went critical between 1973 and 1979. Even within that first generation, there are differences, for example in plant layout: at Kursk and Chernobyl, the two reactors that share auxiliary systems are closer together.¹⁰⁷ Construction of the first-generation RBMKs began before the first edition of the “General Regulations for Nuclear Power Plant Safety” took effect.¹⁰⁸ These reactors did not have a backup system in place should cooling pipes break. They effectively had no emergency core cooling system, nor an accident localization system.¹⁰⁹

Units 3 and 4 at Leningrad (commissioned in 1979 and 1981 respectively) do have an emergency cooling system, an accident localization system, as well as an improved emergency power supply. They represent the first units of the second-generation RBMK, the generation that comprises the largest number of reactors in operation. It includes units 3 and 4 at Kursk, units 3 and 4 at Chernobyl, and units 1 and 2 at Smolensk. These six units were commissioned between 1978 and 1983. Some design features were further improved, and these reactors boast even more safety features. Unit 3 at Smolensk featured improved techniques for maintenance and control. And finally, the two RBMKs at the Ignalina nuclear power plant in Lithuania achieved a significantly increased thermal capacity, and thus electrical output, without increasing the size of the reactors. The increase was achieved instead by better heat removal and by increasing the flow rate in the pipes’ steam sections. Control, protection, and safety systems were modernized, and some automated control system processes changed.

Finally, the fifth unit at the Kursk nuclear power plant, whose construction was halted after the Chernobyl disaster and has proceeded in starts and fits ever since, embodies the third generation of RBMKs. The most significant design change realized there was reducing the graphite volume and the shape of the graphite columns and so the steam coefficient of reactivity.¹¹⁰ This reactor constituted an important showcase for the Soviet nuclear community: scientists hoped it would prove that RBMKs were not inherently dangerous reactors, as some claimed after Chernobyl. This reactor would demonstrate that RBMKs could be made safe. For several years in the early 2000s, nuclear officials considered a new reactor design, the multiloop reactor MKR, for the country's nuclear future. Based on the RBMK design and the experience accumulated by operating these reactors, in theory it would retain these reactors’ “positive features, [eliminate] the negative ones and [absorb] all the favourable potential features of inherent UGR [uranium graphite reactor] safety.” ¹¹¹

Soviets often touted the RBMK as the “national feature of domestic nuclear power,” because this particular design was unique to the Soviet Union.¹¹² The RBMK allowed the nuclear industry to expand based on an existing supply industry and experience with production reactors while it avoided the problems associated with the manufacture of unique pressure vessels and steam generators.¹¹³

Gas-Cooled Reactors

The pressurized water reactor emerged as the most common design internationally, which allowed various countries to exchange operating experience. Still, Laboratory No. 2 (and later LIPAN) developed two other functional reactor designs that to some degree competed with the VVER and the RBMK. Between 1947 and 1955, with an eye on the United Kingdom, which chose gas-cooled reactors for its civilian program, Kurchatov's reactor designers worked on gas-cooled reactors with helium as the primary coolant. Aleksandrov himself proposed a reactor called Sharik that combined a graphite moderator, uranium dioxide fuel elements, and helium coolant.¹¹⁴ His lab originally scheduled tests at Obninsk, but in 1955 that work lost its funding to the VVER. Only in 1957 did designers resume work on gas-cooled reactors, this time focusing on high-temperature reactors with carbon dioxide coolant as the British were doing. Goncharov notes that the sixth five-year plan (1956–1960) contained the development of gas-cooled reactors with an electrical power output of 50–100 MW for commercial nuclear power plants. Construction of a gas-cooled third reactor at the Beloiarsk site started in 1955 but was halted in 1963 in favor of a fast neutron design (the BN-600).¹¹⁵ A group of scientists under Nikolai Ponomarev-Stepnoi experimented with spherical fuel elements for high-temperature gas-graphite reactors, and in 1968 planners considered gas-cooled designs along with the VVER and the RBMK for the Kursk and Chernobyl sites.¹¹⁶

The erratic development of gas-cooled reactors continued into the late 1970s, when designers drew on experiments with spherical fuel elements for another graphite-helium design idea, the ABTU reactor. In May 1974, the government decided to build such a high-temperature, gas-graphite reactor with an electrical power output of 50 MW near Obninsk. But ultimately, the Soviet Union never built any gas-cooled design except as a research reactor.¹¹⁷

Heavy-Water Reactors

Research on heavy-water reactors started at Laboratory No. 2 during the atomic bomb project, with the goal of producing plutonium.¹¹⁸ Heavy water (deuterium) works particularly well as moderator because it slows neutrons down without absorbing many, and it can be combined with natural (not enriched) uranium. But heavy water was a scarce resource in most of Europe, and expensive to produce. In 1949, a heavy-water research reactor began operation at Laboratory No. 3, which was subsequently renamed the Institute of Theoretical and Experimental Physics (ITEF).¹¹⁹ In 1957, together with Czechoslovakian engineers, Soviet nuclear specialists started building the first heavy-water moderated power reactor, the model KS-150, at a site near Bohunice (which today is in Slovakia).¹²⁰ This joint project was part of the Council of Mutual Economic Assistance (CMEA) nuclear program, and the reactor went critical in December 1972. Notably, this was the first and only time the Soviets built a reactor type outside the borders of the USSR before first building and operating it within the country.¹²¹ The Bohunice reactor (A-1) suffered a major accident in 1979, and neither Czechoslovakia nor the Soviet Union pursued the heavy-water design for future nuclear power plans.¹²² Instead, Soviet nuclear planners decided to focus on developing the VVER and researching breeder reactors.

Competition and Standardization: Selecting “Soviet” Reactors

Given that Soviet decision makers and planners had a panoply of choices in the mid-1960s, why did they finally choose the VVER and the RBMK as the standard Soviet designs for the emerging nuclear industry?

The remainder of this chapter presents a more nuanced view of the period from the early 1960s to the mid-1970s, when the decision to produce electricity on a large scale in nuclear reactors provoked significant organizational restructuring in both the nuclear weapons sector and the power industry. Administrators selected the VVER and RBMK as major political and economic reforms brewed domestically. In addition, renewed international scientific exchange that offered insights into, and allowed comparisons with, trends in reactor engineering elsewhere in the world informed their choices.

Beginning in the early 1970s, the nuclear power industry finally acquired momentum all over the Soviet Union, as well as in allied socialist countries.¹²³ Between 1975 and 1986, thirty-three reactors started up in the USSR (among them sixteen VVERs and fourteen RBMKs, for a total capacity of 27,764 MW), compared to only thirteen reactors (among them six VVERs and one RBMK, for a total capacity of 4,014 MW) during the twenty years from 1954 to 1974 (figure 4.3).¹²⁴ In the face of this tremendous growth, why didn't Soviet economists and policymakers push just one reactor design? Or—conversely—why did they not choose more than two? After all, in 1962, the authorities had approved four designs for industrial-scale construction.¹²⁵

Figure 4.3 Between 1954 and 1993, fifty-five nuclear power reactors were started up in the former Soviet Union: twenty-nine VVERs, seventeen RBMKs, and nine others. Until the early 1980s, the RBMK made major contributions to nuclear electricity, while the VVER caught up only by 1987. Shown are the types of reactors that started delivering electricity to the grid each year, and the growth in nuclear power generation (in GW) overall. The figure takes into account the shutdown of units 1 and 2 at Beloiarsk in 1981 and 1989 (100 and 200 MW, respectively), the shutdown of units 1 and 2 at Novo-Voronezh in 1984 and 1990 (210 and 365 MW, respectively), the shutdown of the two 440 MW units at the Armenian nuclear power plant in 1989 as well as the subsequent restarting of the second unit in 1993, and finally, the loss of Chernobyl's unit 4 in 1986 and the shutdown of unit 2 in 1991 (1000 MW each).

Source: Graphic design by Dane Webster.

As we have seen, Soviet planners often avoided putting their eggs in one basket and instead supported multiple parallel research directions.¹²⁶ Then too, designing reactors for power generation in the 1950s and 1960s was a surprisingly uncoordinated effort, despite the planned character of the economy and the tight economic constraints at the time. Although existing infrastructure and comparatively long experience operating military graphite-water reactors no doubt contributed to the decision to adopt the RBMK as one of the two standard reactors in the Soviet Union, what counted as rational in the 1960s Soviet Union may not appear rational to us in hindsight. It is important to remember that this decision could have been made differently. While economic, technical, and diplomatic imperatives certainly mattered, these imperatives also competed with each other—consequently no single factor can on its own explain the specific trajectory the Soviet nuclear industry took. By documenting in detail the technical, economic, and institutional histories of these decisions, as well as the historical contingencies, constraints, and experiences that shaped these decisions, we can begin to acknowledge that Soviet scientists, engineers, and planners did not take decisions of such magnitude lightly. The extended negotiations I chronicle here ultimately formed the collective judgment of a community of experts. In light of Chernobyl's dramatic failure, domestic and international parties turned their scrutiny on this judgment. To more fully understand why Soviet nuclear experts made the choices they did, it is pivotal to draw together the topics we have touched on in previous chapters: the context of a planned, command-administrative economic system, the historical setting of the Cold War, and the emerging institutional structures—domestically and internationally—that governed nuclear energy.

Justifying the Choice

Soviet decision makers’ 1965 decision to select the RBMK as one of the civilian program's reactors depended on several assumptions. They assumed, first, that they could convert a reactor designed to produce plutonium into a power reactor; second, that this reactor would technologically and economically surpass the graphite-water reactors already operating; and third, that the design and construction bureau that had built the military reactors could quickly produce enough new civilian reactors that they would—in combination with the pressurized water reactor—actually fulfill the plan to rapidly expand the nuclear industry. In other words, instead of choosing technically outstanding designs from the available prototypes or waiting until operating reactor types could prove or disprove their economic merit, Soviet planners chose designs they thought would meet ambitious plan targets for nuclear power generation. By 1980, seven RBMK-1000s were producing electricity, which gave their designers cause for pride, and a little poetry:

Говорят что в СССР	They say that in the USSR
Будут толъко ВВЭР	there will be only the VVER
Но энергию пока	but until now, as we can see,
Нам дают РБМК.	the RBMK supplies our energy.127

Logistical Challenges and Eager Allies

In 1964, the VVER looked like a winner. Not only was it already operating successfully, with a larger version under construction, but it was also the design other countries preferred. And yet, by 1965, nuclear energy officials realized that mass production of VVERs would take longer than anticipated. The one factory in the entire Soviet Union that could manufacture the VVER's pressure vessels, the Izhora works in Leningrad, was already inundated with other assignments.¹²⁸ A new, gigantic factory to produce these vessels was under construction but plagued by construction glitches, delays, and insufficient funding. The VVER, it turned out, would not be able to carry the nuclear industry's rapid expansion alone (figure 4.4).¹²⁹

Figure 4.4 Builders of Atommash in Volgodonsk, a city in Southern Russia. The giant factory, which dwarfs the city built next to it, was intended to mass-produce nuclear reactor equipment. The picture shows workers who participated in the construction of the factory.

Source: Photograph by L. Nosov, 1977. Provided to Wikimedia Commons by the Russian International News Agency (RIA Novosti) as part of a cooperation project. RIA Novosti archive, image #587071/L. Nosov/CC-BY-SA 3.0.

The industry needed a second reactor type, one that different branches of industry could supply with materials and equipment.¹³⁰ The RBMK promised just that.¹³¹ It required graphite, cement, and pipe suppliers, while VVERs depended on high-quality, factory-based steel forging. The supply industry for the graphite-water line already existed, because military reactors relied on the same ingredients. By contrast, the industry supplying the VVER equipment was still in its early stages.

Officials thought parallel development of the VVER line and a second reactor line that drew on an already-established industry would let nuclear power expand rapidly. A quick review of Leningrad's history with regard to nuclear power illustrates how that decision turned out. The government first decreed a nuclear power plant near this city on March 16, 1956. It envisioned a nuclear heating plant with a VVER, a 200 MW pressurized water reactor. But a year later, on April 4, 1957, leaders dropped these plans. When the nuclear industry reemerged in the 1960s, they changed the site, the type of reactor, and the ministry in charge of construction. As mentioned above, it was ultimately Sredmash that would build the country's first set of RBMKs 50 miles from Leningrad at Sosnovyi Bor.¹³²

Another argument often made in favor of the RBMK was its relative ease of assembly. Its parts could be installed on site, the concrete produced locally, and even the nuclear core of the plant could be largely assembled on site.¹³³ When engineers scaled up the VVER from 440 to 1000 MW, they had hardly any leeway to enlarge the vessel. Instead, they had to rely on different materials to increase the core's performance.¹³⁴ The RBMK's modular design, by contrast, allowed engineers to start out with a large 1000 MW reactor and to scale it up later without the constraints of a factory schedule or railroad specifications.

Finally, reactor selection had an international dimension as well. Several of the Soviet Union's Eastern European neighbors were pressuring the USSR to provide them with nuclear assistance. In the wake of Eisenhower's “Atoms for Peace” speech at the United Nations General Assembly in December 1953, the Soviet Union—reluctantly—signed agreements that obligated the country to start delivering nuclear power reactors (in addition to research reactors) to the German Democratic Republic, Czechoslovakia, Hungary, and Bulgaria.¹³⁵ And even though the Soviets heavily delegated manufacturing tasks to individual factories in Eastern Europe, obligations beyond their own borders put additional strains on the domestic supply line for the VVER.The limitations in industrial capacity, the logistical challenges, and the international obligations may explain why the VVER could not carry the expansion of the Soviet nuclear industry alone, but they do not explain why the RBMK was chosen as the second type. Another graphite-water design was already in operation at Beloiarsk, and a technologically innovative, scaled-up version was under construction.¹³⁶ Other existing reactor prototypes, such as the heavy-water and gas-cooled designs discussed earlier, relied on different branches of industry just as the RBMK did. The width of train bridges seems a strangely tame obstacle for a state that aspired to nothing less than world revolution.¹³⁷ On-site assembly, while likely cheaper than factory manufacturing, had its obvious downside: less control over the purity of materials and over how well workers put the parts together. Overall, historical accounts that use economic justifications to explain Soviet reactor design choices seem to be missing significant pieces of the puzzle.

The Expertise Factor

When nuclear experts and planners vetted reactor designs, they also considered the availability of experienced personnel. They selected reactor types with military origins because they saw the military-style discipline enforced at plutonium production facilities and during naval reactor operations as a distinctive asset. They expected that members of that workforce would now help a growing number of civilian engineers master nuclear power reactor operations.

And yet, as I have mentioned, even the engineers who worked at naval or military production reactors had often trained in general engineering programs and learned about nuclear facilities only on the job. These engineers’ experience would not translate seamlessly, and they would not automatically find themselves in a position of authority vis-à-vis ambitious civilian engineers.

Actually, few people anywhere in the country had much experience with nuclear reactors for electricity production. Despite its tremendous symbolic value, the reactor at Obninsk was little more than a test and training plant, and only the operating personnel at Beloiarsk and Novo-Voronezh had some limited experience with operating power reactors. Many of the early managers and operators from these two initial plants went on to outstanding careers in the nuclear power industry.¹³⁸ But while decision makers argued for the VVER in part by pointing out that Novo-Voronezh personnel had experience operating a pressurized water reactor, they did not acknowledge that Beloiarsk personnel (mostly energetiki) had similarly accumulated experience with graphite-water reactors, experience that compared favorably with that of the technical experts familiar with military reactors.

As I showed in chapter 3, when Minenergo began to develop curricula that prepared engineering students and credentialed engineers for careers in the nuclear industry in the late 1950s, the emerging cooperation gradually began to produce a new cadre of nuclear power experts.

Designers’ Dilemmas and the Global Stage

Converting graphite-water reactors designed for producing weapons-grade plutonium to machines optimized for electricity generation involved immense technical challenges—and so did modifying naval propulsion reactors to power reactors of entirely different size, purpose, and core physics. Aleksandrov cooperated closely with Dollezhal and Slavskii to promote the VVER and RBMK designs. The three men provided the RBMK in particular with the patronage it needed to become an unexpected front runner.

Soviet scientists and reactor engineers eagerly advertised their early advances in nuclear technology—for example, by taking Western visitors on tours of select facilities.¹³⁹ They closely observed the design choices France, Britain, and the United States made for civilian reactors.¹⁴⁰ Recognition abroad always brought prestige for Soviet science at home, and Soviet nuclear specialists utilized international contacts to promote their ideas with domestic decision makers. But different groups leveraged such international connections in different ways. For instance, VVER advocates invoked international scientific consensus about the pressurized water design being a safe, economical, and deservedly popular option. They emphasized international harmonization in reactor engineering, and they argued that choosing pressurized water reactors would facilitate cooperation and accelerate learning among countries operating similar nuclear reactors.¹⁴¹

Supporters of the RBMK, by contrast, insisted that their design was uniquely Soviet and condescendingly referred to the VVER as “the American reactor.” They stressed that Soviet scientists and engineers had developed graphite-water reactors without any assistance from abroad.¹⁴² As in other nations, however, exactly what was Soviet about the Soviet program was tricky to identify in the context of international exchange (both licit and illicit).¹⁴³ Nevertheless, the alleged indigenous purity of the RBMK design was a powerful argument in the struggle for legitimacy within the Soviet nuclear power program. Like their French counterparts, Soviet reactor designers used similarities with, and differences from, other countries’ reactor designs as rhetorical resources to promote and contest specific proposals.¹⁴⁴ But unlike in France, where different institutions developed competing designs, both Soviet designs originated in the same design institute, Kurchatov's Institute of Atomic Energy.

Technology Triumphs

Decision makers made technical superiority the ultimate justification for the two reactor types they chose. The argument that both the RBMK and VVER were technologically advanced has even survived the Chernobyl disaster.¹⁴⁵ Those who promoted this view defined technical excellence in terms of imaginative solutions appropriate for the specific context of the Soviet nuclear industry, but also in terms of universal criteria such as demonstrated operation (prototypes) and safety features.

But the argument that the VVER and RBMK were the most advanced reactor technologies at the time, and in particular, that the adoption of the RBMK was a natural development, the normal outcome of a logical process, does not hold. As I pointed out above, when the Soviet government decided to equip future nuclear power plants with RBMKs, Sredmash had not even approved the technical design. Modifying the design of plutonium production reactors turned out to be a complex task that took much longer than planners anticipated. In addition, international observers in Geneva were unenthusiastic about the “Siberian” reactor, a direct predecessor of the RBMK.¹⁴⁶ Nuclear experts elsewhere considered the RBMK design neither technologically novel nor particularly worrisome by international standards—overall, they saw it as a functioning but dull machine. In the Soviet Union, however, planners saw its potential both as an official propaganda tool to advertise Soviet technical prowess and uniqueness, and, less publicly, as a backup option to satisfy national security interests.

Conclusion

The dominant historiography of the Soviet nuclear power program needs revision. In contrast to accounts that foreground the economic or technical advantages of certain reactor designs, I argue that those who chose the RBMK and VVER did not know at the time which of the many reactor types available would offer economic advantages or technical superiority. The RBMK was neither the cheapest nor the technologically most sophisticated design available, nor did it have the most substantial track record. Soviet nuclear experts (and experts in other nations) chose reactor designs not because they were the best or most functional ones available. The designs they picked worked, and they probably worked at least as well as others, but factors such as the material constraints of Soviet industry, familiarity with operating similar designs, and the organizational status of different research institutes better explain why some technical configurations acquired momentum and eventually succeeded, while others faltered.

Critics of the RBMK attribute the design's selection to dysfunctional organizations, individual career ambitions, and deliberate recklessness; they argue, with a dash of nationalist rhetoric, that this choice set the world on a direct course toward the Chernobyl disaster. I disagree. We don't need to accept (or reject!) all the details of official Soviet historiography on the nuclear power industry to acknowledge that all technological development involves ambitious engineers, political alliances, and, yes, risk taking. The degree to which strong, independent regulatory agencies, incentives for competition, or public debate keep such factors in check will vary by country, industry, and historical period. A line of reasoning that triumphantly identifies in retrospect where things went wrong obscures the fact that what we consider good and safe always depends on context. The RBMK was a design that only made sense in the specific context outlined above. In this specific context, however, selecting the RBMK made very good sense.