Midnight in Chernobyl

by Adam Higginbotham

On September 29, 1966, the Soviet Council of Ministers, in Moscow, issued a decree approving the construction of the first in a new generation of giant water and graphite nuclear reactors, which would become known by the acronym RBMK, the reaktor bolshoy moschnosti kanalnyy, or high-power channel-type reactor. Developed from the Ministry of Medium Machine Building’s military workhorse, the plutonium-and-power-producing Ivan the Second, it was a direct descendant of the pioneering Atom Mirny-1 reactor, reimagined on an Olympian scale.

  Twelve meters across and seven meters high, the core of the RBMK was a massive cylinder, larger than a two-story house, composed of more than 1,700 tonnes of moderating graphite blocks, and stacked into 2,488 separate columns, each drilled from top to bottom with a circular channel. These channels contained more than 1,600 heat-resistant zirconium-alloy pressure tubes, each of which held a pair of metal assemblies packed with sealed rods of fuel: 190 tonnes of enriched uranium dioxide, compressed into ceramic pellets roughly the diameter of a man’s little finger. Once the reactor went critical and the uranium began to heat, releasing the energy of nuclear fission, the fuel assemblies were cooled by water pumped into the core from below. Under enormous pressure—sixty-nine atmospheres, or a thousand pounds per square inch—the water rose to 280 degrees centigrade and turned to a mixture of water and superheated steam, which was then piped out through the top of the reactor to giant separator drums. These directed the steam to turbines to generate electricity, while the remaining water returned to the beginning of the coolant loop to start its journey through the core once more.

  The power of the reactor was regulated by 211 boron carbide–filled control rods, most around five meters long, which could be raised or lowered into the reactor core to increase or decrease the rate of the nuclear chain reaction—and thus the level of heat, and energy, it generated. To help protect the plant and its staff from the radiation seething within, the reactor core—the active zone—was surrounded by a massive annular tank filled with water, contained within a steel jacket and surrounded by a giant box packed with sand. All this was further encased in a concrete vault more than eight stories high and crowned with a diadem of metal boxes filled with a mixture of iron shot and the neutron-retarding mineral serpentinite. A biological shield, a shallow stainless steel drum seventeen meters across and three meters deep, known as Structure E—or, more affectionately, Elena—sat on top of the vault like a giant lid. Filled with pebbles and rocks of serpentinite and nitrogen gas, Elena weighed two thousand tonnes—as much as six fully laden jumbo jets—and was held in place almost entirely by gravity. Pierced by ducts providing passage for the fuel channels, and surmounted by hundreds of narrow pipes carrying away steam and water, Elena was concealed beneath two thousand removable steel-clad concrete blocks, which capped the vertical fuel channels and formed the floor of the reactor hall. This tessellated metal circle, the visible face of the reactor during day-to-day operation, was known by the plant staff as the pyatachok, or five-kopek piece.

  * * *

  The RBMK was a triumph of Soviet gigantomania, a testament to its creators’ unrelenting pursuit of economies of scale: twenty times the size of Western reactors by volume, it was capable of producing 3,200 megawatts of thermal energy, or 1,000 megawatts of electricity, enough to keep the lights on for half the population of Kiev. Soviet scientists proclaimed it the “national” reactor of the USSR—not only technologically unique but the largest in the world. Anatoly Aleksandrov, the bald-headed director of the Kurchatov Institute of Atomic Energy, had personally taken credit for its design, which he filed as a classified invention with the Soviet patent office. In contrast to its principal Soviet competitor, the VVER—a complex piece of engineering derided by its detractors as “the American reactor” because of its similarity to the pressurized water reactors favored in the Unites States—the parts of the RBMK could be made in existing factories and didn’t require any specialized tooling. Its modular construction—hundreds of graphite blocks stacked into columns—meant that it could easily be put together on-site and scaled up to become even more powerful if necessary.

  Aleksandrov also saved money by dispensing with the containment building, the thick concrete dome built around almost every reactor in the West, intended to prevent radioactive contamination escaping from the plant in the event of a serious accident—but which, because the RBMK was so enormous, would have doubled the cost of building each unit. The subdivision of the reactor into 1,600 pressure tubes was adopted as a less costly solution designed to contain each pair of fuel assemblies in their own thin metal jacket—a feat of Byzantine plumbing that, its inventors argued, made a serious incident exceedingly unlikely. They also devised an accident suppression system that could cope safely with a simultaneous rupture in one or two of these tubes by safely directing the resulting release of high-pressure radioactive steam downward, through a series of valves, and into giant water-filled tanks in the basement beneath the reactor, where it would be cooled and securely contained.

  A break in the pressure tubes was one of the worst accidents the designers had ever prepared to encounter with the RBMK—a so-called maksimal’naya proektnaya avariya, or maximum design-basis accident. This designation also encompassed other potential calamities, including earthquakes, a plane crashing into the plant—or a complete rupture in one of the large-diameter water pipes in the reactor coolant circuit, which would deprive the core of water and trigger a meltdown. To guard against this last eventuality, the designers devised an emergency cooling system powered by compressed nitrogen gas, and reactor operators at every level of the industry were drilled to maintain a continuous supply of water to the core at all costs.

  Worse accidents were theoretically possible, of course: engineering calculations suggested that if more than 2—and as few as 3 or 4—of the 1,600 pressure tubes ruptured simultaneously, the sudden release of high-pressure steam would be enough to lift all two thousand tonnes of Elena and the pyatachok off their mounts, severing every one of the remaining steam lines and pressure tubes and resulting in a devastating explosion. Yet the designers saw no need to prepare for such a calamity, which they regarded as outside the realm of reasonable probability. Nonetheless, they granted the scenario its own designation: the beyond design-basis accident.

  The Ministry of Medium Machine Building ordered the first-draft plans of the RBMK to be drawn up by a heavy machinery plant in Leningrad, which also built tanks and tractors. But when they received the blueprints, Sredmash dismissed them as technically unsound. One scientist from the Kurchatov Institute warned that the design was too dangerous to be put into civilian operation. Another recognized that the hazards of the positive void coefficient made the new reactor inherently prone to explosion, and—although his superiors attempted to have him dismissed from the institute because of his dissent—he began a letter-writing campaign that eventually reached the Central Committee of the Communist Party and the Soviet Council of Ministers.

  But by then, the government—adhering to the rigid needs of central economic planning—had already issued its decree that four of the new behemoth reactors be built. So the designers of NIKIET scrambled to perform a drastic overhaul on the RBMK blueprints, transforming it from a schizoid contraption that could manufacture both plutonium and electricity into a tame generator of power for the civilian grid. Implementing these modifications was difficult and complex work and took far longer than expected: primitive Soviet computing technology made calculating the expected performance of the reactor laborious and produced unreliable results. It was not until 1968 that the new reactor design, now called the RBMK-1000, was complete. So, to save time, Sredmash decided to skip the prototype stage entirely: the quickest way to find out how the new reactors would work in industrial electricity generation would be to put them directly into mass production.

  Construction began on the first RBMK reactor in the Soviet Union at a Sredmash installation on the Gulf of Finland, outside Leningrad, in 1970. In the meantime, a pair of technica
l and economic institutes in Kiev considered possible locations for the first nuclear power station to be built in Ukraine, and quickly narrowed it down to two. When the first proposed site was earmarked for a fossil fuel plant, the Ukrainian Council of Ministers decreed that the republic’s new 2,000-megawatt atomic energy station would be constructed at the other: on a large patch of sandy riverbank near the village of Kopachi, in the Kiev region, fourteen kilometers from the town of Chernobyl.

  * * *

  The first RBMK unit at the Leningrad station started up on December 21, 1973, just one day before energetiki across the USSR celebrated their own national holiday, the Day of the Power Engineer. The proud fathers of the RBMK-1000, Anatoly Aleksandrov of the Kurchatov Institute and Nikolai Dollezhal from NIKIET, were both there to witness it come to life. By then, building was also under way on a second unit in Leningrad, and construction workers had broken ground at RBMK stations in Chernobyl and Kursk. But the initial Leningrad reactor had not even reached full power when it became clear that the designers’ determination to rush their brainchild from drawing board to full-scale production had come at a steep cost. Serious design faults dogged the RBMK from the outset. Many became apparent immediately; others would take much longer to come to light.

  The first problem arose from the positive void coefficient, the drawback that made Soviet graphite-water reactors susceptible to runaway chain reactions in the event of a loss of coolant, and that, in the RBMK, had been exacerbated by attempts to make the reactor cheaper to run. To make it more competitive with fossil energy power stations, the RBMK had been deliberately designed to maximize the electricity output of the uranium fuel it burned up. But it was only when they started up Leningrad Unit One that the designers discovered that the effects of the positive void coefficient grew worse as more of the fuel was burned; the longer it was in operation, the harder the reactor became to control. By the time it reached the end of each of its three-year operational cycles and was shut down for preventive maintenance, the RBMK would be at its most unpredictable. The designers made modifications, but instabilities remained. Yet neither Aleksandrov nor Dollezhal sought to explore these problems further, nor even to fully understand them—and provided no safety analysis of the void coefficient in the manuals accompanying each reactor. The results of the experiments in Leningrad made it obvious that there were important differences between the way they had predicted the reactor would perform in theory and how it worked in practice. But the designers decided not to examine these results too closely. Even as it went into full-scale commercial operation, nobody knew how the RBMK would behave during a major accident.

  A second failing of the reactor resulted from its colossal size. The RBMK was so large that reactivity in one area of the core often had only a loose relationship to that in another. The operators had to control it as if it were not a single unit but several separate reactors in one. One specialist compared it to a huge apartment building, where a family in one flat might be celebrating a raucous wedding, while next door another was observing a funeral wake. Isolated hot spots of reactivity might build deep inside the core, where they could prove hard to detect. This problem was especially pronounced during start-up and shutdown, when the reactor was operating at low power—and the systems designed to detect reactivity within the core proved unreliable. During these crucial periods, the engineers at their desks in the control room became almost totally blind to what was happening inside the active zone. Instead of reading their instruments, they were forced to estimate the levels of activity in the core, using “experience and intuition.” This made start-up and shutdown the most demanding and treacherous stages of RBMK operation.

  A third fault lay in the heart of the reactor’s emergency protection system, the last line of defense against an accident. If the operators faced a situation calling for an emergency shutdown—a major coolant leak or a reactor runaway—they could press the “scram” button, activating the ultimate stage of the unit’s five-level rapid power reduction system, known in Russian as AZ-5. Pushing this button would drive a special bank of twenty-four neutron-absorbing boron carbide control rods—as well as every one of the remaining 187 manual or automatic control rods that remained withdrawn at the time—simultaneously into the core, quenching the chain reaction throughout the reactor. Yet the AZ-5 mechanism was not designed to bring about an abrupt emergency stop. Dollezhal and the technicians of NIKIET believed that suddenly cutting off the electricity generated by the reactor would be disruptive to the operation of the Soviet grid. And they thought that such an immediate shutdown would be necessary only in the extremely unlikely event of a total loss of external power to the plant. So they designed the AZ-5 system to only gradually reduce the reactor’s power to zero. Rather than dedicated emergency motors, the system was driven by the same electric servos that moved the manual reactor control rods, used by the operators to manage reactor power during normal operation. Starting from their fully withdrawn position above the reactor, it would take between eighteen and twenty-one seconds for the AZ-5 rods to descend completely into the core; the designers hoped that the rods’ slow speed would be compensated for by their great number. But eighteen seconds is a long time in neutron physics—and an eternity in a nuclear reactor with a high positive void coefficient.

  Adding to this disquieting list of major design defects, the construction of the reactors also suffered from the shoddy workmanship that plagued Soviet industry. The full start-up of Leningrad’s Reactor Number One was delayed for almost a year after fuel assemblies became stuck in their channels and had to be returned to Moscow for repeated testing. The valves and flow meters in other RBMKs, used to regulate the crucial supply of water to each of the more than 1,600 uranium-filled channels, proved so unreliable that the operators in the control room often had no idea to what extent the reactors were being cooled, or if they were being cooled at all. Accidents were inevitable.

  On the night of November 30, 1975, just over a year after it had first reached full operating capacity, Unit One of the Leningrad nuclear power plant was being brought back online after scheduled maintenance when it began to run out of control. The AZ-5 emergency protection system was tripped, but before the chain reaction could be stopped, a partial meltdown occurred, destroying or damaging thirty-two fuel assemblies and releasing radiation into the atmosphere over the Gulf of Finland. It was the first major accident involving an RBMK reactor, and the Ministry of Medium Machine Building set up a commission to investigate what had gone wrong. Afterward, the official line was that a manufacturing defect had led to the destruction of a single fuel channel. But the commission knew otherwise: the accident was the result of the design faults inherent in the reactor and caused by an uncontrollable increase in the steam void coefficient.

  Sredmash suppressed the commission’s findings and covered up the accident. The operators of other RBMK plants were never informed of its true causes. Nevertheless, the commission made several important recommendations, to be applied to all RBMK-1000 reactors: develop new safety regulations to protect them in the event of coolant loss; analyze what would happen in the event of a sharp rise in steam in the core; and devise a faster-acting emergency protection system. Despite their apparent urgency, the reactor designers failed to act on a single one of these directives, and Moscow promptly ordered more of the reactors to be built. The day after the Leningrad meltdown, the Soviet Union’s Council of Ministers gave its final approval to construct a second pair of RBMK-1000 units in Chernobyl, expanding the station’s projected output to an impressive 4,000 megawatts.

  * * *

  On August 1, 1977, more than seven years after Viktor Brukhanov had watched the first stake being driven into the snow-covered ground beside the Pripyat, and two years later than planned, Reactor Number One of the Chernobyl nuclear power plant at last went critical. The plant’s young operators were overwhelmed with pride as they prepared to bring the Ukrainian republic’s first nuclear station online. They remained at their p
osts day and night as the first fuel assemblies were loaded and the reactor was slowly brought up to full power and, finally, connected to the transformers. At 8:10 p.m. on September 27, the scientists and designers from the Kurchatov Institute and NIKIET joined the plant specialists to celebrate as Ukraine’s first nuclear electricity ran into the 110- and 330-kilovolt lines and out to the Soviet grid. Together they sang the couplet with which atomshchiki across the Union hymned the success of the Soviet Reactor: A poka, a poka tok dayut RBMK! “For today, for today, current flows from the RBMK!”

  But the Chernobyl operators soon discovered that the reactor on which they had lavished so much attention was an unforgiving mistress. The inherent instabilities of the RBMK made it so difficult to manage that the senior reactor control engineers’ work proved not only mentally but also physically demanding. Making dozens of adjustments every minute, they were never off their feet and sweated like laborers digging a ditch. Rumors reached them that up in Leningrad, the Sredmash reactor engineers had doubled up on the control desk, “playing duets” to cope with the complexity of the task. The reactor operators worked the panel so hard that the switches governing the control rods quickly wore out and had to be replaced constantly. When one former nuclear submarine officer first took his seat at the desk in Chernobyl’s Unit One, he was horrified by the colossal size of the reactor and how antiquated the instrumentation was.

  “How can you possibly control this hulking piece of shit?” he asked. “And what is it doing in civilian use?”

  At their first planned maintenance shutdown, the Chernobyl operators found that the serpentine plumbing of the reactor was riddled with faults: the water-steam coolant pipes were corroded, the zirconium-steel joints on the fuel channels had come loose, and the designers had failed to build any safety system to protect the reactor against a failure of its feed-water supply—eventually, the Chernobyl engineers had to design and fabricate their own. Meanwhile, in Moscow, the reactor designers continued to discover further troubling flaws in their creation.