In 1980 NIKIET completed a confidential study that listed nine major design failings and thermohydraulic instabilities which undermined the safety of the RBMK reactor. The report made it clear that accidents were not merely possible under rare and improbable conditions but also likely in the course of everyday operation. Yet they took no action to redesign the reactor or even to warn plant personnel of its potential hazards. Instead of engineering new safety systems, NIKIET simply revised the operating instructions for the RBMK-1000. After decades of accident-free operation of military reactors, the atomic chieftains of NIKIET and the Kurchatov Institute apparently believed that a well-written set of manuals would be enough to guarantee nuclear safety. The designers assumed that, so long as they followed the new instructions closely, human beings would act as promptly and unfailingly as any of the plant’s electromechanical safety systems.
But the staff of Soviet nuclear power plants, faced with ever-increasing production targets and constantly malfunctioning or inadequate equipment, and answerable to a bewildering and dysfunctional bureaucracy, had long become accustomed to bending or ignoring the rules in order to get their work done. And the updated instructions they received from NIKIET were neither explicit nor explained. One of the new directives stipulated that a minimum number of control rods should henceforth be maintained within the core at all times—but NIKIET did not emphasize that this limit on the operational reactivity margin, or ORM, was a crucial safety precaution intended to prevent a major accident. Deprived of information about why such rules might be important, the operators went on with their work as usual, ignorant of the potentially catastrophic consequences of breaching them.
Meanwhile, every accident that did occur at a nuclear station in the Soviet Union continued to be regarded as a state secret, kept even from the specialists at the installations where they occurred.
* * *
Early on the evening of September 9, 1982, Nikolai Steinberg was sitting at the desk in his third-floor office between Chernobyl Units One and Two, overlooking the vent stack shared by the two reactors. Steinberg, a thirty-five-year-old with a short goatee and an easy charm, had worked in Chernobyl since 1971, arriving straight from the Moscow Power Engineering Institute as a graduate in nuclear thermal hydraulics and one of a new breed of bright-eyed atomshchiki. He had spent more than two years studying the RBMK at college before the first of the reactors had even been built, watched the first two units of the Chernobyl plant emerge from the ground, and was now chief of the turbine department for Units Three and Four. When Steinberg saw steam rising from the top of the vent stack, he knew it meant trouble: a broken pipe inside the reactor at least, and certainly a release of radiation. He picked up the phone.
But when he got through to the control room of Unit One to warn the operators there to shut down the reactor, the shift supervisor brushed him off. When Steinberg persisted, the supervisor hung up on him. The engineer gathered his staff and waited to be summoned for the emergency. But no call came. Almost six hours later, at midnight, he and his men got into their cars and drove home to Pripyat.
When he returned to work the following morning, Steinberg heard that there had indeed been a problem in Unit One, but—despite his seniority and experience—could learn nothing further. Director Brukhanov and the chief engineer of the plant initially insisted that whatever had happened had caused no radioactive releases, and local KGB officers took measures “to prevent the spread of panic-mongering, provocative rumors, and other negative manifestations.” In fact, radioactive contamination, carried on the wind and brought down by rain showers, had reached Pripyat and spread as far as fourteen kilometers from the plant. It included iodine 131, fragments of uranium dioxide fuel, and hot particles containing zinc 65 and zirconium-niobium 95 consistent with partial destruction of the reactor core. Levels of radiation in the village of Chistogalovka, five kilometers from the station, were hundreds of times higher than normal. But a team from Soyuzatomenergo—the USSR’s atomic energy authority—contested these findings. Contaminated areas immediately around the plant were simply sluiced with water and covered with soil and leaves. In Pripyat, decontamination trucks dispensed foam on the streets, and Lenina Prospekt was discreetly spread with a fresh layer of asphalt.
A subsequent inquiry revealed that there had been a partial meltdown in Unit One. When the reactor had been brought back online after maintenance, one of the temperamental cooling valves had remained closed; the uranium fuel in the channel overheated, and it ruptured. No one was killed, but the damage took eight months to repair. Workers carried blocks of reactor graphite away in buckets and were exposed to significant amounts of radiation. The chief engineer took the blame and was demoted and reassigned to a job in Bulgaria. The incident was classified top secret, and those directly involved were forced to sign gag orders by the KGB. Nikolai Steinberg would wait years before learning the truth about what had happened.
In the years that followed, there would be even more serious accidents at nuclear plants elsewhere in the Soviet Union, and all of them would be covered up. In October 1982 a generator exploded at Reactor Number One of the Metsamor plant in Armenia; the turbine hall burned down, and an emergency team had to be airlifted in from the Kola Peninsula, more than three thousand kilometers away in the Arctic Circle, to help save the core. Less than three years later, during the start-up of the first reactor at the Balakovo plant in Russia, a relief valve burst, and superheated steam at 300 degrees centigrade escaped into the annular compartments surrounding the reactor well. Fourteen men were boiled alive. Both incidents were concealed, and word reached the operators at other stations only through the atomshchiki rumor mill and hints in Pravda.
Yet the most dangerous suppression of all originated once more within NIKIET, the central nuclear design bureau in Moscow where the RBMK-1000 had been conceived. In 1983, on top of the myriad drawbacks of the reactor that had emerged since it went into operation, the reactor designers learned of one more: a curious design fault in the rods of the AZ-5 emergency protection system. The first conclusive evidence appeared at the end of the year, during the physical start-up of two of the newest RBMK reactors to be added to the Soviet grid: Unit One of the Ignalina nuclear power plant in Lithuania, and Unit Four in Chernobyl, the most advanced of the RBMK-1000 line.
While conducting tests before the two reactors could be brought into normal operation, the start-up teams of nuclear engineers in Ignalina and Chernobyl noticed a small but disturbing glitch. When they used the AZ-5 scram button to shut down the reactor, the control rods began their descent into the core, but instead of completing a smooth shutdown, the rods initially had the opposite effect: for a brief moment, reactor power rose instead of falling. The specialists discovered that the severity of this “positive scram” effect depended on the conditions inside the reactor at the moment the shutdown began—particularly on the ORM, the measurement indicating how many of the 211 control rods were withdrawn from the core. If more than 30 of these rods remained inserted when the scram began, the AZ-5 mechanism worked as intended, and the reactor would shut down quickly and safely. But when the total number of inserted rods fell below 30, the behavior of the reactor at shutdown became increasingly unpredictable, and the AZ-5 system struggled to do its job. With only 15 rods in, the technicians found that initial dampening of fission inside the reactor was marginal; it took six seconds before reactivity began to fall. And under some circumstances—7 rods or fewer—pressing the AZ-5 button might not shut down the reactor at all, but instead trigger a runaway chain reaction. If this happened, the increase in reactor power following an AZ-5 trip might be so great that it would no longer be possible to halt the reaction before the entire reactor was destroyed.
The source of the positive scram effect lay in the design of the control rods themselves, an unintended consequence of NIKIET’s desire to “save neutrons” and make the reactor more economical to run. Like all the manual control rods used to manage the reactor during normal operat
ion, the AZ-5 emergency rods contained boron carbide, a neutron poison that gobbles slow neutrons to reduce the chain reaction. But even when fully withdrawn from their water-filled control channels, the tips of the rods were designed to remain at the ready, just inside the active zone of the reactor—where, if they contained boron carbide, they would have a poisoning effect, creating a slight but constant drag on power output. To stop this from happening, the rods were tipped with short lengths of graphite, the neutron moderator that facilitates fission. When a scram shutdown began and the AZ-5 rods began their descent into the control channels, the graphite displaced neutron-absorbing water—with the effect of initially increasing the reactivity of the core. Only when the longer boron-filled part of the rod followed the graphite tip through the channel did it begin dampening reactivity.
It was an absurd and chilling inversion in the role of a safety device, as if the pedals of a car had been wired in reverse, so that hitting the brakes made it accelerate instead of slowing down. With further experiments, the technicians confirmed that the positive scram effect of the rod tips could create local criticality in the lower part of the giant RBMK core—especially if the operators tripped the AZ-5 system when the reactor was running at less than half power.
Alarmed, the director of the nuclear reactor department at the Kurchatov Institute wrote to NIKIET detailing the anomalies in the AZ-5 system and the need to examine them more closely. “It seems likely,” he warned, “that a more thorough analysis will reveal other dangerous situations.” Nikolai Dollezhal, the chief designer at NIKIET, replied with vaporous assurances: they were aware of the problem, and measures were being taken. But they weren’t. Although some partial modifications to the AZ-5 mechanism were approved, even these proved costly and inconvenient, and were executed piecemeal, just one RBMK reactor at a time. Gradually, Chernobyl’s Units One, Two, and Three were approved for the alterations. But Unit Four, already so close to completion, would have to wait for its first scheduled maintenance shutdown, in April 1986.
In the meantime, NIKIET circulated a notification of the positive scram effect to the senior managers of all RBMK plants. But, swept along in a blizzard of bureaucracy, tangled in secrecy, the news never reached the reactor operators. Nonetheless, as far as Anatoly Aleksandrov and the other nuclear chiefs were concerned, the redoubtable RBMK-1000—the Soviet national reactor—had been troubled by nothing more than temporary setbacks. By the time Viktor Brukhanov put his final signature on the paperwork to acknowledge completion of the fourth reactor of the V. I. Lenin nuclear power plant on the last day of 1983, the world still spoke of only one nuclear accident. And that humiliation belonged entirely to the United States.
* * *
Early in the morning of March 28, 1979, a handful of water-purifying resin beads smaller than mustard seeds blocked a valve in the secondary coolant loop of Reactor Two of the Three Mile Island nuclear power station near Harrisburg, Pennsylvania. Over the next twenty-four hours, the ensuing cascade of minor equipment malfunctions and human error led to a serious loss of coolant, which partially uncovered the core. The reactor began to melt down and contaminated the containment building with thousands of gallons of radioactive water. The staff had no choice but to vent radioactive gases directly into the atmosphere. Although no one was harmed by the released radiation—contained entirely in a cloud of short-lived isotopes of inert gases that drifted out over the Atlantic Ocean—news of the accident caused widespread panic. The roads of the tristate area were clogged as 135,000 people fled their homes in Pennsylvania. President Jimmy Carter—who had served as a nuclear engineer in the US Navy and knew a disaster when he saw one—attended the scene. The international antinuclear movement, which had slowly been gaining momentum over the previous decade, could not have asked for a more fearsome totem of a dangerous technology slipping its leash. In the United States, the development of the nuclear power industry, already dogged by rising construction costs and growing public apprehension, halted almost overnight.
As bad as it made the United States look, news of Three Mile Island was censored inside the USSR, for fear it could tarnish the ostensibly spotless record of the peaceful atom. Publicly, Soviet officials attributed the accident to the failings of capitalism. Academician Valery Legasov, Aleksandrov’s immediate deputy at the Kurchatov Institute, published an article insisting that the events in Three Mile Island were irrelevant to the USSR’s nuclear industry because its operators were far better trained and its safety standards higher than those in the United States. Privately, Soviet physicists began to analyze the likelihood of severe accidents at atomic power stations and revised their existing nuclear safety regulations. But neither Sredmash nor NIKIET made any attempt to bring the RBMK reactor into line with these new rules.
In January 1986 the new issue of Soviet Life—a glossy, English-language magazine resembling the stalwart American title but published by the Soviet embassy in the United States—featured the Chernobyl nuclear power plant as the centerpiece of a ten-page report on the wonders of nuclear energy. The special section included interviews with the residents of Pripyat, the city “Born of the Atom”; color photographs of the plant; and pictures of smiling station staff. Legasov coauthored another essay in which he boasted, “In the thirty years since the first Soviet nuclear power plant opened, there has not been a single instance when plant personnel or nearby residents have been seriously threatened; not a single disruption in normal operation occurred that would have resulted in the contamination of the air, water, or soil.”
In a separate interview, Vitali Sklyarov, the Ukrainian minister of energy and electrification, assured readers that the odds of a meltdown at the plant were “one in 10,000 years.”
5
* * *
Friday, April 25, 11:55 p.m., Unit Control Room Number Four
Beneath the sickly fluorescent strip lights of Control Room Number Four, a rancid haze of cigarette smoke hung in the air. The midnight shift had only just arrived, but the mood was growing tense. The turbine generator test scheduled to finish that afternoon had not yet begun. The station’s deputy chief engineer for operations, Anatoly Dyatlov, was now entering his second day without sleep. He was exhausted and unhappy.
The test was intended to check a key safety system designed to protect Reactor Number Four during an electrical blackout. A total loss of external power providing electricity to the station from the grid was something for which the RBMK designers had planned. It was one scenario of the so-called design-basis accident—in which the plant suddenly lost power, and the giant coolant pumps that kept water circulating through the reactor core came to a rumbling halt. The station had emergency diesel generators, but it would take between forty seconds and three minutes for these to start up and get the pumps going again. This was a dangerous gap—long enough for a core meltdown to begin.
So the reactor designers devised what they called a “rundown unit”: a mechanism to use the momentum of the unit’s turbines to drive the pumps for those crucial seconds. The rundown unit was a crucial safety feature of Reactor Number Four and was supposed to have been tested before it was approved for use in December 1983. But Director Brukhanov had granted approval to skip the test, to meet his end-of-year deadline. And although similar trials had been attempted since, they had failed each time. By the beginning of 1986, the test was more than two years overdue, but the reactor’s first scheduled maintenance shutdown offered an opportunity to conduct the trial in real-world conditions. At two o’clock on Friday afternoon, with new modifications made to one of the unit’s two huge turbine generators, Turbine Number Eight, it had—at last—been ready to begin.
But then the central dispatcher of the Kiev electrical grid intervened. Factories and enterprises throughout Ukraine were still in a frenzy of last-minute activity to meet their production quotas and win bonuses before the May Day holiday, and they needed every kilowatt of electricity the Chernobyl nuclear power plant could supply. The dispatcher said that Unit Four
could not go off-line to begin the test until after the peak load had passed—9:00 p.m. at the earliest.
By midnight on Friday, the team of electrical engineers waiting to monitor the test were threatening to cancel their contract and return to Donetsk if it didn’t start soon. In Control Room Number Four, the staff who had been briefed on the test program had reached the end of their shift and were preparing to go home. And the physicist from the plant’s Nuclear Safety Department—expected to be on hand to help the reactor operator through his part in the test—had been told the experiment was already complete. He hadn’t shown up at all. Stepping up to the instruments on the senior reactor control engineer’s desk, twenty-five-year-old Leonid Toptunov, just two months into his new job, prepared to pilot the capricious reactor through a shutdown for the first time in his life.
But Deputy Chief Engineer Dyatlov was determined to press on. If the test wasn’t completed that night, it would have to wait at least another year. And Dyatlov didn’t like to wait. At fifty-five, Anatoly Dyatlov looked every inch the austere Soviet technocrat: tall and gaunt, with sharp cheekbones, sparse, gray hair swept straight back from his high forehead, and narrow Siberian eyes that even in photographs seemed to glint with malice. A veteran physicist who had come to Chernobyl after fourteen years working on naval reactors in the Soviet Far East, Dyatlov was one of the three most senior managers at the station with nuclear expertise. He oversaw the operation of both Units Three and Four, with responsibility for hiring and training personnel.
Midnight in Chernobyl Page 9