The son of a peasant who had lit the buoys each night on the Yenisei River near the penal settlements of Krasnoyarsk, Dyatlov had run away from home at fourteen. He went to vocational school and worked as an electrician before winning a place at MEPhI, the Moscow Engineering and Physics Institute. He graduated in 1959 and was posted to a stronghold of the Soviet military-industrial complex: the Lenin Komsomol shipyard in the remote city of Komsomolsk-on-Amur. As the head of the classified Laboratory 23, Dyatlov managed a team of twenty men who installed the reactors of the Yankee- and Victor-class nuclear-powered submarines before they put to sea.
By the time he arrived in Chernobyl in 1973, he had overseen the assembling, testing, and commissioning of more than forty VM reactor cores. These small marine reactors, versions of the VVER, were nothing like the massive graphite-moderated behemoths being constructed in Chernobyl. But Dyatlov, a fanatical specialist, threw himself into learning everything he could about the RBMK-1000. He was there for the commissioning of each one of the four new units in Chernobyl and now worked ten-hour days, six and sometimes seven days a week. He walked to the plant every day from his home in Pripyat—he found that walking helped to keep dark thoughts at bay—and jogged to keep fit. He was rarely in his office but prowled the corridors and gangways of the plant day and night, inspecting the equipment, checking for leaks and errant vibrations, and keeping tabs on his staff. A stickler for detail, Dyatlov knew his job and prided himself on his knowledge of the reactor and its systems—of mathematics, physics, mechanics, thermodynamics, and electrical engineering.
But the manner Dyatlov had developed while running a top-secret military laboratory did not prepare him well for managing the operators and engineers of a civilian nuclear power station. He had no tolerance for shirkers or those who didn’t follow his orders to the letter. Even those colleagues he brought with him from Komsomolsk found him hard to work with. He could be high-handed and peremptory, peppering his speech with curses and Soviet navy slang, muttering to himself about the inexperienced technicians he dismissed as chertov karas—fucking goldfish. He demanded that any fault he discovered be fixed immediately and carried a notebook in which he recorded the names of those who failed to meet his standards.
The deputy chief engineer believed he was always right and held stubbornly to his own convictions on technical matters, even when overruled from above. And Dyatlov’s long experience in the shipyard, and of the chimerical construction targets imposed in Chernobyl, had taught him that the categorical dictates of Soviet bureaucracy and the gray netherworld of Soviet reality were very different things.
Dyatlov had fulfilled every autodidactic expectation of Soviet Man, dedicating himself to his work by day and steeping himself in culture by night; he loved poetry and knew by heart all eight chapters of Pushkin’s epic Eugene Onegin. Away from work, he could be good company, though he had few close friends. Only long afterward would his secret emerge: before arriving in Chernobyl, Dyatlov had been involved in a reactor accident in Laboratory 23. There was an explosion, and Dyatlov was exposed to 100 rem, a huge dose of radiation. The accident, inevitably, was covered up. Later, one of his two young sons developed leukemia. There could be no certainty that the two events were linked. But the boy was nine when he died, and Dyatlov buried him there, beside the river in Komsomolsk.
Although the specialists who labored under Dyatlov at the Chernobyl plant may have disliked the way he treated them, many admired him, and few doubted his expertise. Eager to learn, they believed that he knew everything there was to know about reactors. And, by crushing dissent and conjuring an air of infallibility, Dyatlov—like the Soviet state itself—expected his underlings to carry out his commands with robotic acquiescence, regardless of their better judgment.
Yet the deputy chief engineer admitted a peculiar reservation about the reactor on which they all worked. For all his hours spent poring over the latest technical revisions and regulations, for all his mastery of thermodynamics and physics, Dyatlov said that there remained something unfathomable about the RBMK-1000: a nuclear enigma even he could never fully understand.
* * *
The control room of Unit Four was a large, windowless box, around twenty meters wide and ten meters deep, with a polished stone floor and a low suspended ceiling punctuated by the recessed fluorescent lights and ventilation ducts. Usually it was manned by a team of just four. At the back, the shift foreman had his own desk, from where he could observe the three operators who ran the unit, stationed at three long, gray, steel control panels arranged in a broken arc across the width of the room. On the left sat the senior reactor control engineer, known by the Russian acronym SIUR. On the right was the senior turbine control engineer. And in the center, linking the activity of the other two, the senior unit control engineer, who maintained the water supply that kept it all going—hundreds of thousands of cubic meters of it, flowing around the reactor’s primary loop: from the pumps, through the reactor, to the steam separators, out to the turbines, and back again. The three men’s control panels were festooned with the hundreds of switches, buttons, gauges, lamps, and annunciator alarms needed to manage the primary processes of spinning electricity from nuclear fission.
Forming a wall in front of the desks was a floor-to-ceiling bank of instrument panels showing the status of all three of the systems on illuminated dials, closed-circuit TV screens, and trembling pen-trace drums, which slowly logged data across scrolling bands of paper. Concealed behind the panel, and in anterooms to the left and right, were thousands of meters of cable leading away into darkness and banks of computer cabinets filled with glowing valves and clicking relays: the complex but antiquated technology that linked the control panels to the reactor itself.
As young Leonid Toptunov took over at the senior reactor control engineer’s desk, he faced two giant backlit displays reaching almost to the ceiling and showing the operating conditions within Reactor Number Four. One displayed the status of every one of its 1,659 uranium-filled fuel channels; the other was formed of 211 glowing dials arranged in a circle three meters across. These were the Selsyn monitors, which indicated the position of the boron carbide control rods that could be raised or lowered into the reactor to moderate its chain reaction. Beneath Toptunov’s hands lay the panel of switches with which he could select groups of rods, and the joystick that moved them in and out of the core. Nearby, a reactimeter showed the thermal output of the reactor in megawatts, displayed in glowing digital figures. Just behind him stood Shift Foreman Alexander Akimov, who would be responsible for supervising the test under the direction of Deputy Chief Engineer Dyatlov. In the strict technical hierarchy of the power plant, Akimov, an experienced reactor control engineer, was the senior member of the operational staff in the room. Dyatlov’s role was administrative: no matter how deep his nuclear expertise, he could no more take the controls of the reactor engineer’s desk than an airline executive could step onto the flight deck of one of his passenger jets and fly the plane himself.
Akimov, a gangling thirty-two-year-old with thick glasses, a receding hairline, and a small mustache, was a committed Communist and one of the most knowledgeable technicians at the plant. He and his wife, Luba, were the parents of two young boys, and he spent his spare time reading historical biographies or hunting hare and duck with his Winchester rifle on the Pripyat marshes. Akimov was clever, competent, and well liked, but his colleagues agreed that he was easily pushed around by those above him.
Control Room Number Four had now grown busy. In addition to Toptunov and the other two operators manning the turbine and pump panels, members of the previous shift lingered at their posts, along with others who had come to watch. In an adjacent room, the turbine specialists from Donetsk were on hand to monitor the rundown of Turbine Generator Number Eight. Dyatlov paced the floor.
As soon as the Kiev grid dispatcher gave permission, the operators had resumed the reactor’s long, controlled power descent and now held it steady at 720 megawatts—just above
the minimum level required to perform the test. But Dyatlov, perhaps assuming that a lower power level would be safer, was adamant that it be conducted at a level of 200 megawatts. Akimov, a copy of the test protocol in his hands, disagreed—vehemently enough for his objections to be noted by those standing nearby, who heard the two men arguing even over the constant hum of the turbines from the machine hall next door. At 200 megawatts, Akimov knew that the reactor could be dangerously unstable and even harder to manage than usual. And the program for the test stipulated it be conducted at not less than 700 megawatts. But Dyatlov insisted he knew better. Akimov, defeated, agreed reluctantly to give the order, and Toptunov began to decrease power further. Then, at twenty-eight minutes past midnight, the young engineer made a mistake.
When Toptunov had assumed responsibility for the reactor at midnight, the unit’s computerized regulation system was set to local automatic control, which allowed him to manage regions of the core individually—but was usually switched off when operating the reactor at low power. So Toptunov began the process of transferring the system to global automatic—a form of nuclear autopilot that would help him keep the RBMK on a steady course as the men prepared for the start of the test. Before completing the change, he was supposed to choose a level at which the computer would maintain reactor power in the new operating mode. But, somehow, he skipped this step. The reactor proved as unforgiving as ever. Bereft of fresh instructions, the computer defaulted to the last set point it had been given: near zero.
Now Toptunov watched in dismay as the glowing gray figures on the reactimeter display began to tumble: 500 . . . 400 . . . 300 . . . 200 . . . 100 megawatts. The reactor was slipping away from him.
A series of alarms sounded: “Failure in measuring circuits.” “Emergency power increase rate protection on.” “Water flow decrease.” Akimov saw what was happening. “Maintain power! Maintain power!” he shouted. But Toptunov could not stop the numbers from falling. Within two minutes, the power output of Unit Four had plunged to 30 megawatts—less than 1 percent of its thermal capacity. By 12:30 a.m., the reactimeter display was almost at zero. Yet for at least four more minutes Toptunov took no action. While he waited, neutron-scavenging xenon 135 gas began to build in the core, overwhelming what little reactivity remained. The reactor was being poisoned, plunging into what the operators called a “xenon well.” At this point, with the reactor’s power stalled at its minimum, and more xenon accumulating all the time, nuclear safety procedures made the operators’ course quite clear: they should have aborted the test and shut down the reactor immediately.
But they did not.
Later, there would be conflicting accounts of exactly what happened next. Dyatlov himself would maintain that he was absent from the control room when the power first fell—although he could not always recall exactly why—and issued no instructions to the operators at the senior reactor control engineer’s desk during the crucial minutes that followed.
The recollections of others present at the time would be quite different. According to Toptunov, Dyatlov not only witnessed the power fall but also—enraged—told him to withdraw more control rods from the reactor to increase power. Toptunov knew that to do so could certainly increase reactivity but would also leave the core in a dangerously unmanageable state. So Toptunov refused to obey Dyatlov’s command.
“I’m not going to raise the power!” he said.
But now Dyatlov threatened the young operator: if he didn’t follow orders, the deputy chief engineer would simply find another operator who would. The head of the previous shift, Yuri Tregub—who had stayed behind to watch the test—was well qualified to operate the board and right there at his elbow. And Toptunov knew that such insubordination could mean that his career at one of the most prestigious facilities in the Soviet nuclear industry—and his comfortable life in Pripyat—would be over as soon as it had started.
Meanwhile, the reactor continued to fill with poisonous xenon 135, falling deeper and more inextricably into the well of negative reactivity. Finally, six long minutes after the fall in power had begun, Toptunov, terrified of losing his job, gave in to Dyatlov’s demands. The deputy chief engineer withdrew from the console, mopping sweat from his brow, and returned to his position in the center of the room.
But reviving a poisoned reactor is not easy. At first, Toptunov struggled to find the right balance of manual control rods to withdraw. Standing behind him, Tregub noticed the young technician raising them disproportionately from the third and fourth quadrants of the core. The power continued to languish close to zero. “Why are you pulling unevenly?” the veteran engineer asked. “You need to pull from here.” Tregub began to prompt him on which rods to choose. As the buttons of the control panel clattered beneath Toptunov’s right hand, his left tugged at the joystick. The atmosphere in the control room once more grew tense. Tregub stayed at Toptunov’s side for twenty minutes, and together they managed to coax the reactor up to 200 megawatts. But then they could go no further. Xenon poisoning continued to gobble positive neutrons within the core—and they were running out of control rods to extract. More than a hundred of them already stood at their upper-limit stops.
By 1:00 a.m., Toptunov and Tregub had hauled the reactor back from the brink of an accidental shutdown. But to do so they had withdrawn the equivalent of 203 of the unit’s 211 control rods from the reactor core. To pull such a large proportion of rods without the authorization of the chief engineer of the plant was forbidden. Yet the engineers knew that the computer system monitoring the number of rods in the core—the operational reactivity margin—wasn’t always accurate, and they remained unaware of its importance to the safe running of the reactor. They did not suspect that the simultaneous reinsertion of so many rods into the core could trigger a reactor runaway. At this moment, only a careful stabilization of the reactor, followed by a slowly managed shutdown, might have headed off disaster.
Yet, now, two more of the giant main circulation pumps connected to the reactor came online. Although part of the original test program, the addition of these extra pumps had never been intended for such a low power level. Driving more cooling water into the core, they further upset the delicate balance of reactivity, water pressure, and steam content within the reactor. Running the pump system from his central control desk, twenty-seven-year-old Senior Unit Control Engineer Boris Stolyarchuk struggled to correct the water levels in the steam separator drums, as the pumps thundered toward their maximum capacity, forcing fifteen cubic meters of high-pressure coolant into the reactor every second. The rush of water absorbed an increasing number of neutrons in the core, dampening reactivity, and the reactor’s automatic regulation system compensated by withdrawing yet more control rods. A few moments later, the water was moving so fast around the cooling loop that it was entering the core at close to boiling and turning to steam, making the reactor more susceptible to the positive void effect if there was the slightest increase in power.
Now the time to begin the generator rundown had finally come. Some of the operators were clearly nervous. Yet Anatoly Dyatlov felt nothing but calm. The test would proceed, regardless of the fine print of the experiment protocol or the qualms of his subordinates. Ten men now stood ready around the desks and control panels of Control Room Number Four, eyes on their instruments. He turned to Akimov.
“What are you waiting for?” he asked. It was 1:22 a.m.
* * *
Simulating the effects of a total power blackout on a single unit of the Chernobyl power plant was a deceptively simple process, and many in the control room mistakenly regarded the rundown test as a matter mostly for the electricians. The reactor’s role seemed almost incidental. The test program closely duplicated one conducted on Unit Three in 1984—which, although it had failed to produce the desired results and keep the circulation pumps running, had nonetheless concluded without incident. Nikolai Fomin, the chief engineer, had ordered that test himself and without clearance from above and saw no reason to behave differen
tly this time. He did not notify the State Committee for Nuclear Safety, NIKIET, or the specialists at the Kurchatov Institute in Moscow of his plans. He didn’t even bother to tell Director Brukhanov that the test was taking place.
Emboldened by his previous experience, Fomin made two important changes with the new test: this time, all eight of the unit’s main circulation pumps would be connected to the reactor, increasing the amount of water driven through the primary circuit during the rundown. But he had also ordered a special piece of equipment, an electrical box designed to be patched in to the control panel circuitry for the test, which would reproduce the effects of the design-basis accident at the touch of a button. The new test program—drafted a month earlier by Gennadi Metlenko, the head of the team of electrical engineers from Donetsk, and approved by Fomin and Dyatlov in April—appeared straightforward.
First, the operators would cut off the supply of steam from the reactor to the turbine, which would begin to spin down. At the same moment, they would press the design-basis accident button. This would send a signal to the reactor’s safety systems that all external power to the plant had been lost, triggering the start-up of an emergency diesel generator and connecting the rundown unit of Turbine Number Eight to the main circulation pumps. If all went well, the electricity produced by the coasting turbine would keep the pumps going until the diesel generator could take over. The technicians expected the experiment to take less than a minute. It would begin with an order from Metlenko, who would record the results with an oscilloscope. It would conclude with the operators bringing the reactor to a routine shutdown by tripping the AZ-5 system for a full emergency stop.
By 1:23 a.m., at his desk in the control room, Leonid Toptunov had successfully stabilized the reactor at a power level of 200 megawatts. Dyatlov, Akimov, and Metlenko stood in the center of the room, awaiting the moment to begin. Upstairs at mark +12.5, in the cavernous three-story pump room alongside the reactor vault, Senior Coolant Pump Operator Valery Khodemchuk stood at his post, engulfed in the thunderous roar of all eight main circulation pumps working at once. At the bottom of the reactor core, the pressurized water was now entering the inlet valves at a temperature just a few degrees below boiling. And directly above them, 164 of the 211 control rods had been withdrawn to their upper-limit stops.
Midnight in Chernobyl Page 10
