Fukushima: The Story of a Nuclear Disaster Read online

  One scenario TEPCO had not anticipated, however, was that the vent might be needed during a station blackout, when the valves required to open it could not be operated remotely from the control room. Consequently, the emergency guide did not explain how to operate the valves manually. Nor was the vent equipped with filters to remove radiation from the steam if an emergency release was required. The designers of boiling water reactors believed that filters were unnecessary because radioactive steam would be naturally scrubbed by water in the torus before being vented. But this filtering mechanism had never been demonstrated under real-world conditions, and some experts doubted its effectiveness.

  These were two more holes in an emergency plan that was turning out to be full of them. “When on-site workers referred to the severe accident manual, the answers they were looking for were simply not there,” investigators would later write. “[T]hey were thrown into the middle of a crisis without the benefit of training or instructions.”

  Nobody was sure if the ability to relieve the rapid buildup of pressure inside a reactor was within reach. In a BWR accident, venting—a controlled release of radioactivity—is a last-ditch move to stave off a far worse disaster: core melting and failure of the containment, which could result in larger, uncontrolled releases of radioactivity. The amount of radiation released during venting depends on the extent to which the core has been damaged. If it is badly damaged, radiation levels in the steam could be deadly. At this point, no one knew the status of the Unit 1 core, so the relative risks and benefits of venting were not clear. But guided by scant data and instinct, the engineers knew that some sort of intervention to stabilize the reactor was needed—and needed quickly. They moved on two fronts: to bring in additional water supplies and to prepare, somehow, to vent.

  Venting, in addition to being technically difficult, was fraught with political and public relations implications; as a result, both TEPCO management and the government in Tokyo would demand a say. Ultimately, however, the decision to vent rested with the most senior official on the scene. At Fukushima Daiichi that was Masao Yoshida, fifty-six years old, who had become the boss ten months earlier.

  The superintendent of Fukushima Daiichi, Masao Yoshida. Tokyo Electric Power Company

  Like a ship’s captain, the site superintendent knows the equipment intimately, has firsthand knowledge of the unfolding crisis, and is best positioned to assess the options. Under TEPCO’s emergency plan, Yoshida was to make the calls with input from utility executives. At the moment, however, TEPCO’s top executives were missing in action.

  Chairman Tsunehisa Katsumata was in Beijing on a business trip with Japanese media owners. President Masataka Shimizu had been sightseeing with his wife in western Japan. Shimizu had received an earthquake alert on his cell phone, but his efforts to return to Tokyo that afternoon were thwarted. He had traveled only partway, hoping to make the final leg of the trip home in a TEPCO helicopter. But Japan’s civil aviation law bars private helicopters from flying after 7:00 p.m. Late in the evening, he won approval from the government to fly aboard a Japan Air Self-Defense Force (SDF) airplane, but, twenty minutes after the 11:30 p.m. takeoff, the defense minister (unaware of the official authorization) ordered the plane to turn around and instead stand by for disaster relief duties. Shimizu was left on the tarmac. It would be 10:00 a.m. the next day before he and Katsumata made it back to TEPCO headquarters.

  Even if TEPCO’s executives had been on hand, the decision to vent wasn’t solely the utility’s to make. Government approval is not required by law, but it was widely understood at TEPCO that government officials from several agencies needed to be brought into the loop. Although no one knew exactly what was taking place inside Fukushima Daiichi, everybody wanted a say. As a result, the decision-making process about venting Unit 1 came to a near standstill.

  For the nuclear power establishment, the decision to vent radioactive steam holds serious implications. Releasing radiation into the environment demonstrates unequivocally to the public that this form of generating electricity is not as clean and safe as the industry’s public reassurances and promotional campaigns proclaim.

  By late afternoon, conditions appeared serious enough that Japanese regulators decided to notify international authorities. Shortly after 4:45 p.m., NISA alerted the International Atomic Energy Agency (IAEA) in Vienna that Fukushima Daiichi had reached the state of “near accident at a nuclear power plant with no safety provisions remaining.” Under the IAEA’s seven-level scale of nuclear accidents, seven being the most serious, that constituted a level 3 event.

  Prime Minister Kan put off declaring a nuclear emergency, despite a request that he do so by the heads of both NISA and METI. The events at Fukushima Daiichi weren’t waiting, however. Just before 6:00 p.m., a work crew was sent to the fourth floor of the Unit 1 reactor building, hoping to learn more about the status of the isolation condenser. They wore no protective clothing. As they arrived at the double doors of the reactor building, their dosimeters shot off the scale, and they hurried back to the control room. This strongly indicated that the Unit 1 core was now exposed and fuel rods had already ruptured. Full-scale melting of the core would soon begin.

  But the official word from TEPCO was vague and outdated. At 5:50 p.m., the utility issued a press release announcing the “malfunction” of the diesel generators and the resulting loss of backup power more than two hours earlier. “There have been no confirmed radioactivity impact to [the] external environment,” the English-language version of the announcement said. “Further details are in the process of being confirmed.”

  Across Japan, the scope of the natural disaster that had hit the Tohoku coast was still sinking in. More than one hundred thousand members of the Self-Defense Forces, the Japanese equivalent of the National Guard, had mobilized; local disaster agencies were struggling to grasp where to focus their attention; in some communities, whole neighborhoods had simply vanished, their residents washed out to sea. In places along the coast, tsunami survivors were still trapped atop buildings where they had fled the oncoming water. It was cold and darkness was falling. The human toll was staggering. As Kan would later say, “The focus was on saving lives.”

  Japan’s prime minister Naoto Kan announces the emergency response efforts on March 11. After declaring that the nuclear power plants in the region had automatically shut down, he said, “At present, we have no reports of any radioactive materials . . . affecting the surrounding areas.” Cabinet Secretariat, Government of Japan

  That rescue effort took on an added dimension shortly after 7:00 p.m., when Kan declared the nuclear emergency requested two hours earlier. At 7:45, chief cabinet secretary Yukio Edano alerted the nation and the world that an emergency had been declared. “Let me repeat that there is no radiation leak, nor will there be a leak,” Edano said in a reassuring voice. The prime minister’s office apparently was unaware of the high readings taken earlier at the Unit 1 reactor building. That news certainly wasn’t coming from TEPCO. At 9:00 p.m., the company issued another press release warning of a possible power shortage.

  Shortly afterward, Yoshida got what he thought was a reprieve: the water gauge on Unit 1 suddenly started working, indicating the water level was still almost eight inches above the top of the fuel. (In all likelihood, the gauge was providing an inaccurate reading because it was not calibrated for extreme conditions.) Not long after receiving that bit of apparent good news, however, he got the bad news. The radiation levels inside the Unit 1 reactor building had risen so high that entry was forbidden, seriously complicating any emergency repairs. The radiation readings were positive proof that the fuel core now was exposed and most likely melting. That finding was passed on to Tokyo along with Yoshida’s alarming prediction that the Unit 2 water level and RCIC status were unknown and that the fuel there could also be uncovered soon.

  By then, authorities had ordered the emergency evacuation of those living within about a two-mile (three-kilometer) radius of the reactors. Officials o
f the towns of Okuma and Futaba dispatched sound trucks and local firefighters to go door-to-door with the announcement. Many of the residents were still reeling from the earthquake and tsunami, searching for missing loved ones, or scavenging for their possessions. They were told to leave immediately. Those living a little farther out, between two and six miles (three to ten kilometers) from the reactor, were directed to stay indoors. All they were told was that there were problems at Fukushima Daiichi.

  As people fled, the first of about a dozen power supply trucks were rumbling toward Fukushima Daiichi, dispatched that afternoon from TEPCO headquarters and from other utilities. These mobile generating units might provide the power so desperately needed. But the challenge of even getting to the plant was enormous. The drivers were forced to navigate roadways battered by the natural disaster and clogged with traffic leaving the area. By 11:00 p.m. the first trucks had made it, and workers attempted to connect the generators but had difficulty locating functioning electrical power panels. Unfortunately, some of the cables were too short and the plugs incompatible.8 False alarms of another tsunami interrupted the task, forcing workers to flee to higher ground. After twenty-four hours, only one generator was operating.

  During the first hours of the accident, as crews at Fukushima Daiichi scrambled, government and utility officials in Tokyo lacked a similar sense of urgency, a response some later attributed to a failure of leaders there to understand what was happening at the plant. They seemed to feel that there was adequate time to decide a course of action. Prime Minister Kan, however, became increasingly frustrated that the venting was not happening.

  The delay in venting was the result of a number of factors. One was the difficulty of the emergency evacuation. There was an informal agreement with the government of Fukushima Prefecture that until nearby residents were safely relocated, venting would not occur. But even if workers had tried to start the venting immediately, they would have had problems. Because the vent valves could not be operated from the control room, they had to be opened manually. It took hours to figure out where the valves were physically located and which could be opened by hand.

  By the time the evacuation was declared complete and the decision to vent was finally reached on the morning of March 12, conditions at the plant had worsened. Accessing and opening the vent valves located deep inside the dark, intensely hot, and now radioactive reactor building was a far more dangerous mission than it would have been the previous evening. This was a scenario no accident drill had covered.

  In Tokyo, Kan’s irritation over the slow flow of information—and its accuracy—was mounting. In addition to the two hundred TEPCO technical advisors on duty at company headquarters, four hundred plant personnel under Yoshida’s direction manned Fukushima Daiichi’s emergency response center. Communication between headquarters and the response center was occurring via TEPCO’s in-house videoconferencing system.

  The government, on the other hand, had no similar ability to communicate with the plant. In Tokyo, NISA obtained information from phone conversations with TEPCO. (A videoconferencing system was not set up until March 31, when the government and TEPCO created a joint response center.) Dissatisfied, Kan and his advisors asked TEPCO to assign staff members to the prime minister’s office for briefings, and Kan and his aides eventually even began calling Yoshida for answers, an action for which Kan would be later accused of micromanaging.

  Just before midnight on March 12, the plant’s emergency team was able to get a pressure reading of the Unit 1 drywell using a portable generator in the control room. The team found that it exceeded the design maximum operating pressure. At 12:49 a.m. on March 12, Yoshida decided the pressure in Unit 1 was likely so high that venting now must take place. TEPCO president Shimizu, who still hadn’t returned from his vacation, agreed with the decision at about 1:30 a.m. But TEPCO also wanted the government’s blessing. At this point, an unanswered question was whether TEPCO might have to vent Unit 2 as well as Unit 1.

  Unit 1 was the only reactor of the six that relied on isolation condensers for emergency cooling. Apparently, even the shift team assigned to Unit 1 was unfamiliar with that design. Had team members been trained in the system, they would have recognized that it was not operating and that Unit 1 had been deprived of water for hours.

  Meanwhile, the situation at Unit 2 also remained a mystery. Although the RCIC system had started up at the time of the earthquake, once DC power to the control panels was lost no one could tell whether the RCIC had continued to successfully inject water into Unit 2. Yoshida feared it hadn’t and that the top of the fuel was about to be exposed.

  A crew wearing breathing gear and protective clothing ventured into the RCIC room in the basement of the Unit 2 reactor building to determine its status. The first trip was inconclusive. A second team was dispatched. This crew said it believed the RCIC was functioning based on pressure measurements. With that news, Yoshida decided that Unit 1 warranted first priority; Unit 2 could wait. Word was sent to Tokyo. Kan and Banri Kaieda, head of the Ministry of Economy, Trade and Industry, agreed.

  TEPCO managing director Akio Komori, who had once worked at Fukushima Daiichi, joined Kaieda and the head of NISA at a joint press conference shortly after 3:00 a.m. to announce the venting. If it was meant to provide a reassuring message, it fell far short. Just before the press conference began, the three men found they had differing information about the status of the reactors. Uncertain of what was actually occurring at Units 1 and 2, they decided that Komori would announce the venting but not identify the reactor involved. When questioned by the media, he became confused. A few minutes later, government spokesman Yukio Edano took the podium to say that radiation would be released but the public should remain calm.

  Edano was getting his information from NISA, which was getting it from TEPCO. When later asked to assess the accuracy of information coming from his office to the Japanese public at this time, Kan would say that NISA officials were “choosing their words carefully,” and as a result Edano was being misled.

  The Unit 1 drywell, at twice its design pressure, was likely approaching a failure point. Venting had to happen, and it had to happen now. Engineers at the plant hurriedly calculated possible radiation exposure from the release. As the preparations continued, a worker was sent to check radiation levels at the Unit 1 reactor building. When he opened the door, he saw “white smoke” inside and quickly left without taking a reading. The smoke, whatever it was, clearly showed that something was leaking somewhere. At about 4:00 a.m., radiation levels near the plant’s main gate were measured at 0.0069 millirem (0.069 microsieverts) per hour. Twenty minutes later, they had jumped nearly tenfold to 0.059 millirem (0.59 microsieverts) per hour. The Unit 1 drywell was now venting itself.

  RADIATION AND THE BODY: DOSES, DAMAGE, AND DEBATE

  Radioactive materials emit ionizing radiation—that is, radiation energetic enough to detach electrons from atoms, turning them into charged particles (called ions). Exposure to ionizing radiation can have different effects on the human body, depending on the extent and nature of the damage it causes on the cellular level.

  The relative biological damage in the human body resulting from radiation exposure is measured in units called sieverts. The United States, unlike most of the rest of the world, uses a rem (“radiation exposure man”) as its standard measure. One sievert is equivalent to 100 rem.

  One class of radiation effects is known as “deterministic,” meaning that a certain level of exposure will almost always cause a particular outcome. Deterministic effects generally result from levels of radiation high enough to kill cells, causing widespread damage to tissues or organs. Depending on the nature of the injury, such doses range from tens to hundreds of rem delivered over a short period. The resulting injuries include burns, cataracts, thyroid nodules, hair loss, gastrointestinal distress, low blood counts, and cardiovascular disease. Recent studies have also identified statistically significant excess risks of certain circulatory diseases
at low doses. These studies suggest that the mortality from such diseases due to low-dose radiation exposure may be comparable to that from cancer.

  As the dose increases or wider areas of the body are exposed, the victim may develop an illness known as “acute radiation syndrome.” Although this sometimes can be cured with treatment, high enough doses—above several hundred rem delivered in a brief period—will almost certainly result in death within days or weeks. Following the Chernobyl accident, twenty-eight people—plant workers and firemen in close proximity to the damaged reactor—are known to have died in this manner. Death from acute radiation syndrome would be classified as an “early fatality,” occurring within a few weeks or months after exposure to a nuclear plant release.

  Deterministic effects feature a dose “threshold” below which a particular effect will not occur. This is because cells must sustain a certain amount of damage before the cell dies. In addition, a certain number of cells must be affected before enough tissue damage occurs to cause clinical symptoms.

  The other major class of radiation effects is known as “stochastic,” or random. Ionizing radiation can cause DNA damage that might produce changes in cellular behavior leading to cancer, but does not definitely cause such changes. Cancer risk does rise with increasing doses, however, because the more DNA lesions there are, the higher the chance that one of them will lead to cancer.

  It is currently estimated, based on studies of survivors of the atomic bombings at Hiroshima and Nagasaki, that a dose of ten rem delivered at once will raise an individual’s lifetime risk of fatal cancer by about 1 percent on average. This risk is higher for children and other groups of people (for instance, those with certain genetic variations) who are more sensitive to the effects of radiation than the average adult. Doses delivered over lengthy periods may be less effective at causing cancer, but there is much uncertainty about whether this is true. Because most cancers take many years, even decades, to appear after exposure to radiation, deaths due to stochastic effects are called “latent cancer fatalities.”