Fukushima: The Story of a Nuclear Disaster Page 5

  Increases in cancer risk associated with low doses of radiation are small compared to the background level of cancer in humans. Thus, in epidemiological studies it is hard to detect a direct cause-and-effect relationship between radiation in the low-dose range (generally below about five rem) and cancer. However, there is broad scientific consensus that all radiation exposure, no matter how small, produces some increased risk of cancer. It is biologically plausible that even a single particle of ionizing radiation could cause enough damage in a cell to induce cancer.

  Nonetheless, a small and vocal group, including some scientists, believes that the absence of observable evidence for an increase in cancer risk at low doses implies that ionizing radiation is harmless below some threshold. The logic is that at these doses, the rate of damage is so low that DNA repair mechanisms can combat it successfully. Some even believe in the theory of “hormesis,” which holds that low doses of radiation are actually beneficial and can stimulate the immune system like a vaccine.

  Although these arguments have been reviewed and largely rejected by authoritative scientific bodies, such as the National Academy of Sciences’ Biological Effects of Ionizing Radiation (BEIR) VII Committee, advocates of a threshold continue to cite them as justification for the belief that the harmful effects of radiation in general, and nuclear energy in particular, are exaggerated.

  At the extremes of pressure and temperature that the drywell was now experiencing, the bolts and seals used to make it leak-tight were giving way, allowing radioactive steam and hydrogen gas to escape directly into the reactor building. This was the last barrier preventing the release of radioactivity into the atmosphere. At least the pressure inside the containment was dropping, although not enough to obviate the need for operator-controlled venting. The plant operators needed to reduce the containment pressure by venting through the torus instead of the drywell; then some of the radioactivity might be filtered as the gases passed through water. Otherwise, unfiltered releases directly from the drywell would continue.

  To vent the Unit 1 containment, workers needed to manually open two valves in different locations. The first was on the second floor of the reactor building. Without electric power, workers would have to open it with a hand crank—provided they could reach it. The second valve was in the basement torus room. Ordinarily the basement valve required compressed air to operate. By poring over plant drawings, however, workers had located a wheel handle on the valve that they could use to open it manually.

  They now mapped out the route they would take to get to it. Their path would be through the dark and hot reactor building. In the past few hours, the plant site had been rattled by twenty-one aftershocks, raising fears of another tsunami. What if another wall of water were to sweep in and surround them all?

  But the more immediate threat was invisible: the rising radiation level. Crews now had to wear protective gear and breathing equipment if they ventured out of the emergency response center. Soon, radiation levels rose even inside the Units 1 and 2 control room, and operators there had to wear full face masks and protective clothing. Most of the control room team moved to the Unit 2 side of the room and crouched down on the floor, where the radiation was a little lower, to do their work.

  Workers wearing protective clothing and respirators inside the Unit 2 control room on March 26, 2011, after power is restored at the plant. Tokyo Electric Power Company

  Meanwhile, workers continued their efforts to force water into the Unit 1 core by any means available. The diesel-driven fire pump idled for hours, but it remained useless because operators could not depressurize the reactor vessel and the pump was not powerful enough to inject any water into it. Finally, the pump shut down and could not be restarted.

  And then, an apparent miracle happened, but it was a mixed blessing. At around 2:00 a.m., operators were able to recover some instrumentation using batteries and discovered that the reactor vessel pressure had dropped significantly all by itself. On one hand, the pressure, although still high, was now low enough to give the fire engines on-site a chance of getting water to the core. On the other hand, the depressurization was an ominous sign that the reactor vessel had sprung a leak somewhere—although it is not clear that anyone appreciated that at the time.

  At about 4:00 a.m., workers finally managed to connect a fire hose to a portal on the Unit 1 turbine building, providing a pathway for one of the fire engines to pump freshwater into the reactor core. It took almost two more hours to establish a consistent flow rate, but for the first time in nearly fifteen hours, water was reaching the core. However, it was too little, too late. The fire engine pump pressure was still too low to force much water into the vessel. And in any event, it is likely that by then the core had already melted through the bottom of the reactor vessel and dropped to the containment floor. If so, much of the water being injected into the core was probably flowing out into the containment.

  The whole exercise had taken much longer than anyone had anticipated in part because no plant worker knew how to operate the fire engines maintained for emergencies at the reactor site. A contractor had to be convinced to help with the arduous task. With rising radiation levels, ongoing aftershocks, and the threat of another tsunami, the contractor was reluctant to agree.

  In Tokyo, at both the utility and the prime minister’s office, everyone was asking the same questions: When was the venting going to begin? Why was it taking so long? When Prime Minister Kan asked the TEPCO official assigned to his office to explain the delay, the only answer he got was: “I don’t know the reason.”

  Shortly after 6:00 a.m., Kan decided to find out for himself. In the midst of the growing crisis at the plant, Yoshida was informed that the prime minister was en route via helicopter. Before Kan left Tokyo, he ordered authorities to widen the evacuation zone around the plant from about two miles (three kilometers) to six miles (ten kilometers). Without careful coordination, however, that would put a lot of people on the roads when the venting finally did take place. Nonetheless, while Kan was airborne, METI minister Banri Kaieda ordered TEPCO to vent.

  At 7:11 a.m., Kan, accompanied by Haruki Madarame, the chairman of the Nuclear Safety Commission, landed at Fukushima Daiichi. Yoshida explained the difficulties to Kan, who calmed a bit when he discovered that he and Yoshida had attended the same college, the Tokyo Institute of Technology. Kan repeated the order: vent. Yoshida promised that would happen by about 9:00 a.m. Fifty minutes after he arrived, Kan headed back to Tokyo and Yoshida back to managing the disaster. Although Kan’s dramatic fly-in garnered much publicity (and eventually was portrayed by his critics as dangerous meddling), Yoshida remained the man calling the shots.

  Three two-person teams suited up with protective clothing. They were about to enter a dark, highly radioactive reactor building that they might have to flee at any moment because of the intermittent earthquake tremors. They knew they would have just a few moments to accomplish their task before they would reach their allowable radiation dose and have to retreat. The makeup of the teams reflected the radiation exposure risks everybody recognized: young employees were excluded from the mission.

  At 9:02 a.m., the plant was notified that the evacuation of residents was complete and that the venting could now begin. (The information about the evacuation turned out to be incorrect.) The first team entered the second floor of the Unit 1 reactor building, flashlights providing the only illumination as team members searched for and located the hand crank. With tools about the size of those used to change a car tire, they cranked the vent valve open a quarter of the way before they hastily left. In the ten or so minutes they had been inside, each man had received a radiation dose of two and a half rem (twenty-five millisieverts), one-fourth of the total dose they were permitted to receive under emergency conditions for an entire year. When members of the second team entered the torus room in the basement of Unit 1 to open the second valve, they found dose rates so high that they were unable to reach it, and they fled. Even so, one of the oper
ators exceeded the emergency dose limit of ten rem (one hundred millisieverts). Entry efforts were then abandoned.

  A new plan was devised; it, too, proved problematic. There was another, larger valve that might be opened from the control room with battery power and a compressed air supply. But the available compressor didn’t work without electricity, and nobody had a portable unit. Once again, contractors’ offices on and off the plant site were searched. More time elapsed.

  Workers then attempted to open the small air-operated valve in the torus room from the control room in the hope that there was still enough residual air in the valve to make it work. For venting to be successful, this valve would have to remain open long enough to create pressure in the vent sufficient to burst a rupture disk, which served as the last barrier between the radioactive gas and the environment. When radiation levels spiked at the main gate and at several monitoring posts at about 10:40 a.m., operators thought their venting attempt had worked and the disc had ruptured. However, soon afterward radiation levels began to fall and it was no longer clear that there had been significant venting.

  Finally, shortly after noon, a portable compressor was located and jury-rigged so it could be connected to the system that normally supplied air to plant equipment. Together with temporary DC power, this allowed operators to open the large air-operated valve from the relative safety of the control room. At 2:00 p.m., the compressor was started and the large valve opened. Pressure dropped inside the containment, and presumably inside the reactor vessel as well, allowing operators to inject water at a higher rate.

  Another sign that the operators might have succeeded came when NHK cameras trained on the plant from a distance captured white smoke emerging from the exhaust stack shared by Units 1 and 2 and rising high above the sky-blue reactor building. Yoshida and the exhausted crew in the emergency response center thought they had finally caught a break. At 3:18 p.m., he notified TEPCO in Tokyo of the venting.

  Captured by an NHK camera positioned about twenty miles away, the explosion inside Unit 1 at 3:36 p.m. on March 12 destroys the roof of the reactor building and blows out a panel in the adjacent Unit 2 reactor building. NHK

  Yoshida had other progress to report as well. Recognizing that freshwater supplies were limited, he had ordered workers to come up with a method for utilizing seawater to cool the reactors. After freshwater supplies were exhausted shortly before 3:00 p.m., he gave the order to prepare to use the Pacific. It was a complicated endeavor—workers had to position three fire engines with interconnecting hoses to pump seawater transferred into a pit—but by 3:30 p.m., the lines to inject seawater into the Unit 1 reactor were almost in place. A steady supply of vital cooling water for the core was now within reach.

  And then, at 3:36 p.m.—almost exactly twenty-four hours since the tsunami had roared in and flooded the plant—a powerful explosion ripped through the Unit 1 reactor building, blasting off the roof and sending debris everywhere. The impact blew out a panel in the neighboring Unit 2 reactor building. Inside the emergency response center, stunned workers watched the explosion on television, not sure exactly what had blown up.

  2

  MARCH 12, 2011: “THIS MAY GET REALLY UGLY . . .”

  At 9:46 a.m. on March 11 local time, the U.S. Nuclear Regulatory Commission officially entered what it calls Monitoring Mode, a heightened state of readiness to respond when a nuclear incident is unfolding. It had been exactly nine hours since the earthquake tremors were first detected at Fukushima Daiichi.

  NRC personnel with a wide range of expertise assembled in the agency’s Operations Center, a low-ceilinged room crowded with desks and computer monitors. The NRC is currently housed in three modern mid- and high-rise office buildings in Rockville, Maryland, just outside Washington, DC. The headquarters complex is often simply called White Flint, after a nearby Metro stop.

  Overnight, news media had begun carrying stories about the natural disaster in Japan, where local time was fourteen hours ahead of Washington. Now, details about mounting problems at the Fukushima Daiichi reactors vied with accounts about the devastation from the earthquake and tsunami. The NRC staff was scrambling to learn more.

  Earlier that morning, the National Oceanic and Atmospheric Administration had issued a tsunami warning for the U.S. Pacific coast, triggering concern about a possible threat to the Diablo Canyon Power Station in California and other coastal nuclear facilities. (When the wave did arrive at about 8:30 a.m. local time, it amounted to a surge no larger than the normal tides; U.S. nuclear operations were not affected.) The information available from Japan at that time, however, convinced NRC officials that the disaster there had now risen far beyond any West Coast consequences. Responsibility shifted from the NRC’s western regional office to headquarters. Ultimately more than four hundred staff members at White Flint and around the country would be drawn in.

  The Operations Center at the Nuclear Regulatory Commission (NRC) headquarters outside Washington, D.C. More than four hundred NRC staff members nationwide were involved in analyzing the accident at Fukushima Daiichi. U.S. Nuclear Regulatory Commission

  Details were arriving from a variety of sources; most of the news was secondhand. Japan’s Asahi Shimbun was reporting that a state of emergency had been declared and evacuations were planned. CNN reported that radiation levels were increasing. On the other hand, Reuters quoted the World Nuclear Association in London, an industry trade group, as stating that the situation in Japan was “under control.”

  Among NRC staff members across the agency, e-mails were flying furiously as they combed the Web and TV networks for information, sharing their finds with each other and struggling to put the sometimes conflicting pieces together. Soon, however, those pieces appeared to add up to an ominous accident, as their e-mails reveal:

  9:21 a.m.: “We are safe and lucky this time.”

  9:42 a.m.: “[M]y understanding is that this is a Station Blackout and if they don’t get some kind of power back it’s only a matter of time before core damage. This is a really big deal.”

  More details were shared—and analyzed from a distance of nearly seven thousand miles. Within minutes of the early morning press briefings in Tokyo announcing plans to vent radiation, the implications of that decision were obvious at White Flint.

  3:07 p.m.: “This may get really ugly in the next few days.”

  Even as they struggled to comprehend the toll on human lives and property of the natural catastrophe that had struck Japan, the men and women of the NRC knew that the disaster unfolding inside the reactors half a world away had implications for White Flint. The fate of Fukushima Daiichi could dramatically affect the commission’s own future, which is closely tied to the fortunes of the nuclear industry.

  Among the messages crisscrossing the NRC as the crisis deepened was one that put this concern in a nutshell:

  “Is this safety bad or economics bad?” asked an e-mail sent to Brian Wagner of the NRC’s Office of New Reactors.

  “Both,” Wagner replied.

  Like many regulatory agencies, the NRC occupies uneasy ground between the need to guard public safety and the pressure from the industry it regulates to get off its back. When push comes to shove in that balancing act, the nuclear industry knows it can count on a sympathetic hearing in Congress; with millions of customers, the nation’s nuclear utilities are an influential lobbying group.

  Over the years since its establishment in 1974, the NRC has been accused by many critics of favoring the industry’s point of view when it comes to adopting higher safety standards. To the critics, when the question was “safety” versus “economics,” it seemed that economics often won. The NRC’s supporters responded that nuclear power generation was already so safe that tighter requirements were rarely worth their cost to industry.

  The crisis now beginning in Japan plainly had the potential to undercut the safety argument. What the NRC staffers could not yet know was how clearly Fukushima Daiichi would demonstrate the dangers that arise when r
egulators become too close to the industry they oversee.

  Japan is not the United States; the relationships between government and business in the two nations are as different as other aspects of their cultures. But the history of nuclear power in Japan, and the incestuous practices that fostered it, do provide dramatic evidence that giving the nuclear industry the benefit of the doubt can lead to unimaginably dire consequences.

  • • •

  At the moment an NRC staffer speculated things “may get really ugly,” it was 5:07 a.m. March 12 in Japan. Inside the Seismic Isolation Building at Fukushima Daiichi, Masao Yoshida was struggling to figure out how to vent the Unit 1 reactor. Prime Minister Kan, frustrated with a lack of information, was about to head to the plant himself. And thousands of Japanese embarked on the first of what would be a torturous series of migrations as they evacuated from a two-mile radius around the plant.

  This wasn’t the first time Japanese citizens had had to flee a runaway nuclear accident. Nor was it the first time they had had reason to question the response of nuclear authorities.

  In September 1999, inside a nondescript industrial building in Tokaimura, a village about seventy miles northeast of Tokyo, three employees of the JCO Company were preparing a small batch of 18.8-percent-enriched uranium fuel—containing a far higher concentration of uranium-235 than the typical fuel used in power reactors—for use in an experimental fast neutron reactor. The workers were untrained in handling material of this enrichment level and were preparing the fuel in stainless steel buckets. In a hurry, they poured the solution into a tank, bypassing safety controls. The mixture went critical, initiating a self-sustaining chain reaction that cycled on and off for hours, periodically emitting high levels of gamma and neutron radiation. The industrial building provided little radiation shielding because under normal conditions there was no need for it.