Fukushima in review: A complex disaster,a disastrous response

Lapses in preparedness

The Fukushima crisis revealed the dangers of building multiple nuclear reactor units close to one another. This proximity triggered the parallel, chain-reaction accidents that led to hydrogen explosions blowing the roofs off of reactor buildings and water draining from open-air spent fuel pools—a situation that was potentially more dangerous than the loss of reactor cooling itself. Because of the proximity of the reactors, Masao Yoshida, the director of the Fukushima Daiichi Nuclear Power Station at the time of the accident, was put in the position of coping simultaneously with core meltdowns at three reactors and exposed fuel pools at three units.

The first 24 hours of the accident—from the total loss of power to the station on March 11, when standard cooling stopped and damage to cores began, to the forced injection of seawater the following day—were critical. The most significant mistakes on that first day may have been the misjudgment of the state of the isolation condenser, and the 15-hour delay in venting Unit 1’s containment area to ease rising pressure.

At the outset of the accident, a Tepco worker misjudged the backup cooling situation at Unit 1. He failed to notice that the valve of the unit’s isolation condenser, or IC—a battery-powered emergency cooling system—was either fully or partially closed after the plant lost power. Steam usually spews out when the IC is activated; because there was no sign of steam, the worker hastily assumed that the IC system had lost its cooling water. For fear of the mechanical damage that could occur if the system were run without water, the worker removed the IC from service for about three hours, starting at around 6:30 p.m. on March 11.

The delays in venting reactor containments in the immediate wake of the accident may have been the result of assuming that battery-powered emergency cooling systems were in operation. ⁶ If this backup cooling system were working, there would be less buildup of pressure and hydrogen in reactor containments and less need to vent them. Beyond this, most likely the Tepco worker was distracted by a deteriorating situation at Unit 2, as well. And even after top Tepco officials made the decision to vent, there was a more than seven-hour delay, the exact causes for which remain unclear. ⁷ Whatever the reason for the delay, it resulted in the creation of more hydrogen in the Unit 1 reactor vessel—hydrogen that later was vented into the reactor building, where it exploded, blowing the building’s roof off and exposing the reactor’s spent fuel pool to the environment.

Many human errors were made at Fukushima, a point elaborated on in great detail in the interim report of the Japanese government’s Investigation Committee on the Accident at the Fukushima Nuclear Power Stations, and our investigative commission has incorporated some of these findings in its report as well. But the role of human error in the Fukushima nuclear accident was not limited to the misjudgment of any one worker, like the one who misjudged the backup cooling situation at Unit 1. The technical chief, the plant director, and the nuclear energy section of Tepco’s headquarters all failed to ascertain the true operational situation of the IC system at Unit 1.

In part, that failure may stem from Tepco’s most recent abnormal operating procedures, which were created in 1994 and do not address the possibility of a prolonged, total loss of power at a nuclear plant. When on-site workers referred to the severe accident manual, the answers they were looking for simply were not there. And those who misjudged the condition of the emergency cooling system had never actually put the system into service; they were thrown into the middle of a crisis without the benefit of training or instructions.

Tepco bears the primary responsibility for the incompetent handling of the aftermath of the disaster. Behind the failure to prepare adequately for a major accident are problems inherent in Tepco’s management structure and culture. For example, neither the chairman nor the president—Tepco’s top two managers—was at the head office between Friday, March 11, and 10 a.m. on Saturday, March 12, the most crucial period for dealing with the accident. According to the explanation given by Tepco, company Chairman Tsunehisa Katsumata was traveling in China on a business trip, and company President Masataka Shimizu was in Nara, a historical town in the western part of Japan, sightseeing with his wife, when the disaster happened. The closure of three of the main Japanese transportation arteries—the Chuo motorway, the Tomei motorway, and the Tokaido Shinkansen, or bullet train—leading back to the Tokyo area prevented Shimizu from returning by land, and a tragicomic series of miscues related to air transport kept him in western Japan until mid- morning on Saturday. ⁸ Tepco was consequently unable to make prompt organizational decisions and wound up losing the government’s trust with regard to information sharing and decision making.

But government regulators—the Nuclear and Industrial Safety Agency (NISA) and the Nuclear Safety Commission (NSC) which oversees NISA’s activities as the safety authority—share the blame for the poor response at Fukushima. NSC did not include provisions for an extended loss of power in its accident-management policy. NSC regulatory guidelines for reviewing safety designs of light water nuclear facilities specify that the potential for an extended station blackout need not be considered, as it is reasonable to expect that transmission lines will be restored or emergency power systems repaired quickly.

At Fukushima, however, transmission lines for external power sources were not restored until March 17, and the emergency power systems could not be quickly repaired. The accident was the result of an extended loss of electric power, so the regulatory guidelines— issued by NSC and executed by NISA—which stated that a station blackout need not be considered played a large and negative role in the events that transpired.

The Japanese government’s unpreparedness also played a role in exacerbating the Fukushima disaster. Perhaps the most obvious indication of this systematic breakdown in preparedness involves the nuclear emergency response headquarters, or off-site center, for the Fukushima plant. Such centers were established in the wake of a 1999 criticality accident at the Tokai nuclear fuel conversion facility that exposed hundreds of workers, responders, and residents to excess radiation levels; the centers were intended to serve as frontline headquarters for the coordination of responses to nuclear accidents. In March 2011, however, the off-site center for Fukushima, originally planned to be the base to cope with nuclear disasters, was inoperative throughout the crisis because of the destruction wreaked by the earthquake and tsunami—roads were blocked and electricity was out. Further, even without these logistical problems, the center could not be used because it was not fitted with such basic protections as air-purifying filters.

Another preparedness breakdown involves SPEEDI, the Japanese government’s System for Prediction of Environmental Emergency Dose Information. The system was touted as able to provide forecasts for the diffusion of radioactive materials during a nuclear event, but it remained largely unused during the crisis because the Nuclear Safety Commission and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) were reluctant to release predictions claiming that the simulated results were based on what several government officials interviewed by our commission called “unreliable emission source term.” Despite widespread environmental contamination by radioactive material between March 11 and March 15—the time when the central government made decisions about evacuating residents—SPEEDI data were not officially provided to top leaders in the Prime Minister’s Office until March 23. Evacuation orders were therefore issued without the benefit of SPEEDI forecasts.

In hindsight, March 15 turned out to be a crucial turning point; an early morning accident at Unit 2 led to a dramatic rise in the diffusion of radioactive materials from that site. This quashed any hope of containing the radioactivity. SPEEDI was developed in 1984 for exactly this kind of situation; the system was intended to help governments decide precisely when to evacuate residents—and from which specific areas. The failure to use SPEEDI suggests that the heavy investment in time and money to develop this system were for naught.

The trap of the absolute safety myth

Why were preparations for a nuclear accident so inadequate? One factor was a twisted myth—a belief in the “absolute safety” of nuclear power. This myth has been propagated by interest groups seeking to gain broad acceptance for nuclear power: A public relations effort on behalf of the absolute safety of nuclear power was deemed necessary to overcome the strong anti-nuclear sentiments connected to the atomic bombings of Hiroshima and Nagasaki.

Since the 1970s, disaster risk has been deliberately downplayed by what has been called Japan’s nuclear mura (“village” or “community”)—that is, nuclear advocates in industry, government, and academia, along with local leaders hoping to have nuclear power plants built in their municipalities. The mura has feared that if the risks related to nuclear energy were publicly acknowledged, citizens would demand that plants be shut down until the risks were removed. Japan’s nuclear community has also feared that preparation for a nuclear accident would in itself become a source of anxiety for people living near the plants.

One example of the power of the safety myth involves disaster drills. In 2010, the Niigata Prefecture, where the 2007 Chuetsu offshore earthquake temporarily shut down the Kashiwazaki-Kariwa Nuclear Power Plant, made plans to conduct a joint earthquake and nuclear disaster drill. But NISA advised that a nuclear accident drill premised on an earthquake would cause “unnecessary anxiety and misunderstanding” (Committee’s interview with Hiroyuki Fukano, NISA director, February 7, 2012) among residents. The prefecture instead conducted a joint drill premised on—heavy snow.

The word used then to describe risks that would cause unnecessary public anxiety and misunderstanding was “unanticipated.” Significantly, TEPCO has been using this very word to describe the height of the March 11 tsunami that cut off primary and backup power to Fukushima Daiichi.

But research on the Jogan tsunami of 869 AD has shown that such heights should not be considered “unanticipated” along the part of the Japanese coast that includes the Fukushima nuclear complex (Minoura et al., 2001). In fact, even Tepco’s own nuclear energy division understood that there was a risk of large tsunamis at Fukushima (Sakai et al., 2006). However these probabilities were ultimately dismissed through the internal discussion of the division on the grounds that they were “academic.” Regulatory authorities, too, had encouraged the company to incorporate new findings regarding tsunami risks into its safety plans, but such measures were not made mandatory.

So the March 11 tsunami was not unexpected. Yet Tepco perpetuated the safety myth and did very little to back-fit the existing nuclear systems to incorporate the latest scientific findings and technological innovations for improved safety. Making such changes, the nuclear community felt, would be an admission that existing safety precautions and regulations were insufficient and that nuclear plants did not possess “absolute safety.” In this way, power companies found themselves caught in their own trap.

A problem of governance?

Japan’s nuclear safety regulatory regime has been under the dual jurisdiction of the Ministry of Economy, Trade, and Industry (METI), which promotes nuclear energy use, ⁹ and the Science and Technology Agency (now part of MEXT). Supposedly, NISA enforces and NSC double-checks the regulatory review—however, both agencies use the same guidelines in their review. It should come as no surprise that the problems with this clunky structure have been criticized by the international community for quite some time.

In June 2007, for example, the International Atomic Energy Agency (IAEA) demanded clarification of NISA’s regulatory role and the role to be played by the NSC, especially in the development of safety-assessment guidelines (IAEA, 2007). In response, the NSC issued a chairman’s statement in 2008 dismissing the recommendations and noting that Japan has been praised highly for regulations that are, on the whole, outstanding in the context of international standards and that are functioning effectively to ensure nuclear safety (Nuclear Safety Commission, 2008).

The NSC attitude typifies Japan’s decades-long approach to nuclear safety. Having great confidence in its technical capabilities, the Japanese nuclear community did not take the need for improvement of safety regulations seriously before Fukushima; there was little accountability, given the unclear jurisdictions, complicated turf wars, and mountains of red tape that connected and divided NISA, METI, NSC, AEC, MEXT, and other entities with nuclear responsibilities. This is not to mention the sweetheart relationships and revolving door that connected the regulatory bodies and electric companies, academics, and other stakeholders in the nuclear community. The country’s regulatory regime and culture of safety assurance are out of sync with global standards, evolving as they did in isolation and within the complex vacuum of Japan’s safety regulation agencies.

As a regulatory agency, NISA lacked the philosophy, capacity, and personnel to properly fulfill its role, and it has consequently failed to train true safety regulation professionals. Top NISA officials were unable to answer the questions posed by members of the crisis response team at the Prime Minister’s Office and offered no proposals to bring the accident under control. Instead NISA simply conveyed to Tepco the government’s requests for updates on conditions at Fukushima Daiichi, which was pointed out almost unanimously by the politicians we interviewed.

The response to the Fukushima nuclear accident revealed a crisis in the administration and enforcement of safety regulations. To correct the problem, bureaucratic sectionalism and duplicate safety regulations must be eliminated, and the government needs to create a nuclear regulatory authority that is not part of a ministry that promotes nuclear energy use. Toward this end, the government is considering the creation of a new nuclear safety agency that would replace NISA and NSC and be positioned as an external organ of the Environment Ministry. Determining whether it can truly be independent or better able to enforce safety regulations efficiently and appropriately will require further review.

Even if a new agency is created, a serious challenge will remain: staffing the new entity with real experts. The March 2011 crisis was a painful reminder that few people are qualified and capable of grappling with an actual nuclear disaster. Prime Minister Kan was so frustrated with the ineptitude of top NISA officials that he called in six outside experts to advise him on technical matters in the wake of the Fukushima accidents. Why were there no top officials at NISA with professional knowledge? The answer to that question lies primarily in Japan’s bureaucratic and organizational culture, under which NISA officials are regularly reshuffled—along with members of related agencies—after serving in a position for only a few years. Effective regulations will not be formulated and enforced unless top officials in a regulatory agency are experts in their respective fields and serve longer terms.

Taking responsibility: The public and private sector clash

Although the peaceful use of nuclear power is part of the Japanese government’s energy policy, nuclear power generation is undertaken within the private sector. This approach is not unique and is the model used in other countries with nuclear power, including the United States. But once Tepco’s weaknesses in crisis management, decision making, and governance were laid bare, many wondered whether the company was really capable of operating a nuclear power plant. Regardless of the structure Japan establishes to conduct nuclear power generation in the years ahead, there is a clear need for the government to take responsibility for nuclear power management and safety regulation. What happened in the early days of the Fukushima catastrophe illustrates that need.

When, on March 14, Tepco indicated that it might pull all its workers from the Fukushima plant and leave the failing facilities abandoned, Prime Minister Kan and other top political officials stormed into Tepco’s headquarters in Tokyo and demanded that a joint response headquarters be established. Clearly, the government bears the ultimate responsibility for bringing a nuclear accident under control.

Unfortunately, this wasn’t the government for the job. According to first-hand accounts by five top officials in the Prime Minister’s Office, the prime minister raced into Tepco’s headquarters at 5:35 a.m. on March 15 and told more than 200 workers in its operation room that abandoning the reactors and spent fuel pools would have devastating effects over several months, creating 10 to 20 sources of radiation, each releasing two to three times the contamination discharged at Chernobyl. It didn’t matter how much it cost to contain the disaster, the prime minister said: Withdrawal was out of the question when Japan’s survival was at stake. Kan mentioned that the United States or Russia would not have any choice but to intervene in the Japanese government’s effort to control the nuclear disaster if Tepco did nothing. He went on to say that Tepco was not allowed to accept defeat and that the company bore the ultimate responsibility. The workers, he said, should put their lives on the line to salvage the situation. Moreover, he concluded, if Tepco did withdraw, the company would certainly be bankrupted by the accident. When our commission asked Kan during an interview on January 14, 2012, whether he really asked Tepco employees to put their lives on the line, Kan didn’t directly answer the question, noting that there was no law that could prevent Tepco from withdrawing from Fukushima. But in his speech at Tepco on March 15, 2011, he appealed to the workers’ sense of duty, asking them to remain on-site, protect the plant, and, thus, protect Japan as a nation.

Even in a crisis, a Japanese leader cannot order private-sector employees to die. Similarly, a senior METI official responsible for nuclear energy told our commission that the ministry has no authority to order its officials to remain on-site if doing so would cost them their lives. Because of these constraints, as a final recourse, the government used members of the Self-Defense Forces (SDF)—who must obey orders—to save Fukushima. SDF personnel directed the efforts to inject water into the reactors and fuel pools despite rising radiation levels. They dutifully performed their mission.

Crisis management and leadership

In assessing Fukushima one year on, the deficiencies in crisis management must be noted, as should the magnitude of the limitations under which the prime minister and other top leaders were operating. The key to effective crisis management is the speed with which the bureaucratic machinery can be shifted to operate in emergency mode. In general, well-run bureaucracies place the highest priority on compliance with the law, fairness, efficiency, and bottom-up decision making. In times of crisis, however, flexibility, improvisation, clear identification of priorities, and top-down decision making become more important.

Within the first days of the disaster, Fukushima Daiichi Director Masao Yoshida ordered that seawater be injected to cool Unit 1. But a hydrogen explosion, perhaps due to a much-delayed venting of the containment vessel, hampered the operation. Debris from the blast obstructed access to water lines, so workers could not make the necessary repairs, and some instruments and machines were damaged at the site, further degrading the working environment.

During a discussion of the seawater injections, Kan asked about the possibility of re-criticality of the damaged fuel in the Unit 1 reactor; Haruki Madarame, chairman of the Nuclear Safety Commission, replied that after a seawater injection, such a possibility “could not be denied.” Kan was not convinced by the argument, and the discussion on seawater injection became tangled. It took two more hours until they finally decided to start the injections at Unit 1. This delay had an enormous and unusual impact on Tepco.

The company’s chief liaison with the government, Ichiro Takekuro, notified the head office that further water injections should be avoided until the government decided on a course of action. Tepco President Shimizu relayed this information to Fukushima Daiichi Director Yoshida, who insisted on restarting injections as soon as possible, leading to a heightening of tensions between the head office and the on-site staff.

During a teleconference, Yoshida called the employee in charge of the seawater injections to his side and whispered in his ear—so the microphone for the teleconference with the head office would not pick up his voice—that though he would now order a halt to the seawater injections, the employee should disregard the order and continue. Thereupon, Yoshida loudly declared to all teleconference participants that water injections would be interrupted.

Yoshida’s kabuki play successfully helped Tepco avoid further confrontations with the government, while ensuring that the cooling of the reactors would continue; at this point, the company’s Fukushima Daiichi plant team was working independently of their headquarters. That the on-site director in Fukushima needed to go to such lengths to avoid a further deterioration of the nuclear crisis shows the extent to which relations between the Prime Minister’s Office and Tepco and communication between Tepco’s headquarters and the company’s on-site managers had broken down.

Because all the ordinary means of filling the pools had been disabled by the earthquake and tsunami, the only option was to mobilize the police, the SDF, and the fire department—agencies that are usually first on the scene during emergencies. None of these groups, however, had ever been called upon to spray water into fuel pools, and they had never received training for such a dangerous and difficult task. Unsurprisingly, information sharing did not go smoothly: SDF did not have a site map of the nuclear plant because Tepco employees feared this would violate security regulations.

But the biggest problem with the government’s crisis management was probably the amateurish level of its crisis communications. To be sure, information was, for the most part, insufficient, and there was little time to assess its reliability before dissemination. Still, the government’s crisis communication efforts often were abysmal.

When, at a press conference held on the second day of the crisis, NISA Deputy Director Koichiro Nakamura acknowledged the possibility of a core meltdown, Chief Cabinet Secretary Yukio Edano demanded that reactor updates only be communicated with approval from the Prime Minister’s Office. Nakamura was dismissed from his post later that evening, and his assessment of a potential meltdown was rejected. The Prime Minister’s Office then functioned as a micromanager, only further complicating the process. Kan personally visited the plant, circumventing the NISA director and directly contacting lower-ranking NISA managers with questions about minor technical details.

Moreover, Kan’s often-abrasive comments and questions could seem like cross-examinations; they made many officials and advisers shrink under his direction. In a December 17, 2011, interview with our commission, NSC Chairman Madarame said the prime minister became overly excited after a March 14 hydrogen explosion at Unit 3 that led to the injury of some SDF soldiers on-site. For a couple of days, Kan and other officials were driven by a fear that public disclosures of radiation levels would cause widespread panic. This gave the impression that the political leadership had fallen into a sort of “elite panic.” ¹⁰

Despite the government’s efforts to downplay the seriousness of the situation, this hydrogen explosion made the severity of the crisis clear, and public confidence in the government rapidly declined. The majority of the general population had no idea of the meaning behind the reported radiation levels. There was no yardstick against which to judge whether or not the levels were dangerous. The government made no effective effort to educate or soothe the public in this regard. For example, in many evacuation zones, men wearing white protective uniforms would arrive at a house and order the residents to evacuate; they often did this without explaining the reasons for the evacuation.

Resilience

Though much of the response to Fukushima was an utter failure, the reality is that it actually could have been worse had lessons not been learned from previous crises. Tepco had certainly failed to put in place adequate tsunami countermeasures at Fukushima Daiichi, such as reinforcing the embankment and protective system for cooling system water pumps—but some measures taken after the earthquakes at other nuclear plants such as Onagawa, owned by Tohoku Electric Power Company, and Tokai, owned by Japan Atomic Power Company, paid off, and these facilities were able to escape the total loss of power. After the 2007 earthquake in Chuetsu, a seismically strengthened building was built at Fukushima Daiichi; this building, though it did receive some radioactive contamination in 2011, was unharmed by the earthquake and tsunami. Had it not been available for use as an emergency headquarters, the accident could not have been brought under control.

Quality crisis management includes the drawing of lessons from disaster. This process involves studies into the causes of accidents and responses to those causes, the charting of new goals for minimizing the risk of disaster, and the building of a national consensus around realizing those goals. Ultimately, the final outcome of studies of Fukushima Daiichi should be an intense effort to build up the resilience of the country, its organizations, and its people, so future disaster can be averted or responded to effectively.

When it comes to nuclear disasters, no two are exactly the same. So legislation and manuals do little to add clarity or direction to the situation. At Fukushima Daiichi, the problems were not with the law or the manual, but with the humans who formulated the “anticipated” risks that fell in line with corporate and political will—but did not represent the actual risks the nuclear plant faced and posed.

A personal aide to Prime Minister Kan confided to the independent investigation that he felt the Japanese have all the luck, or must have been blessed to be able to live through this ordeal, given its enormity and the truly existential threat it posed. In light of the worst-case scenario submitted to Kan, the remark is especially poignant. After all, just as no two disasters are identical, no two measures of luck will ever be exactly the same.