MARCH 15 THROUGH 18, 2011: “IT’S GOING TO GET WORSE …” - Fukushima: The Story of a Nuclear Disaster (2015)

Fukushima: The Story of a Nuclear Disaster (2015)

4

MARCH 15 THROUGH 18, 2011: “IT’S GOING TO GET WORSE …”

Jim Trapp and Tony Ulses were back on the phone with their colleagues at White Flint before dawn East Coast time on March 15. The original plan for the pair to spell each other off until reinforcements arrived had fallen apart; there was just too much going on. Both were exhausted; it had been a roller-coaster day.

At first, it had seemed that the flood of bad news from Fukushima Daiichi was slowing. The latest pressure readings indicated that the core of the Unit 2 reactor had not breached the reactor vessel, and operators were able to reestablish water injection. Other pressure readings contradicted the assumption that the mysterious noise heard at Unit 2 had been a rupture of the torus, although the data were inconsistent, suggesting an instrument failure. Units 1 and 3 were stable, with seawater injections proceeding smoothly.

The details even elicited a bit of optimism from Brian McDermott a short time later when he briefed his boss, Gregory Jaczko. “[L]ast night we thought we had a big problem,” he told Jaczko, referring to Unit 2. “And this morning, it, it suggests we have less of a problem.”

That assessment was short-lived.

Overnight, concern about the status of Unit 2 had led the NRC’s Protective Measures Team to run computer simulations for that unit based on worst-case assumptions: a fully molten core and a ruptured containment. The simulations projected levels of radiation exposure based on the prevailing winds blowing toward Tokyo. Readings picked up at the Yokosuka naval base and elsewhere seemed to suggest that the release was worse than everyone had thought.

Because the NRC’s RASCAL program can only estimate radiation doses within fifty miles of a release site, it was of little use in interpreting the data collected at Yokosuka, about 190 miles from the plant. Even so, the findings were passed along to the State Department and the U.S. Embassy in Japan. The calculations indicated that, if winds continued to blow steadily in one direction, then evacuation would be warranted for everyone up to fifty miles downwind of the plant. This was based on the U.S. Environmental Protection Agency’s (EPA) protective action guides, or PAGs. (According to the PAGs, members of the public should be evacuated from any area where they could receive more than one rem, or ten millisieverts, of radiation exposure in a four-day period.) Based on the RASCAL estimate, the NRC was recommending to State that U.S. citizens evacuate if they were within fifty miles of the plant, four times the distance the Japanese were recommending. Approximately three hundred Americans resided inside this fifty-mile zone, along with about 2 million Japanese.

From a health standpoint, it seemed a prudent call. Politically, however, a recommendation so at odds with what Japan was telling its citizens was sure to upset Tokyo. The State Department was in an awkward position diplomatically, hoping the NRC might find a way to support the twelve-mile evacuation with a press release—even suggesting the language it wanted the NRC to use, reaffirming the Japanese evacuation recommendations. The NRC’s experts stuck by their guns: ask Japan to tell us why we’re wrong, they countered.

For the NRC, a fifty-mile evacuation advisory for Japan could also have political ramifications at home. The commission currently requires that emergency evacuation plans be developed only for the area within ten miles of a nuclear reactor. Safely evacuating an area twenty-five times as large would be difficult, if not impossible, at many reactor sites without detailed advance preparation—and plenty of time.

While debate continued about the status of Fukushima Daiichi Unit 2 and its implications, Trapp and Ulses had more bad news to report: the crisis at the Unit 4 spent fuel pool. It appeared, they said, that the situation could be just as dangerous—and baffling—as that in Unit 2, if not more so.

The hydrogen explosion at Unit 4 had occurred almost simultaneously with the suspected explosion in Unit 2 (about 6:00 a.m. March 15, Japan time), but the NRC heard nothing about a problem at Unit 4 until a couple of hours after the report of a possible containment breach at Unit 2. (Masao Yoshida at the plant also had been late in learning. He had not known of the explosion for almost two hours—until his workers made it back to the Seismic Isolation Building.) The Unit 4 reactor, out of service with its fuel in the pool, had not even been on anyone’s radar. In fact, only a few hours before the Unit 4 explosion, Marty Virgilio had reported in a status briefing that there were no concerns about any of the spent fuel pools, although given the extended loss of power, the team needed to “keep an eye on” them. Even after information about the Unit 4 crisis began trickling in, Unit 2 remained at the top of the priority list—but not for long.

As the NRC staffers pieced together data from a variety of sources, a disturbing picture began to emerge that challenged the team’s complacency toward the spent fuel pools. The massive blast that had blown off the Unit 4 reactor building’s roof and damaged walls there looked like yet another hydrogen explosion. But where had the hydrogen come from? The most obvious source would be a chemical reaction between steam and overheating fuel rods in the spent fuel pool. But that would mean that the water level had dropped to expose the fuel much more quickly than anyone anticipated.

The spent fuel pool was designed to hold 1,300 to 1,400 tons of water, approximately half the volume of an Olympic-sized swimming pool. If the pool lost cooling, it would have taken a couple of days for the water to reach the boiling point. When the water began boiling away, the pool would lose around one hundred tons of it per day. About one thousand tons of water would have to boil off before the fuel would be exposed. By a back-of-the-envelope calculation, it should take well over a week to get to that point. At the time of the explosion, the cooling pumps for the Unit 4 pool had lacked power for less than four days. If the spent fuel had become uncovered in such a short time, then more than eight hundred tons of water must have been lost in some way other than boiling. This could have been a result of shaking during the earthquake. (One worker who was on the Unit 4 refueling floor when the quake struck later reported seeing waves in the pool and being drenched as water sloshed out.) There was another possibility: a leak or crack in the pool itself, caused by the earthquake or the explosion in Unit 3.

No one could verify what was actually going on in Unit 4, or in any of the other spent fuel pools for that matter, because the gauges that measured the water level and temperature in the pools were useless without electrical power.

The worst case seemed to be confirmed when the NRC started to receive reports that a few hours after the blast a fire was observed burning in the Unit 4 reactor building. Initial fears were that this showed the zirconium cladding on the fuel had indeed ignited, making it possible that the fuel itself was melting and releasing a massive amount of radioactive cesium. With the roof and walls of the Unit 4 reactor building blown apart, the radiation would go straight into the atmosphere. No one could get close enough to investigate; however, the fire burned itself out shortly before noon. Its source was subsequently identified by NISA as lubricating oil used in a generator on a level below the fuel pool. The quake might have damaged piping or a storage tank, and a spark apparently set the oil ablaze. If the zirconium cladding had been burning, the fire would have lasted for days.

Radiation measurements between the Units 3 and 4 reactor buildings had soared to forty rem per hour, making even brief forays near the buildings extremely dangerous. They also seemed to indicate that, even if there had not been a zirconium fire in the Unit 4 spent fuel pool, something bad had happened there: perhaps pieces of spent fuel had been dispersed by the explosion, or there was no water left in the pool to shield the fuel.

Meanwhile, in the hours after the mysterious event at Unit 2 and the explosion at Unit 4, radiation at the plant gate had spiked to its highest level since the accident began, and an ominous cloud of “white smoke” was seen drifting from Unit 2. The prevailing winds continued blowing inland, toward populated areas. News that the wind direction was rotating clockwise, away from Tokyo but toward smaller cities and agricultural areas northwest of the plant, was small consolation. Since it was no longer clear that there had been a containment rupture at Unit 2, reports of white smoke notwithstanding, it was looking more plausible that Unit 4 was the source of the increased emissions.

Now the NRC had a new challenge: trying to understand what was going on with the Unit 4 spent fuel pool. The question of whether the pool still held any water had serious implications for estimating the amount of radiation that potentially could be released. This would soon become a thorny issue for the NRC team.

There was no doubt at White Flint that water had to be added to the Unit 4 pool immediately. That meant plant operators would have to figure out a way to deliver tons of water per hour into a spent fuel pool five stories off the ground, filled with rubble, and emitting levels of radiation that could be lethal within minutes. In the NRC’s view, getting water into the spent fuel pools should become the number one priority at Fukushima Daiichi. However, the Japanese did not yet see it that way.

WHERE FUKUSHIMA DAIICHI GOT THINGS RIGHT

Fukushima Daiichi focused world attention on spent fuel pools and their associated risks. Fukushima also demonstrated a safer means of managing the spent fuel risk: dry storage.

There were 408 spent fuel assemblies in dry storage at Fukushima. Although jostled by the earthquake and submerged temporarily by the tsunami, these assemblies survived without the need for helicopters dropping water from above or fire trucks spraying water from below.

Damage inside the dry cask storage building at Fukushima Daiichi was extensive …

Damage inside the dry cask storage building at Fukushima Daiichi was extensive. Although the casks (one is visible, center) were jostled and submerged by the tsunami, they remained unharmed. Tokyo Electric Power Company

Dry storage was first used in the United States in 1986, and it is now practiced at plants across the country. After spent fuel assemblies have stayed in a pool for an initial period, workers transfer them into dry storage on-site. Typically, about fifteen tons of spent fuel are placed inside a sealed metal canister, then placed within a concrete and steel cask (the core at a large reactor holds nearly one hundred tons of fuel). Passive cooling—air flow from the chimney effect—removes the heat produced by the fuel.

Passive cooling does not work for densely packed spent fuel pools. The heat produced by the fuel assemblies—particularly those discharged from the reactor core within the past five years—requires continual cooling of the water in the pools.

A large radiation release from a pool could result in thousands of cancer deaths and hundreds of billions of dollars in decontamination costs and economic damage.

The crowded spent fuel pools at U.S. reactors pose hazards. An accident or terrorist attack could cause a loss of water from a pool or interrupt vital cooling systems. As Fukushima Daiichi demonstrated, it can be challenging to pump water up five levels into a pool for a boiling water reactor. And unless a hydrogen explosion rips apart walls and roofs, it can be difficult to drop water into a pool from above or spray water in from the side.

Transferring more assemblies into dry storage is the better way to manage risks posed by spent fuel. Reducing the spent fuel pool inventory accomplishes four things: It (1) reduces the heat load in the pool, (2) adds more water to the pool for every assembly removed, (3) allows the remaining assemblies to be spread out, and (4) reduces the amount of radioactive material in the pool.

The first two of these give workers more time to intervene in the event of a problem, increasing their chances of success. The third restores margin against inadvertent criticality within a spent fuel pool and provides additional space between assemblies for cooling air or water flow. And the fourth limits the size of the radioactive cloud that can be emitted from a pool. Thus, both the probability and the consequences of a spent fuel pool accident are lowered by moving fuel assemblies into dry storage.

There are lessons to be learned from what went wrong at Fukushima. There are equally important lessons to be learned from what went right.

NRC chairman Gregory Jaczko checked in with the Operations Center at White Flint at around 9:00 p.m. on March 15. Jack Grobe was on duty and gave his boss a rundown. Fresh NRC troops were due to arrive in Japan to provide relief to Jim Trapp and Tony Ulses. (By now, about two hundred people from a variety of federal agencies had arrived at the U.S. Embassy in Tokyo to assist.)

Reliable information was still hard to come by, Grobe told Jaczko, and the NRC was relying on “snippets” of information from NISA and U.S. and Japanese news media. “It’s been extremely frustrating,” Grobe said, adding that he hoped to glean more from industry sources in the United States. Also, the DOE’s Radiological Assistance Program teams were now on the ground and would soon be providing crucial land and aerial monitoring. That, Grobe said, “will be a huge benefit.”

Jaczko reminded Grobe that it would take at least two hours to obtain high-level authorization if an expanded evacuation order for U.S. citizens was deemed necessary. “I don’t anticipate a need for that,” Grobe told his boss. “[T]hings seem to be reasonably stable.”

Jaczko was due on Capitol Hill the next day for hearings in the House and Senate. He and his staff had been briefing lawmakers since March 11, but this would be the first opportunity for members of Congress to question the chairman in public. Media coverage was a certainty.

The day wasn’t over yet. At about 10:00 p.m., the NRC team watched Japanese television footage of smoke pouring from Unit 3. TEPCO announced that radiation levels were rising. A short time later, U.S. media began reporting that TEPCO was pulling its workers from the plant. Although it was unclear if that meant everyone, this was extremely troubling news. Without sustained intervention, the plant condition could quickly deteriorate.

Jim Trapp and Tony Ulses, who had called in to White Flint, confirmed that radiation levels were high. “Unit 4 is in shambles,” Trapp told his colleagues. He handed the phone to one of Ambassador John Roos’s assistants, who asked if the NRC headquarters knew more about the status of workers at the plant. Marty Virgilio read aloud a Washington Post bulletin: “The skeleton crew remaining at Fukushima Daiichi Nuclear Power Plant is being evacuated because of the risks they face from dangerous radiation levels.” (This wasn’t quite accurate; it later turned out that the workers had been pulled from the heavily contaminated control rooms but remained on-site, available to return if needed. Nonessential personnel had departed the plant, but they eventually returned.) The line went dead briefly and then Ulses explained the interruption: “We just had an earthquake here.”

Inside the Seismic Isolation Building, superintendent Yoshida and his fellow workers often had to rely on reports from the emergency crews making brief forays out onto the plant grounds to know what was taking place. Only then were they able to assess just how bad things had gotten. After one explosion, the men inside could hear debris raining down on the roof of the bunker-like building.

“I really felt we might die,” Yoshida told a reporter in a rare interview, nearly a year and a half after the accident.1

Working conditions in the Seismic Isolation Building rapidly worsened. Personal dosimeters, used to measure radiation levels, were in short supply. Many had been destroyed by the tsunami or rendered useless when their batteries died. Those still functioning were reserved for workers forced to venture outside. Protective clothing also became scarce. Meals consisted of biscuits and dried food, and water was rationed. Crews slept in the hallways.

To support his people, Yoshida said, he passed out cigarettes in the heavily used smoking room. He encouraged them to record their names on a large whiteboard as a memorial in case they did not survive. In the circumstances, he told the interviewer, “it was clear from the beginning that we couldn’t run.”

Tapes from the videoconference link between the Seismic Isolation Building and TEPCO headquarters, released by TEPCO more than a year later, reveal a chaotic scene. Tempers flare as Yoshida and his team are besieged with orders and counterorders from TEPCO headquarters, which was also getting instructions from the prime minister’s office.

Yoshida would later tell government investigators he believed TEPCO’s chain of command during the accident was “disastrous.” When he was ordered on March 12 to suspend injecting seawater—the order he ignored—Yoshida said he thought: “What the hell is the office talking about?”

By March 18, his patience had all but run out. When officials at headquarters directed him to send crews out to conduct another check, Yoshida snapped: “My people have been working day and night for eight straight days… . I cannot make them be exposed to even more radiation.”

In the wee hours of March 16, East Coast time, Charles “Chuck” Casto called White Flint from Tokyo, where he had arrived a short time before. Casto, a deputy regional administrator in the NRC’s Atlanta office, had been chosen to head the NRC team in Japan.2 (He would spend almost a year there.) Along with eight other NRC staff members, Casto joined Trapp and Ulses in the U.S. Embassy. The NRC crew set up shop in a large conference room.

Charles Casto

Charles Casto

Bill Borchardt

Bill Borchardt

Marty Virgilio …

Marty Virgilio. U.S. Nuclear Regulatory Commission

The first thing that Dan Dorman, in charge overnight at White Flint, asked Casto for was an update on the Unit 4 spent fuel pool. Casto promised to get what he could. “As you know,” he said, “the communications channels are very limited.” Even though he had arrived so recently, Casto had picked up the high level of concern pervading the embassy. “[T]hey’re worried about the 170,000 [sic] Americans over here,” he said. “That’s their primary goal. They’re here to protect them… . [T]hat’s got to be our focus.”

Marty Virgilio, who was on the line in White Flint, concurred. “[T]he top priority is to support the ambassador and the U.S. citizens there. I would say second is support for the Japanese government in the recovery of these reactors. And, then, third is gathering insights and lessons learned for us, so that we can assess the implications for our [reactor] operating fleet and applicants.”

Late the previous evening, a RASCAL simulation had estimated radiation doses from a spent fuel fire in a pool that had lost all its water, as well as its roof, and released 50 percent of its radioactive contents to the atmosphere. Just as for the case of Unit 2 with no containment, the calculation indicated evacuation would be needed at least out to fifty miles downwind, RASCAL’s limit. So now two possible scenarios would warrant a fifty-mile evacuation, and the NRC still had no way to know whether either was occurring.

But a decision had to be made soon on evacuation advice to Ambassador Roos, and the NRC team needed to judge whether its disaster scenarios were plausible. There would be hell to pay if the United States expanded the evacuation zone and it turned out that the Unit 2 containment remained intact and there was no spent fuel fire at Unit 4.

The lack of solid data about the status of Fukushima Daiichi frustrated everyone. “It was like trying to investigate a homicide and not having access to the crime scene,” Casto would later say. “There was just so much misinformation, lack of information, and you didn’t know what information to trust.” The NRC crew at White Flint agreed to try harder to get details out of the U.S. and international nuclear industry. GE Hitachi, for instance, had its own emergency operations center in Tokyo; if need be, Casto suggested, Virgilio could ask top-level GE officials in the United States for access.

Casto had spent a portion of the fourteen-hour flight to Tokyo buried in a couple of NRC technical publications (known as “NUREGs”) relevant to the crisis. For decades, the NRC has funded contractors at U.S. national labs to run computer simulations of possible reactor accidents. Casto was struck by how closely these simulations matched the real-world situation at Fukushima. As he told his colleagues, “of course, that Mark I containment is the worst one of all the containments we have … this NUREG tells you that in a station blackout you are going to lose containment. There’s no doubt about it.”3

About ninety minutes later, John Monninger, an engineer with Casto’s group, called White Flint to report that the NRC team had been invited to a meeting with TEPCO and Japanese nuclear regulators. TEPCO apparently wanted guidance on dealing with spent fuel pools. Monninger also shared bad news: the pools in Units 1 and 2 were “boiling down.” Those in Units 3 and 4 were displaying zirconium-water reactions, meaning that the fuel was exposed. Further, the explosion at the Unit 4 reactor building had “leveled the walls, leveled the structure for the Unit 4 spent fuel pool all the way down to the approximate level of the bottom of the fuel,” Monninger said. “So, there’s no water in there whatsoever.”

TEPCO was talking about dumping sand in the pool, which would reduce the radiation levels but do nothing to cool the overheated fuel rods. The NRC group agreed that a combination of water and boron (to reduce the risk of criticality in the pool) seemed the best answer. The headquarters team members promised Monninger that they would come up with further suggestions before their colleagues in Tokyo left for the meeting in about forty-five minutes.

While the NRC was projecting radiation dose levels under various scenarios, the DOE’s Office of Naval Reactors was doing its own assessments, and both agencies were feeding their recommendations to the U.S. ambassador. The two groups weren’t seeing eye to eye, it seemed. To the Office of Naval Reactors, the radiation readings detected on naval vessels and at bases far from Fukushima indicated that the event was much worse than the NRC had painted it. While the NRC was estimating the radiation release from damaged fuel in a single reactor, experts at Naval Reactors saw a far greater danger; they believed the NRC was “undershooting” the potential release. As Stephen Trautman, deputy director of Naval Reactors, told the NRC by phone, “[I]f we end up losing one of these plants, there’s a good possibility we’re going to lose all of these plants.” The two sides agreed that the threats from the spent fuel pools also had to be factored in.

For U.S. officials, the evacuation decision was a delicate balancing act. On March 13, Ambassador Roos had put out a press release telling Americans in Japan to “follow the instructions of Japanese civil defense authorities,” who were evacuating only a 12.4-mile (twenty-kilometer) zone around Fukushima. Now the ambassador was getting information that suggested the advice might not be sufficient. Yet if the United States were to unilaterally launch a wider evacuation, it could anger an important ally.

Within the embassy itself, fears were high. When Casto made trips to the cafeteria, he was often stopped by staff members, including pregnant employees worried about their unborn children, asking: how bad is it? “There was significant concern,” he recalled.

To arrive at a unified recommendation for Ambassador Roos, a high-level phone conference was arranged with Naval Reactors and the NRC teams in Japan and White Flint. During the call Jaczko asked: “[D]o we think this is going to get better or get worse?” Marty Virgilio responded, “It’s going to get worse if you think about the spent fuel pools. Right now, Unit 4 doesn’t have a spent fuel pool anymore.” Water levels couldn’t be maintained in the Units 1, 2, and 3 fuel pools, he added.

At worst, Jaczko summarized, the accident could involve three reactors “out of control” and possibly up to six spent fuel pools. The ambassador should base any evacuation decision on that information, he said, and then asked: “Does anybody disagree with that?”

“Chairman,” said Admiral Kirkland Donald of Naval Reactors, “I agree with you.”

“If this happened in the U.S.,” added Bill Borchardt, the NRC’s executive director and a veteran of the U.S. nuclear navy, “we would go out to fifty miles. That would be our evacuation recommendation.”

It was about 6:30 a.m. Wednesday, March 16. It was time to call the White House.

Jaczko’s summary to personnel in the White House Situation Room was succinct. In even terser language, a White House official repeated Jaczko’s report to confirm: “Three reactors melting down, six spent fuel ponds go up in flames. And what, then, would be the impact for Tokyo, for example, if the wind kept blowing in that direction?”

“At this point, I think I would still go with the fifty miles right now,” said Jaczko, noting the lack of solid information. However, he cautioned that even if the estimates were only “off by a little bit,” the impact could be “significantly larger.”

Before Jaczko’s scheduled 9:30 a.m. testimony on Capitol Hill, he and Borchardt, who was accompanying the chairman, called in to White Flint for updates. Which posed the larger threat now, Jaczko asked his team: Unit 2 or the spent fuel pools? The pools, he was told.

After Jaczko hung up, Borchardt remained on the line to seek more details. Casto, participating from Japan, said that in addition to the threat posed by the Unit 4 spent fuel pool, one had to assume that conditions at the other endangered reactors and spent fuel pools would also deteriorate because the NRC team wasn’t seeing any “mitigation” of the ongoing crisis.

Borchardt asked: “I just want to make sure that no one takes anything he [Jaczko] says as implying that, if they resolve the issues with Unit 2, life is going to be a lot better. Right?”

“No,” someone else on the call replied.

Discussions at White Flint now focused on how best to deal with the fuel pools. The Japanese were talking about using a helicopter to fly over Unit 4 and dump sand or water into the gaping hole where the roof of the building had been. Casto was growing impatient. “I don’t know what they’re waiting for.”

Radiation levels at sixty-five feet (twenty meters) above the open pool would be “hundreds of thousands of rads,” came the reply. Helicopter pilots who flew over Chernobyl to dump sand on the burning reactor there “flew until they died, basically,” said one member of the NRC team.

Borchardt called again to “triple-check” the fifty-mile evacuation recommendation. Was it still a go? Charlie Miller said yes, based on current meteorological conditions. But if conditions changed, even a fifty-mile evacuation might not be sufficient. If the wind shifted, Miller added, even Tokyo—150 miles away—could be affected.

A short time later, Monninger called White Flint to report that two government ministry employees had asked for help acquiring emergency equipment to move water from the ocean to the reactor buildings, plus water cannons and four or five trucks with aerial booms capable of spraying water at a height of sixty-five to one hundred feet (twenty to thirty meters) to reach the fuel pools. All of this apparatus would need to be positioned quickly because the radiation level near the reactors was thirty rem (three hundred millisieverts) per hour. Robots with remote cameras and radiation measuring equipment also would be useful, the Japanese said.

“One thing, we went back and forth on [the] Unit 4 [spent fuel pool],” said Monninger, “… because we believe the walls were blown out and the water level is at the bottom of the active fuel. That took them back, you know, quite a bit… . They said they don’t have any indications of that. They believe, which would be true, there will be screaming radiation levels if those walls were knocked out. They said the walls … could have been knocked out, but the spent-fuel pool wall may not be the outermost wall.”

The Japanese had shown Monninger a “bird’s eye schematic” of the Unit 4 spent fuel floor. “They had a decent level of confidence that … the spent-fuel pool walls hadn’t been blown out and the success path was to flood that up and not to go with sand, dirt, or whatever.”

Borchardt called again from Capitol Hill looking for any new information. Monninger’s report from the Japanese was repeated: “They don’t think the spent fuel pool for Unit 4 is as degraded as we thought.” Borchardt asked for more details. Cooling in the pool still may be possible, he was told. Monninger added that he believed reactor fuel had been ejected in the explosion at Unit 4 and was now lying around the plant site, which would explain high radiation readings on the ground. The Japanese were bulldozing dirt over the rubble, and the radiation levels dropped by as much as 70 percent.

“So the potential is solely to refill the pool?” Borchardt asked.

“That’s correct,” came the reply. Even so, Borchardt was told, “there’s no change to the recommendations that we’ve made.” The fifty-mile evacuation stood. As best as the White Flint crew could tell, Units 1 to 3 had damaged cores and the spent fuel pool at Unit 4 was almost dry.

The NRC had sound reasons for its suspicions. There was no clear proof, but there was plenty of circumstantial evidence. The Unit 4 pool contained the most, and the hottest, spent fuel among the reactors at Fukushima. It would therefore be the first to boil after cooling was lost. The Unit 4 reactor building had blown apart, as TV replays nearly constantly reminded them; the most likely reason would be the detonation of hydrogen gas. Since the Unit 4 reactor core held no fuel, the most likely source of hydrogen gas was exposed fuel in the pool. If the fuel were uncovered, its metal cladding could ignite—explaining the fire reported in Unit 4. And the high radiation levels reported between the Units 3 and 4 buildings could be caused by debris, perhaps even portions of fuel rods, ejected from a dry Unit 4 pool during the explosion.

Finally, observers reported white vapor emanating from the Unit 3 reactor building while nothing visible wafted from Unit 4. Based on this latest circumstantial evidence, it was reasonable to conclude that the Unit 4 spent fuel pool had boiled or drained itself dry while the Unit 3 pool was boiling toward that outcome.

But still, there was no definitive proof. The Japanese could turn out to be right after all, but it would be many days before anyone could get close enough to Unit 4 to find out.

✵ ✵ ✵

Jaczko had been scheduled to discuss the NRC’s 2012 budget with two House Energy and Commerce subcommittees, but Fukushima was on everyone’s mind. Before Jaczko could testify, he was summoned to a meeting at the White House. Jaczko told President Obama and his national security advisors that the NRC recommended the evacuation of Americans living within fifty miles of Fukushima Daiichi. It was, the chairman said, the advice the NRC would give if this incident were taking place in the United States.

NRC chairman Gregory Jaczko tells a House subcommittee hearing on March 16 that the NRC believed the spent fuel pool at Unit 4 was dry, raising the possibility that the spent fuel could catch fire and release additional radioactive material …

NRC chairman Gregory Jaczko tells a House subcommittee hearing on March 16 that the NRC believed the spent fuel pool at Unit 4 was dry, raising the possibility that the spent fuel could catch fire and release additional radioactive material. The announcement came shortly after the United States advised its citizens living within fifty miles of Fukushima Daiichi to evacuate or remain indoors if evacuation were not possible. U.S. Nuclear Regulatory Commission

A short time later, White House Press Secretary Jay Carney alerted reporters to the expanded evacuation recommendation during a briefing. The State Department had issued a travel warning half an hour before, citing the “deteriorating situation” at Fukushima Daiichi. U.S. citizens living within fifty miles (eighty kilometers) of Fukushima were advised to leave or take shelter indoors if evacuation was not possible. In addition, the warning said, “the State Department strongly urges U.S. citizens to defer travel to Japan at this time and those in Japan should consider departing.”

Washington’s fifty-mile evacuation recommendation was about to get some competition in the news. Before he began his testimony on Capitol Hill that afternoon, Jaczko called in to the White Flint operations center. “[J]ust to repeat, we believe pool No. 4 is dry, and we believe one of the other pools is potentially structurally damaged?” he asked. “That’s correct,” Chuck Casto said from Tokyo. “That’s the best we know.”

“I mean the relevant factor is it’s dry,” said Jaczko. “Yes,” Casto confirmed, “and they can’t maintain [water] inventory at all.”

Jaczko hung up, but the conversation continued between the NRC experts in Tokyo and White Flint. They realized they had neglected to tell the chairman that the Japanese didn’t agree with the NRC’s assessment about the Unit 4 pool. They didn’t call him back, though, because they thought he didn’t want that level of detail. And there were plenty of other problems to worry about. The Unit 2 spent fuel pool now appeared to be in trouble; steam was rising through the hole created in that reactor building’s wall by the Unit 1 explosion, so that pool might be boiling dry. Steam or smoke was coming from the Unit 3 pool, meaning the water level there could be dropping. Unit 1 appeared to have the only pool of the four not in immediate trouble.

When Jaczko took his seat in the large House hearing room, he launched into the latest news from Fukushima. At three reactors there is “some degree of core damage from insufficient cooling,” he said. At Unit 2, cooling was not stable but the primary containment was functioning. Water levels in the Unit 2 spent fuel pool were decreasing. The integrity of the Unit 3 pool had been compromised.

Then he moved on to Unit 4. “We believe that secondary containment has been destroyed and there is no water in the spent fuel pool,” he said, “and we believe radiation levels are extremely high, which could possibly impact the ability to take corrective measures.” Jaczko said the highest priority now was to maintain cooling of the reactors at Units 1, 2, and 3 and to keep water levels up in the spent fuel pools. If cooling functions should fail, he warned, “it would be very difficult for emergency workers to get into the site and perform emergency actions.”

What poses the highest risk at the moment, asked Rep. Henry Waxman, a Californian with a long involvement in nuclear safety issues.

“All of the factors together, really, the combination,” Jaczko said. “There’s the possibility of this progressing further.” If the cooling systems failed, he explained, emergency workers “could experience potentially lethal doses in a very short period of time.”

Jaczko hustled out of the House hearing and across Capitol Hill to the Senate, where Bill Borchardt had been pinch-hitting for his boss before the Senate Environment and Public Works Committee, which had convened specifically for a briefing on Fukushima. The chairman was grilled about the safety of U.S. reactors, especially those in seismically active areas, such as the two in the home state of the committee’s chair, California senator Barbara Boxer.

After Jaczko’s testimony, the committee heard from representatives of the Union of Concerned Scientists and the Nuclear Energy Institute (NEI), the primary nuclear industry lobbying group in Washington. They were asked to address whether a Fukushima-type accident could happen in the United States and if so, what needed to be done to preclude such a disaster.

Tony Pietrangelo, chief nuclear officer of the institute, chose his words carefully: “I think I understand your concern, because I share it, that people are seeing what’s happening in Japan and they’re scared. We can never say that that could never happen here. There’s no such thing as a probability of zero… . But what I would tell you is it doesn’t matter how you get there, whether it’s a hurricane, whether it’s a tsunami, whether it’s a seismic event, whether it’s a terrorist attack, whether it’s a cyberattack, whether it’s operator error or some other failure in the plant, it doesn’t matter. We have to be prepared to deal with those events.”

Edwin Lyman of the Union of Concerned Scientists agreed that a Fukushima- type event could not be ruled out in the United States: “We have plants that are just as old. We have had a station blackout. We have a regulatory system that is not clearly superior to that of the Japanese. We have had extreme weather events that exceeded our expectations and defeated our emergency planning measure[s], [such as] Hurricane Katrina.”

While Pietrangelo said that the industry had to be prepared to deal with severe events like natural disasters, he did not say that the industry actually was prepared to deal with them. In fact, he told the committee that Fukushima had prompted an industry-wide effort to “verify each company’s capability to mitigate conditions that result from severe adverse events,” including total loss of electric power, earthquakes, and floods. At the time, he did not anticipate that those reviews would turn up numerous problems, calling into question the level of preparedness of U.S. reactors.

Pietrangelo informed Boxer that “we already have mobile—diesel-driven mobile pumps on every site in the country that can be moved around the site to provide another contingency measure should we lose a cooling source. And there’s countless other measures like that.”

Those measures are known as “B.5.b,” named after the section of the regulatory orders where they appear.4 They were among the steps that the NRC required utilities to take after the 9/11 attacks to prepare for “loss of large areas of a plant due to explosions or fire”—the kind of damage that might be expected from a terrorist aircraft attack. For example, utilities must store B.5.b equipment far enough away that it might be able to survive a plane attack on a reactor. But there is no requirement that the equipment survive other types of disasters, such as an earthquake. When Lyman asked about the ability of the emergency pumps to survive an earthquake, Pietrangelo said that the pumps are not certified to withstand any earthquake at all, much less a severe one.

In other words, in the face of a Fukushima-scale event, the B.5.b measures could well be worthless. But in the days after Fukushima, both the industry and the NRC cited the B.5.b measures in response to the question they had both feared: “Can it happen here?”

Jaczko’s assertion that the Unit 4 spent fuel pool was dry apparently didn’t sit well with the Japanese. Casto and Tony Ulses were invited to the Kantei, the prime minister’s official residence and office. The Japanese wanted to show the Americans a video of the Unit 4 spent fuel pool—indicating it contained water. Monninger told the White Flint crew: “We think the reason they’re doing that is because, I guess, of maybe statements the NRC has made or maybe the chairman’s hearing testimony … saying that Unit 4 the spent fuel pool was dry.”

A short time later, Casto reported back after watching the video, which the Japanese declined to provide to the Americans. “You know, it’s not very clear. You’re talking about a helicopter that’s trying to do a lot of things at once in a field. And they tried to scan all four [reactor] units. You have to look through a window. And they claim there’s a reflection of water on the Unit 4 spent-fuel pool… . There’s something there. You don’t know if it’s steel or water. They claim it’s a reflection.”

Steam was visible coming from the side of the building nearest the spent fuel pool, said Casto, obviously frustrated. “You’ve got a building that’s had an explosion and has debris everywhere, and you’re trying to look at it with a helicopter that’s flying by in split seconds. You can’t tell anything in there. You know, they claim there’s a glimmer of a reflection but, you know, it’s steaming. Unit 3 is steaming even harder.”

Jaczko, who had joined the call, interrupted Casto. “So, at this point, you no longer believe that the pool is dry?” Casto replied: “I would say, as of five o’clock yesterday, the pool had some water in it.”

“Okay. Now I’ve said publicly the pool is dry,” said Jaczko. “Do you think that’s inaccurate?” Casto replied: “I would say it’s probably inaccurate to say it’s dry.”

“Do you think I need to roll back any of the statements that I made?” Jaczko asked.

“I don’t think so,” Casto said. “It may not have been dry, but it certainly wasn’t full.” There was no need to change the fifty-mile evacuation advisory, the team assured Jaczko.

As debate swirled at White Flint, President Obama and Prime Minister Kan were on the phone. Obama promised continued technical assistance to Japan and the two leaders discussed the steps the United States was taking to protect its citizens. After the conversation, Japanese cabinet secretary Yukio Edano told reporters in Tokyo that the two countries would do a better job of sharing information.

As for the Americans, everyone understood that the United States needed to tread lightly. As Chuck Casto later observed of the situation: “The best science person can’t bring science to it if they don’t have the right diplomacy skills.”

Convinced that the Unit 4 spent fuel pool held water, the Japanese focused their attention on Unit 3. The white smoke seen billowing from that unit the day before—as well as a spike in radiation—made getting water to the fuel there a priority. A three-pronged attack began shortly before 10:00 a.m. on Thursday, March 17.

It consisted of dropping water from Self-Defense Forces helicopters and spraying water from the ground using SDF fire trucks and high-pressure water cannon trucks that the Tokyo Metropolitan Police Department used for riot control.5

An initial attempt to dump water from an SDF helicopter on March 16 had been abandoned because of high radiation levels. Lead plates lined the bellies of the helicopters and the crews wore protective clothing when they took off the next day.

Over the course of ten minutes, two CH-47 Chinook helicopters dropped thirty tons of water on the upper parts of the Unit 3 reactor building during four flights. NHK television cameras, positioned about twenty miles from the plant, captured the drops in footage viewed around the world. Although a small amount of white steam rose from the building, very little water apparently reached the pool. Most was carried off by strong winds, and some was deflected by debris atop the pool. The water cannons were useless and the fire trucks proved inadequate; neither could deliver the water where it was needed. Radiation levels prohibited moving the equipment closer.

TEPCO also focused on restoring power to the plant via a new transmission line, although utility officials were unable to say when that might be completed. The plan called for repowering Unit 2 to restore cooling to its spent fuel pool. Because Unit 2’s reactor building was intact aside from a small hole, it would be difficult to spray water inside (in contrast to Units 1, 3, and 4, which no longer had roofs). But given the extensive damage to the plant’s wiring systems, there was no assurance that off-site power would be of much use.

The first radiation readings gathered by the U.S. National Nuclear Security Administration’s Aerial Measuring System, part of the DOE’s Radiological Assistance Program, were providing independent data. Until now, most of the measurements had come from TEPCO via NISA, and Washington wanted its own. The U.S. sensors not only measured how much radiation was present, but also identified the isotopes emitting it—the nuclear fingerprints of dangerous materials including cesium-137. In addition, a U.S. Air Force Global Hawk drone flew over the plant site, measuring temperatures and gathering high-resolution imagery.

While the Japanese struggled to get water into the pools, the Americans were offering five heavy-duty pumps from the U.S. Defense Threat Reduction Agency that were available for pickup three hours away from the plant. (Earlier, the Japanese had rejected two fire trucks offered by the U.S. Air Force because the vehicles were not registered in Japan and thus could not be legally driven on the roadways, an act of bureaucratic nitpicking that amazed the Americans.) The pumps could deliver seawater to the fuel pools—assuming the pools were still intact. That worried Casto. “The pumping strategy may not be useful at all if there is no spent fuel pool, and there is just a rubble bed in there somewhere,” he said.

Even so, it seemed a logical plan of attack. At 6:00 a.m. on March 17, the Japanese SDF picked up the pumps at Yokota Air Base and set out for Fukushima Daiichi, 196 miles away.

✵ ✵ ✵

Delivering the water would require more than pumps, however, and U.S. experts had spent hours conferring on just what was needed. Pipes and strainers (to keep debris in the ocean water from clogging the pumps or the spray nozzles) had to be acquired. Because of the extremely high radiation, installation of the system had to be completed as rapidly as possible. “[I]t’s got to be on tractor-trailers, minimum disconnect,” one member of the U.S. team said. “They would drive in, hook this thing up, throw the portable submersible pump out in the ocean, and light this system off and leave.” The NRC asked the giant engineering firm Bechtel Corporation to design the complete cooling systems that would be needed and to arrange for procurement and delivery of all the additional equipment to the site, but not to perform the dangerous jobs of installing and operating the systems, which would be left to the Japanese.

The NRC was developing a radiation dose map of the plant site to help installers minimize their exposure. “[I]t’s going to be heroic efforts, because you’re going to have somebody running toward the pool,” said one of the White Flint staffers. Boron would be added to the water to reduce the likelihood of the spent fuel going critical. Twenty thousand pounds of boron had been located at California’s Diablo Canyon nuclear plant and was being flown from Vandenberg Air Force Base to Yokota.

An exhausted Chuck Casto called in to White Flint to say the pumps had arrived at the plant and efforts were now under way to reduce the radiation levels with sand or lead shielding so the system could be installed. “[Y]ou know there are lethal dose rates they’re getting outside that building,” said Casto.

The discussion turned to Unit 2, and Casto was asked about its status: had the primary containment been breached? He did not know. A NISA summary was vague, Casto said, referring only to the possibility of an “incident” in the suppression chamber, based on the sound of an explosion and a subsequent reduction in the measured pressure in the torus.

For a moment, the magnitude of the crisis for the Japanese overwhelmed Casto, who noted that a thousand bodies had washed ashore from the tsunami. “[I]t’s hell over here for that government. I mean, it’s just absolutely hell. And I know we get frustrated with them, but, man, when you think about what they’re faced with, it’s absolutely unfathomable.”

Although the United States was working out details on the pump system, it was a one-sided effort. “We are not getting any takers from the Japanese side of this equation,” Virgilio told his colleagues. Although the SDF had collected the pumps and delivered them to the plant, the Japanese apparently were afraid that trying to hook up the system would be too dangerous because of the high radiation levels. So they were intent on continuing their spraying operations. “They are using helicopters the same way you fight forest fires,” said a member of the American team.

At about 8:00 a.m. Thursday, Dan Dorman, on duty at White Flint, briefed the NRC team via a conference call. Events had moved both Unit 4 and Unit 2 down the list of concerns. The highest priority now was the Unit 3 spent fuel pool, which appeared to have experienced a zirconium-water reaction and a resulting fire, he said. Over the previous twelve hours, the Japanese had tried to drop and spray water into the Unit 3 pool, but without much effect. Seawater injection to the reactor vessels continued in Units 1, 2, and 3, with water levels in the cores at about half the height of the fuel. There were mixed reports on the status of the Units 2 and 3 containments. TEPCO was still trying to restore power to the site. The pumps were in place but it was not certain that they could develop enough pressure to spray water high enough to reach the pools. New radiation measurements had been collected by aircraft overnight and were being analyzed. Radiation levels of 375 rem per hour were measured three hundred feet above the Unit 3 reactor.

To everyone’s surprise, the Japanese asked the NRC to list its priorities on controlling the situation at Fukushima Daiichi. Although the turnabout was welcome, fulfilling the request would put the NRC in the position of directly offering advice, something it had thus far avoided. Making a mistake weighed on Casto. “I don’t want to try any strategy that doesn’t have sound measures, or reasonably sound measures, of success,” he told Virgilio. “Because the last thing we want to do is make the situation worse for the Japanese. We’re going to own this thing if we do.”

Radiation exposure fears were growing beyond the vicinity of Fukushima Daiichi. The NRC was aware that significant levels of fallout could occur outside of the fifty-mile evacuation zone. For data on that, it had to depend on the DOE, which was using its more sophisticated atmospheric modeling tools to develop dose projections extending one hundred and fifty to two hundred miles from the plant, an area that included Tokyo. The DOE was also running simulations to estimate possible doses to the public in Alaska, Hawaii, and the West Coast, and it was getting some alarming results: for instance, doses as high as thirty-five rem to the thyroids of people in Alaska. The NRC didn’t trust the numbers the DOE was coming up with, but it didn’t have the means to disprove them.

In the United States, the Department of Homeland Security was screening passengers returning from Japan. In China, grocery stores in several cities were stripped of iodized salt after rumors spread that a radioactive cloud was moving in. Some buyers mistakenly believed that iodized salt would protect against radiation; others feared that sea salt supplies from the coast would be contaminated. The city of Florence, Italy, sent a plane to Japan to pick up its stranded orchestra and the conductor, Zubin Mehta. (Maestro Mehta would return a month later for a performance to benefit disaster survivors.)

Casto and the NRC squad in Tokyo were swamped with meetings—with the ambassador and embassy staff, with other U.S. officials aiding in the American response, with Japanese government officials. The crew was running on six hours of sleep or less, with sessions stretching well into the early morning hours. Noticeably absent, however, were representatives of TEPCO. The utility seemed intent on going it alone.

Finally, TEPCO officials did agree to sit down with the Americans to discuss a pumping system the United States had proposed for cooling the spent fuel pools. It was the first time the utility had heard of the U.S. plan, Monninger told his colleagues at White Flint. Apparently NISA and METI, which had been involved in the discussions of the U.S. plan, had not bothered to apprise TEPCO.

Yet the United States had already given Bechtel the go-ahead to assemble four separate pumping systems, or “trains,” one for each of the fuel pools. By March 18, the components had been gathered in Australia, awaiting an airlift to Japan. Only then did the U.S. team discover that the Tokyo Fire Department had assembled a similar pumping system with local parts for use at Fukushima Daiichi.

No matter whose system would be used, the problem, of course, was the final leg: getting close enough to lay water lines near the reactors and pools. Radiation levels between Units 2 and 3 now ranged between 450 and 600 rem per hour.

TEPCO’s preferred and potentially less risky approach was to use two concrete pumping trucks, similar to those used to encase the Chernobyl reactor after that accident. The trucks, equipped with booms capable of extending about two hundred feet, could remotely spray seawater into the fuel pools with precision. Two such trucks were in Japan, and additional ones could be obtained from abroad.

Casto was still worrying about pumping water onto the hot fuel. The risk was a massive radioactive steam cloud “that’s going to … panic everybody.”

“We’re in such never-never land, and I don’t really know how to make decisions when you have very little information,” Casto said, and urged his colleagues to seek additional guidance from experts inside and outside the government. “We’re going to need all the great minds together,” he added. “Let’s solve this as an American nuclear industry helping the Japanese nuclear industry, not just the NRC.”

HELPFUL HINTS

As the experts in Japan and the United States struggled to find ways to cool the cores and stem the radiation releases spilling from Fukushima Daiichi, average Americans had no shortage of ideas—nor any reluctance to share them. They flooded the NRC with suggestions via e-mails and phone calls.

“I’m sure y’all have many experts trying to assist in the emergency in Japan,” wrote Sharon H., “but I thought I would write just in case this idea had not been considered”: cornstarch.

“Have you considered … fine ground black pepper to seal the crack at Fukushima Daiichi?” wrote Shawn M. “Yes, common household pepper is very effective as a crack sealant. I once sealed a long crack (8") in an engine block. It held for years and I did not need to replace the engine.”

Among other suggestions: Ping-Pong balls, antifreeze, a ski resort snowmaker to blow snow into the core. Coal ash. An asbestos blanket. Liquid nitrogen. Putty.

Vinny S. proposed using tiles from the space shuttle. “The shuttle program is ending and we could recycle materials and be eco-friendly while ensuring safety at the reactors.”

Others were more insistent: “Get this information to Japan or tell them when they call you,” directed one e-mail. “I was divinely led. God is telling me it will work.” What followed was a detailed list of instructions entailing “vector forces,” a “metrogravitron particulator,” and a device known as the “Deometrian.”

Most suggestions received a standard response from the NRC’s Office of Public Affairs: “We appreciate suggestions that work toward resolving the situation in Japan; it’s reassuring to see how helpful and dedicated private citizens have been in light of this disaster… . Please understand that the NRC has some of the most expert people in the world available to assist the Japanese authorities in whatever way they request.” The NRC suggested the Americans also contact the Japan Atomic Energy Commission.

One public-spirited citizen apparently got a runaround before finally landing at the Office of Public Affairs. His name was Harold Denton. He’s “been calling all over NRC,” wrote staffer Amy Bonaccorso. In case the name didn’t ring a bell, she added that Denton was the “spokesperson for the NRC during Three Mile Island—he also was the person [who] provided daily reports to Pres. Jimmy Carter.” Denton, who also delivered daily briefings to hundreds of reporters during the Three Mile Island accident, wanted to urge the commission to make its experts available to the media, she wrote, because reporters “are reaching out everywhere for info … including retired employees (with old information).” Bonaccorso’s boss promised to call Denton.

Stepping into the information void on March 15 was TV personality Glenn Beck. “Glenn Beck is now explaining the China Syndrome, using a wok as a containment vessel,” an Office of Public Affairs staffer e-mailed his colleagues, to which one responded: “ohmigod.” “Maybe it’s a big mixing bowl, I can’t tell,” said a follow-up e-mail. That wasn’t Beck’s only prop: M&Ms made handy fuel pellets.

That collaboration was already in the works. The next day, Saturday, March 19, a group of experts from industry met with their government counterparts at White Flint. The model, Virgilio explained, was the public-private response to the Deepwater Horizon oil spill in the Gulf of Mexico in 2010. Attending were representatives from the Institute for Nuclear Power Operations, which was organizing the meeting, GE, Exelon, the French nuclear conglomerate Areva, the DOE, and the Electric Power Research Institute. Experts within the nuclear industry had regularly been conferring with the NRC and Japanese authorities as well as conducting their own research. The Americans hoped that by merging the private and public response, TEPCO might become more receptive to outside help.

The Japanese government might also need some convincing. “We’re going to start working through diplomatic channels in-country to try to make sure that what we develop is implementable from a political standpoint,” said Virgilio. Key to any public or private response was reliable data. And that was still in short supply.