Nuclear power plants are some of the most closely guarded places in the world. In order to get to the plant exits on the other side of the complex, the ocean-side workers needed to go back into the main part of the plant through an entry gate. That meant a security check with metal detectors. The guards at the entry gate kept the workers waiting. One worker recalled the panic building among those trapped outside the gate. “Let us out of here! A tsunami may be coming!” they yelled. They worried that if they tried to climb the fence to flee, they could be prosecuted. Eventually, the guards allowed them to pass through.
Inside the complex, fallen office equipment had trapped some workers under desks and they needed to be freed by their co-workers. In the building housing reactor 3, a crane operator was stuck in the cockpit of his crane. His colleagues rushed in with a flashlight to help him find his way down.
On the inland side of the plant, workers gathered at the exit gates. Another worker, who asked to be called Kai Watanabe when he later told his story, remembers people standing in an orderly line to turn in their dosimeters—instruments that measure radiation exposure—and waiting as their supervisors counted the workers to make sure no one was missing. Once they’d made it through the gates, some, including Kai, rushed off to try to reach family members in Okuma and Futaba before the tsunami did.
Hundreds of employees who had been designated as emergency workers headed for an earthquake-proof structure called the seismic isolated building, which was set back from the water. To get in, they had to navigate more than 60 feet of stairs that had been covered in slippery sludge. “When I arrived,” one worker remembered, “a ruptured ground pipe was spraying water like a geyser and had caused a mudslide that covered the stairs.”
On the second floor, in the emergency response center, the plant superintendent, Masao Yoshida, was trying to get a handle on the situation. When the first two tsunami warnings, predicting 10- and 20-foot waves, were announced on TV, he ordered all workers to evacuate to either the seismic isolated building or high ground. The third tsunami warning, predicting a 32-foot wave, arrived too late.
At 3:27 P.M., just forty-one minutes after the earthquake, the first wave crashed into the Fukushima shore. Reactors 1 through 4 sat 33 feet above sea level; reactors 5 and 6 were at 42 feet above. All were safe from the first wave, which was only 18 feet high. But when the second wave barreled in ten minutes later, it was closer to 50 feet tall—about as high as a five-story building. It easily rolled over the plant’s 30-foot seawall, sweeping a tractor trailer into the building complex and pulling a massive oil tank into the ocean when it receded. Water swamped the turbine buildings on the ocean side of the reactors, and climbed 18 feet—about two stories—up the outsides of units 1 through 4.
A maintenance worker was in the reactor building at unit 2 when he heard rumbling in the basement. He raced for the stairs as seawater poured through a service entrance. He was soaking wet when he reached the control room on the top floor. Another worker, trying to reach a building near unit 4, was trapped at a security gate. He was trying to contact the guards using an intercom when the tsunami flooded in. “Just as I thought I was going to die with water encroaching from below,” he later remembered, “a senior employee in the same situation broke the glass of the gate and escaped, and then helped me out by breaking the glass on my side. When I did escape from the confinement, water had inundated to the height of immediately below my jaw. I was really scared.”
A car is tossed like a tub toy by a wave that inundated the North side of the Fukushima Daiichi plant. This photo was taken from the fourth floor of the plant’s radiation waste treatment facility.
Despite the close call, he was fortunate. Two of his co-workers, who had been sent to the basement of unit 4 to check the status of the reactor, drowned.
* * *
Sheltered behind the reactors in the windowless control rooms, the operators couldn’t see or hear the incoming waves. But they knew they were in trouble. In the minutes after the scram, warning lights had lit up and alarms had started blaring. At 3:36, the lights on the control panels in the unit 1 and 2 control room began to flicker. Then they snapped off. The operators were plunged into darkness on one side of the room. On the other, dim emergency lighting cast a glow over the stunned crew. An eerie silence settled.
A nuclear reactor control room looks like something out of a classic Cold War movie. Most reactors were built long before fast, compact computers were available, and Fukushima Daiichi, which was commissioned in the 1970s, was no exception. They are complicated contraptions with hundreds of parts, every one of which feeds back through wires to the control room. Gauges and switches that crowd the control room walls constantly report on the state of the reactor. From there, the team of operators can track every tiny detail of the reactor’s function. On a usual day, operators are making sure that the chain reaction clips along at just the right rate for power production. On March 11, they needed those instruments to prevent a meltdown.
The unit 3 control room at Fukushima Daiichi in September of 2010, about six months before the tsunami.
* * *
The three active reactors at the plant had successfully scrammed. But operators knew that the scram is only the first step in reaching what is known as cold shutdown.
When a uranium atom fissions, the two pieces that are left are radioactive. The broken nuclei hold too many neutrons for the new atoms—that’s why neutrons break loose. But they don’t necessarily fly off immediately. The atoms shed them over time. So, even after the control rods have been inserted into the reactor core and the chain reaction has been stopped, atoms in the fuel rods that have already split continue to release energy, creating what is known as decay heat. It can take up to twenty-four hours for a properly functioning nuclear reactor to reach cold shutdown—the point at which the temperature of the core dips below boiling.
Until then, cooling systems need to run or the water in the reactor will quickly boil off, leaving the fuel exposed. Without water to keep the fuel rods cool, temperatures in the core can climb as high as 5,100°F—hotter than a lava flow. Hot enough to melt the fuel rods and allow them to sink right through the bottom of the reactor.
All three of the scrammed reactors had emergency cooling systems. Units 2 and 3 each had a reactor core isolation cooling system, or RCIC (pronounced “Ricksy”). The RCIC system pumps water from storage tanks called suppression pools into the reactor. The pumps are powered by turbines, which are driven by steam coming from the reactor. But unit 1, the oldest reactor in the plant, had an outdated cooling system known as an isolation condenser.
The isolation condenser is what’s known as a passive system—it doesn’t require outside power. All it needs is for two valves to be opened: one at the top, to allow steam from the core into the system, and one at the bottom, to allow cooled water to drip back into the core.
In the minutes following the earthquake, operators had opened the valves. The system was up and running.
* * *
Both the isolation condenser in unit 1 and the RCIC systems in units 2 and 3 should have continued to work without power. But they’d taken a severe beating from a major earthquake. And in the minutes before they were flooded, the RCIC systems had automatically shut down because there was too much water in the reactors. The batteries that supplied power to the instrument panels in unit 3 were still running. But when the second wave swept through, the operators in the control room for units 1 and 2 lost all power. They had no way to tell what was going on. For people who spent their workdays watching the ups and downs of reactor pressure and heat, and making constant adjustments to keep things running safely, it was a worst-case scenario.
In the emergency response center, Superintendent Yoshida knew they were in serious trouble. He immediately notified the company’s headquarters in Tokyo that two of Daiichi’s reactors, 1 and 2, had likely lost their cooling. But Tokyo was 190 miles away. Between the plant and any outside help lay a region that had ju
st been devastated by a massive flood. The reality was that the employees at Fukushima Daiichi were on their own.
To make matters worse, phone lines that should have allowed the emergency response team to communicate with the operators in the control rooms were down. The only way to communicate was in person, so workers had to carry messages back and forth between the emergency response center and the control rooms. Along their route, manhole covers had been blown off by the force of the waves, turning the roads between buildings into an obstacle course of open holes lurking beneath the standing water. What’s more, hundreds of aftershocks continued to rattle Fukushima, threatening to shake broken pieces loose from buildings at any time. And in the hours following the second wave, five more washed into the site. Workers could never be sure when the next wave would come, making the short trip potentially deadly.
Yoshida and his team started brainstorming. They didn’t have a lot to work with. Much of the heavy equipment outside the reactors had been swept away by the waves. A tractor trailer had been pushed across an access door to unit 4. A storage tank blocked the road leading to units 1 through 4. Everything on the bottom floor of the reactor buildings—including the diesel generators they so desperately needed to run the reactors—had been submerged in salt water. And despite hundreds of hours of emergency training, no one really knew what to do next. Using flashlights in the dark, windowless room, they combed through operating manuals. But the possibility of a nuclear reactor losing every source of power at once had never seemed real, so no one had thought through solutions ahead of time.
A photo taken a little over an hour after the tsunami struck Fukushima Daiichi shows the damage outside the unit 1 building.
A team of nuclear engineers and high-level Tokyo Electric Power Company (TEPCO) staff had gathered around a horseshoe of tables laden with computers in the emergency response center. Sitting at the head table and barking orders to the hundreds of workers who had crammed into the space after the quake, Superintendent Yoshida began to take stock of his reactors. It was an enormous task, made much worse by the fact that each of the six reactors had risk factors that could cause unique problems. While units 4, 5, and 6 were shut down before the earthquake and tsunami, those three reactors still held spent fuel in storage pools, and even spent fuel needs to be kept cool. Fortunately, a backup generator in unit 6 had been installed beyond the reach of the flood. They could run a line to that generator to keep unit 5 running, too. Because unit 3 still had battery power, operators were able to restart its RCIC system and knew that water was flowing into the reactor. But units 1 and 2, they feared, could be in danger. They desperately needed to get a better sense of what was happening inside those reactors.
They had a brief glimpse at 4:42 P.M., about an hour after the power went out. Operators managed to get a reading off a water-level indicator inside the reactor in unit 1. Measuring how much water had been lost since the tsunami, they did some calculations and realized that they had less than two hours before the water level in the reactor dropped below the tops of the fuel rods. Once that happened, the rods would begin to melt.
At 5:19, operators went back to unit 1 to try to get another reading off the water meter. When they opened the double door to the building, the Geiger–Müller counter they were carrying to measure radiation maxed out and they had to abandon their task.
In the meantime, workers had been searching for a power source for the control panels. They fanned out across the plant, searching cars and trucks for batteries that hadn’t been damaged by the flood.
It was a little past 6:00 P.M. before they managed to hook up enough batteries to get the first gauges working. Then they made an unwelcome discovery: the two valves that needed to be open to allow cooled water to flow from the isolation condenser back into the reactor of unit 1 were closed. When the power cut out, the system had misread the signal as a steam leak.
The condenser was not working. They needed to come up with a plan to get cooling water into the reactor.
* * *
The reactor vessel, which holds the fuel, control rods, and coolant, lies at the center of a nuclear reactor building. It’s shaped like a capsule and made of 8-inch-thick steel. The reactor is surrounded by a containment structure, a chamber of steel-reinforced concrete as much as 5 feet thick—strong enough to hold the contents of the reactor if the steel vessel fails. Surrounding that is the giant steel building that we see from the outside, known as secondary containment. In an emergency, the containment should trap gases that might escape from the inner chambers, but we rely on the reactor vessel to keep a nuclear reaction—and the radiation it produces—contained. Like a pressure cooker, it is designed to hold together even as extreme amounts of steam build up inside it.
The steel reactor vessel at the heart of each reactor building is surrounded by a thick concrete containment vessel, designed to hold radioactive materials that might escape from the reactor in an emergency.
Think of it this way: When you boil a pot of water, water molecules in the form of steam escape from the pot and drift into your kitchen. If you put a lid over the pot, though, the steam collects underneath. As more and more steam fills the pot, the water molecules become more and more crowded. They bang into each other, and the pressure in the pot rises, causing the lid to dance.
Pressure cookers are designed with heavy locks that hold the lid on the pot, even when an enormous amount of steam has built up beneath it. A reactor vessel is the same. But with its 8-inch-thick walls, it can withstand far more pressure than your pot.
That’s a problem if you’re trying to get more water into the reactor vessel. The steam inside presses against the walls of the vessel with an enormous amount of force. The pressure of the water coming in has to be stronger, or it won’t be able to push its way into the space.
Before workers could add water to reactor 1, they would have to vent the reactor, releasing steam to lower the pressure. That meant opening the closed valves, and for that, they needed power. The control room sent a message to the emergency response center requesting more batteries so they could open the safety valves and vent the extra steam.
But in the chaos of the emergency response center, the request didn’t seem urgent. The emergency response team, frantically running down scenarios on all six reactors, hadn’t gotten the news that the cooling system in unit 1 was down. Workers had reported that they heard hissing sounds from the reactor, and steam was puffing from a vent on the side of the building—both of which seemed to indicate that the isolation condenser was working. The emergency response team believed unit 2 was in greater danger, and the team members had focused their attention there. The control room operators, powerless in the dark, were left waiting.
* * *
By about 9:00 P.M., the water in unit 1 had boiled down to below the tops of the reactor’s fuel rods, and they were most likely already melting. But no one knew the situation was that dire. At 9:19, operators used battery power to read the water level inside the reactor, and the gauge showed that the fuel rods were still covered by about 8 inches of water. It was a far cry from the 20 feet of water that would normally be found above the fuel rods, but it looked as though they were at least covered.
As it turned out, the water-level indicator was malfunctioning. By then, the melting fuel rods had probably already begun to pool at the bottom of the reactor in a molten mass called corium. It was only a matter of time before the molten corium would eat through the reactor vessel and fall to the floor of the concrete containment chamber.
* * *
In the pitch-black of the control rooms, some of the operators wondered what they were doing there. Most of them had families nearby. Some had managed to send messages before the tsunami, but all contact was lost after the power went out. The stress of waiting helplessly wore on them. There were no working toilets, and very little food. A supervisor remembers some of his workers begging to go home. “We were thrown into confusion over the question Is it worth us staying here
helplessly…? Why are we staying here? I pleaded with them to stay.”
Operators read instruments by flashlight in the units 1 and 2 control room on March 23, 2011.
The question grew more pressing around 10:00 P.M., when radiation levels in unit 1 began to climb. Workers were ordered to stay out of the reactor building. Rising radiation could mean only one thing: that radioactive isotopes had somehow gotten out of the reactor vessel.
Although they did not know it then, the steel reactor vessel had been breached, and the radioactive fuel was oozing into the drywell, the open space in the primary containment chamber. Even worse, though, a chemical reaction that was occurring in the high heat of the reactor core was producing highly flammable hydrogen gas.
DAY 2
meltdown
Saturday, March 12, 2011
Meltdown: Earthquake, Tsunami, and Nuclear Disaster in Fukushima Page 4