Fukushima Daiichi Diary

By PHRED DVORAK And YUKA HAYASHI

A Wall Street Journal examination of the first 24 hours after the Fukushima Daiichi accident shows that disaster piled on disaster, worsening the nuclear crisis faster than anyone had initially thought could happen. Here's a more detailed look at some of the other problems that the stricken plant, regulators and operator Tokyo Electric Power Co. were grappling with.

The Shut-Off Cooling System

Documents released by Tepco Monday showed the isolation condenser— an emergency cooling system installed on Reactor No. 1 before the quake as a final resort in case of a total loss of power—worked only sporadically, if at all. Tepco officials explained that somebody appears to have manually closed the valves on the condenser soon after the March 11 quake—but before the tsunami hit about an hour later—to control the fluctuating pressure inside the reactor. Reopening the valves required battery power, so those valves likely couldn't be opened because the tsunami damaged the backup batteries.

If the valves hadn't been shut, things might have turned out differently. Temperatures in the reactor climbed faster than initially expected, causing more and faster damage. Tepco admitted this week the problems at Reactor 1 were far worse than originally thought. Its new projection shows fuel may have started melting rapidly only five hours after the March 11 quake. By 6:50 a.m. March 12, the fuel was likely in a heap at the bottom of the vessel.

The Offline Off-Site Center

The government's emergency response was to be coordinated out of an off-site center run by regulator Nuclear and Industrial Safety Agency, a 15-minute drive from the Fukushima Daiichi plant. But when Kazuma Yokota, the regulator charged with setting up the center, arrived at the site after the earthquake, he found that phone lines and cellphones were down, the satellite phone wasn't working and the fuel pump supplying the backup power generator was broken. That effectively cut off the office from the outside world during key early hours.

Tepco's emergency-communications systems were still up: Fukushima Daiichi's earthquake shelter was far enough inland to have escaped the tsunami flooding, so its generator powered a video link with Tokyo as well as a special phone system internal to Tepco. But with NISA's off-site center offline, the government was left to shepherd the crisis from Tokyo and depend on Tepco headquarters for its information—something that was to cause crossed signals and mounting frustration later. Mr. Yokota himself had to send staff back to the plant in order to keep tabs on what was going on; the NISA generator wasn't fixed until around 2 a.m. March 12, and it ran out of fuel within a day.

The False Positive at Reactor No. 2

At 8:35 p.m. on March 11, Tepco reported to the government what it thought was the first sign of serious trouble at Fukushima Daiichi: the final emergency cooling system at Reactor No. 2 appeared to have stopped. Experts at NISA threw together a projection showing that fuel in the reactor could start melting by around 1 a.m. By 9:23 p.m., the national government was ordering the evacuation of people within two miles of the plant. The Fukushima prefectural government, unwilling to wait so long, had already issued its own order about a half hour earlier.

In fact, as is now known, the real problem was at Reactor 1. Toward midnight, Fukushima Daiichi engineers were reporting that the cooling system at Reactor 2 had restarted, but had stopped at Reactor 1. Tepco recently admitted Reactor 1's fuel rods were likely exposed by around that time.

Lost in Transit

According to company protocols, the decision to vent radioactive gas from a reactor had to be made by Tepco's president, Masataka Shimizu. Around the time Tepco's engineers and the government were coming to that decision, though, Mr. Shimizu was stuck in the city of Nagoya, about 165 miles west of Tokyo, after trying unsuccessfully to get back to headquarters following the quake. Around 9:30 p.m., Mr. Shimizu had asked for a military transport flight back to the capital only to be denied permission by the Minister of Defense. Officials said the aircraft was needed instead to fly supplies to the tsunami zone. At 12:13 a.m.—around the time Fukushima Daiichi engineers were getting their alarming reactor-pressure reading at Reactor 1—the plane landed again in Nagoya.

Polluting the Air

In a sign of how far-fetched a crisis of this magnitude was considered, Tepco didn't take the extra step of installing a filter on its emergency vent pipe to scrub out radioactive particles. Indeed, the whole vent installation was considered voluntary, since Japanese regulators didn't think nuclear reactors would ever have to deal with the high pressures the vent system was designed to withstand. That meant venting would be accompanied by the release of significant amounts of radioactivity in the air. Before the venting, the government expanded the evacuation zone to six miles from the two miles set the night before. Tepco attributes much of its venting delay to concerns about the evacuation.

"As someone who knows a little bit about nuclear energy, I knew how urgent it was to vent," said Masashi Katayose, a 57-year-old engineer in the Fukushima prefectural government's nuclear safety section. "But as local officials, we can't say, 'go ahead and contaminate our air with radiation.'" When Mr. Katayose and others broke the news of the venting to the governor, he just nodded.

Pump Problems

Tepco workers wanted to make sure they could inject more water into the reactor as they vented. The venting process could aggravate the cooling problems by further lowering water levels in the reactor, because venting releases water in the form of steam and because water boils at higher temperatures when under higher pressure. But that faced obstacles, too, Tepco officials said. At least one of the fire trucks that would normally have been used to pump water into the reactor had been washed away by the tsunami. When the ground crew did get fire hoses hooked up to Reactor 1, they had trouble getting the water in, Tepco officials said.

Tepco didn't manage to start pumping fresh water into the reactor until 5:46 a.m. on March 12. Workers called off the operation around 3 p.m.—half an hour before the hydrogen explosion—after injecting around 80 tons, Tepco documents released Monday showed.

Taking Turns

Radiation levels at Reactor 1 were so high by the time workers went in to open vent valves manually that they had to go in short shifts to prevent high exposure. The Reactor 1 shift manager's stint exposed him to 100 times the radiation an average person gets in a year. Then, several workers followed him one by one, relay style, to turn the crank little by little. At 9:15 a.m. Saturday , the valve was open a quarter of the way, Tepco documents released Monday say.

Radiation levels delayed attempts to open the second, air-operated valve as well, Tepco officials said. They went in short shifts, carrying a portable air compressor and a power source into reactor building and hooking it up in the darkness.

A total of 18 workers were exposed to radiation that day, NISA documents say, though none were reported to have suffered any health problems. The shift manager reported a headache and was seen by doctors, but he was later sent home.