Tokaimura: Japan's forgotten nuclear disaster

There are episodes in nuclear history that serve as terrible reminders of the tremendous power and volatility contained within atomic nuclei. One such reminder is the 1999 Tokaimura criticality accident.

In fiscal 1999, Japan Nuclear Fuel Conversion Co, (JCO) was contracted to manufacture 160 liters of uranyl nitrate solution with an enrichment of 18.8% for use in the Jōyō fast breeder reactor located in Ibaraki prefecture. The reactor is operated by the Japan Atomic Energy Agency, and is a research reactor where experiments in nuclear energy can be carried out.

In September, the factory carried out the uranium refining work, finishing on September 28th. But the solution was not uniformly distributed, and a procedure to homogenize the mixture was scheduled for the next day. On the 29th, the mixing process was successfully completed, but rather than use the official manual called for the use of a "mixing tower" where small amounts of uranium fuel would be mixed with nitric acid to dissolve impurities which could then be decanted out in a specially designed buffer tank. Then in a precipitation tank, ammonia is added to the liquid to capture any remaining waste. The tower and tank process was designed to prevent the possibility of an accidental criticality, but it was quite slow as the throughput was minimal.

The company had created a secret alternative procedure, which was faster, but was also less safe. Ouchi Hisashi, Shinohara Masato, and Yokokawa Yutaka were working on the night of September 30, 1999, following the company manual, which was not approved by the Science and Technology Agency. But the three men did not know this. The alternate procedure skipped the slow mixing in the tower in favor of direct mixing in stainless steel buckets. JCO was completing an order for fuel for the breeder reactor for the first time in 3 years, and the process was taking longer as everyone had forgotten the procedures.

The workers, in an attempt to make up time and meet their shipping deadlines poured the mixture into the precipitation tank, exceeding even the upper limits of the shadow manual. Ouchi was leaning over the tank to make sure the mixture didn't spill. Shinohara was standing nearby, helping pour the solution, while Yokokawa was sitting at a desk four meters away. The workers wore no protective gear for radiation, and the building was effectively a corrugated steel warehouse. The mixing process was exothermic, so the bucket was surrounded by cooling water to prevent overheating.

At 10:35 am, the three workers saw the contents of the bucket produce a bright blue glow. This is Cherenkov radiation, which is what happens when electrically charged particles, like protons or electrons, pass through a clear medium like water at a speed greater than light. The precipitation tank had reached critical mass and the cooling water had acted as a neutron reflector to sustain the criticality. Alarms sounded, and Ouchi and Shinohara immediately felt pain, nausea, and difficulty breathing. All three evacuated the building, but were unsure what they should do next. A worker from the building next door found them and called for an ambulance.

JCO notified the Science and Technology Agency of a possible criticality incident at 11:15 am. The reaction was in full swing at this time, and completely unshielded. There was no procedure for this type of accident, so the fire department responding was not informed, resulting in their exposure. There were no external monitoring stations installed near the plant, so it was 1 hour and 15 minutes before nearby residents were alerted to the crisis, and at that time they were told to shelter in place.

400 meters north of the factory was the Tokai parking area on the Joban Expressway, filled with motorists driving north to view the autumn leaves. They were unaware of the danger, and the express was yet to be shut down. Slowly evacuation orders were issued. Initially 40 households within 350 meters of the plant were ordered to evacuate, and households within 500 meters were advised to evacuate (but not compelled). Households outside that radius but within 10 km were advised to shelter in place. By approximately 8:30 pm, the highways and train lines passing near the plant were suspended.

The efforts to stop the criticality were initially hampered by the fact that the radiation levels inside the building were too high for it to be approached safely. One of the ideas considered was having the Self Defense Forces shoot the tank to drain the water, but fortunately the option of firing rounds at an unshielded critical reaction was rejected.

18 JCO employees volunteered to work in shifts to cut the cooling water pipes and stop the reaction. The reasoning was given "we must handle accidents caused by ourselves." The first team entered to discover that the setup did not match the schematics that JCO had on file. In the diagram, the settling tank was positioned 1 meter from the wall, offering plenty of room for them to work on cutting the cooling pipes which were creating the neutron reflector, but when they arrived, they found that the tank was actually only 30 cm from the wall, making it extremely difficult to get tools into the space. The first team came out after 3 minutes, which was what they had estimated would be a "safe" amount of time to be exposed. However when they returned, their dosimeters revealed that they had received twice the anticipated radiation, subsequent teams were shortened to only 1 minute on site.

The successive teams were able to cut the cooling pipes, drain the water, inject argon gas, followed by pouring boric acid directly into the tank. By 6:30 in the morning, 20 hours after first criticality, the chain reaction was stopped.

The workers who had caused the criticality were rushed to Mito Hospital before being helicoptered to Tokyo for treatment of acute radiation syndrome. Yokokawa, sitting at a desk four meters from the accident received between 1 and 4.5 grays, suffered a drop of white cells to zero, but eventually recovered. Shinohara, who had received between 6 and 10 grays suffered chromosome destruction, which meant his body's cells lost the ability to replicate themselves. He was given a stem cell transplant which was temporarily effective so that he was able to give testimony to police, but relapsed and died from MRSA-caused pneumonia on April 27, 2000, 211 days after exposure. Ouchi, who had been leaning over the bucket to guide the mixing process, received 16 to 20 grays which caused chromosomal destruction. He was moved to sterile containment, and received a donation of stem cells from his sister, which helped temporarily. His skin cells were unable to reproduce themselves and his fluids constantly leaked out of his body. On December 21, he died of multiple organ failure.

In the wake of the disaster, JCO's license to reprocess nuclear fuel was revoked. Its parent company, Sumitomo Metal Mining, was driven to the brink of bankruptcy. The Nuclear Safety Commission implemented a whistleblower system, and the Japanese government passed a law to strengthen compliance with safety regulations and increased the frequency and scope of periodic inspections. The Japanese Ground Self Defense Forces amended its disaster response to include nuclear disaster relief, a system which was put to the test in 2011.

All major newspapers covered the story on their front page on October 1st, with the exception of the Chunichi Shimbun. The Chunichi Dragons baseball team had won their first Central League championship in 11 years the night before, and the baseball news took precedent in that paper (39 pages containing baseball stories vs. 38 pages containing stories about the accident). As a result, there is a belief among Japanese baseball fans that Chunichi winning the league is a harbinger of doom.