JCO Criticality Accident:
The Victims and the Final Report
Victim of Tokai
Hisashi Ohuchi, one of the three seriously exposed JCO employees in the criticality accident at Tokai on September 30, died at 11:21 p.m. on December 21 at the University of Tokyo Hospital. He was only 35. Mr. Ohuchi was the first casuality from acute radiation injury since the dropping of the A-bombs at Hiroshima, Nagasaki and the Bikini tests in which a Japanese crew member of 'Daigo Fukuryumaru' died. It is the first time in the history of nuclear power development in Japan that a life was lost due to an accident, and it is therefore a serious blow to the industry.
Mr. Ohuchi was exposed to 16~20 Sv (equivalent) and died from multiple organ failure. His face, arms, and torso were burnt by radiation. His lymphocyte count dropped to zero, and white blood cells were drastically reduced. Due to the damage to his marrow, affecting his ability to produce blood cells, he underwent a transfusion of peripheral stem cells which were taken from the peripheral blood of his brother on October 6 and 7. His burnt-off skin could not regenerate. His intestines were continually bleeding, and he was given more than 10 L of blood and infusion solution every day. By mid-November the functioning of his liver and kidney had deteriorated and breathing became difficult. Drugs to boost blood pressure were administered after his heart failed temporarily on November 27. His heart was barely functioning by December 18. Death came 83 days after he was exposed. As for the two other employees, Mr. Masato Shinohara, who was exposed to 6~10 Sv, is still receiving treatment at the Institute of Medical Science of the University of Tokyo. Mr. Yutaka Yokogawa, who was exposed to 1~ 4.5 Sv has reached a stable condition for the time being and left the National Institute of Radiological Sciences on December 20.
Final Report Released by the JCO Accident Investigation Committee
Three days after the death of Mr. Ohuchi, the eleventh meeting of the JCO Criticality Accident Investigation Committee was held and the final report on their investigation was released. Almost no criticisms or remarks were made about the final report by the committee members at the last meeting, and it was clear that they had already settled the matter amongst themselves. Although the report is 170 pages long, many of its contradictory claims are left unresolved and it can hardly be called a product of an exhaustive debate. The report was put together in a mere three months, and it is hard to avoid the impression that it was patched up in great haste. It is quite clear that there was an implicit deadline. Following is an examination of this hastily-produced and extremely problematic document.
First of all, the report's analysis of the accident is not based on sound scientific analysis. The estimation of the total number of fissions which occurred during the criticality reaction has not been changed from the estimation reported in the interim report, that is 2.5x1018 . The estimated number corresponds to the fission of about 1mg of uranium 235. However, the estimated contributions to the cumulative radiation dose from the initial burst and the succeeding plateau of the criticality reaction given in the interim report -48% and 52% respectively- were drastically altered to 11.4% and 88.6%. (Fission reaction consists of an initial burst and a prolonged plateau.) The new figures derive from statistical analysis of data from the neutron monitor at the Japan Atomic Energy Research Institute (JAERI)-Naka Research Institute. Then the radiation dose received by local residents was estimated using the neutron monitor data during the plateau period rather than calculating directly from the total number of fissions. However, this estimate can vary dramatically according to how the background count (naturally existing neutron dose from cosmic rays) is estimated and how the precision of the monitors is evaluated. Therefore, the results of the report still contain significant uncertainties and it looks as though the uncertainties were actually used to the benefit of the Committee to minimize the estimated residents' exposure.
Secondly, the report does not address the responsibilities of the Japan Nuclear Cycle Development Institute (JNC), which ordered JCO to provide them with homogenized uranium in the form of nitrate solution. The accident occurred because JCO employees were handling uranium enriched to a relatively high concentration of 18.8% in an amount far above the minimum critical mass. This came about because JNC (then known as PNC or Donen), responding to strong concerns from the US about the danger of the proliferation of nuclear materials, had ordered JCO to supply uranium solution for manufacturing mixed uranium-plutonium oxide (MOX) fuel for the Joyo Fast Breeder Reactor. Originally, plutonium and uranium were prepared separately as powder and then mixed together to manufacture MOX fuel. However, to decrease the risk of plutonium theft/diversion, JNC had to switch to a method where uranium and plutonium are mixed together in the form of solution before providing mixed uranium and plutonium powder. When providing uranium powder to JNC, JCO refined the imported uranium powder to purify the uranium by dissolving the powder and then making it into powder again. However, to provide JNC with uranium solution, JCO had to re-dissolve the refined uranium powder. JCO did not equip itself with additional installations for this new operation and thus had to come up with ways to do the job using the existing process line. In addition to ordering JCO to provide them with uranium solution, JNC demanded that the solution be completely homogenized. In order to homogenize the solution, JCO began a process called 'cross-blending.' In this method, 40 L of uranium solution was divided into ten 4 L stainless steel bottles. Then 1/10 of the solution from each bottle was put into a separate set of 10 bottles. This method was eventually abandoned due to its troublesome nature, and the company began to combine the solution in the buffer column before transferring the solution into the 4 L bottles. The buffer column was designed to prevent criticality and thus JCO employees were able to homogenize the solution without causing an accident. However, for the process that led to the accident, the company used the precipitation tank, which was not designed to prevent criticality, to homogenize the solution. According to one of the three employees directly involved with the process, they had used the precipitation tank to cut time and to simplify the process since the structure of the buffer column made it difficult to transfer the solution from the column to the 4 L bottles. JCO employees did not take the time to see if using the precipitation tank was definitely safe. Our analysis indicates that for uranium powder re-dissolution to prepare the required 370g/ÜPuranyl nitrate solution, JNC or JCO should have constructed a separate re-dissolution and homogenizing facility, which they chose not to do. In view of the fact that JNC is finally responsible for the whole MOX fabrication procedure in a proliferation-resistant manner, JNC, not JCO, should have provided the uranium solution. If JCO was commissioned for the job, as was the actual case, JNC should have supervised the process. Instead of taking responsible measures, JNC had even rushed the delivery date for JCO to provide them with the uranium solution. CNIC believes that the unreasonable demands of JNC had a very significant part to play in the circumstances leading to the events of September 30. Why is it, then, that the extent of JNC's responsibility is not examined at all in the final report? Could this have something to do with the fact that two members of the investigation committee are JNC employees? A final report lacking a solid scientific analysis and a thorough examination of the events leading to the accident is far from convincing. The final report presents no less than 103 countermeasure plans and suggestions. However, the greater the number of suggestions, the greater are our concerns about accidents of a similar scale - or worse - happening in the future.