A criticality accident occurred on September 30, 1999, in a conversion test facility at the JCO Tokai site. The accident was triggered by pouring an 18.8% enriched uranyl nitrate solution into a precipitation vessel beyond the critical mass. The accident continued for about 19 hours before the criticality could be stopped, during which time neutrons and gamma-rays were emitted continuously due to fission reactions. The total number of fission reactions was 2.5×10¹⁸, which was estimated by an activity analysis of the fission products in the solution of the precipitation vessel. The accident gave serious radiation dose to 3 employees and fatal dose to 2 of them. Neutrons and gamma-rays emitted by the accident caused meaningful doses to the residents of the surrounding area of JCO. The dominant dose to the residents and JCO employees was brought by neutrons and gamma-rays from the precipitation vessel, while the contribution of radioactive plume was negligible. The individual dose was estimated for 234 resident, 169 JCO employees and 260 emergency personnel. The maximum doses were 21mSv for the residents, 48mSv for the JCO employees, and 9.4mSv for the emergency personnel, respectively. No deterministic effect, however, has been observed, except for the 3 workers.

View full abstract

The uranium solution in the precipitation tank in the JCO's uranium conversion facility was analyzed in order to evaluate the total number of fissions in the criticality accident. Two analytical groups at JAERI performed chemical analyses independently in order to check the validity of the results: the concentration of the fission products (⁹⁵Zr, ⁹⁹Mo, ¹⁰³Ru, ¹³¹I, ¹⁴⁰Ba, etc), uranium, boron and impurity elements in the solution. The analytical results obtained by the two groups were almost in agreement within the analytical error. The number of fissions per one gram of uranium in the accident was determined to be (1.5±0.1)×10 ¹⁴. Also, the total number of events was evaluated to be (2.5±0.1)×10¹⁸ fissions using the total amount of uranium (16.6 kg) fed into the precipitation tank at the accident.

View full abstract

A few days after the JCO criticality accident in Tokai-mura, a collaborating scientific investigation group was organized to evaluate the environmental impact of the accident. The group consisted of two groups: an environmental research group (radiochemistry group) and a biological research group. This paper overviews the scientific activity of the former group based on 6 sampling campaigns conducted at the JCO campus, Tokai-mura and Naka-machi. Some of the topical results and our remaining tasks concerning the JCO accident are discussed.

View full abstract

A transport calculation of the neutrons leaked to the environment by the JCO criticality accident was carried out based on three-dimensional geometrical models of the buildings within the JCO territory. Our work started from an initial step to simulate the leakage process of neutrons from the precipitation tank, and proceeded to a step to calculate the neutron propagation throughout the JCO facilities. The total fission number during the accident in the precipitation tank was evaluated to be 2.5×10 ¹⁸ by comparing the calculated neutron-induced activities per ²³⁵U fission with the measured values in a stainless-steel net sample taken 2 m from the precipitation tank. Shield effects by various structures within the JCO facilities were evaluated by comparing the present results with a previous calculation using two-dimensional models which suppose a point source of the fission spectrum in the air above the ground without any shield structures. The shield effect by the precipitation tank, itself, was obtained to be a factor of 3. The shield factor by the conversion building varied between 1.1 and 2, depending on the direction from the building. The shield effect by the surrounding buildings within the JCO territory was between 1 and 5, also depending on the direction.

View full abstract

A criticality accident occurred at a uranium conversion facility in Tokai-mura, Japan on September 30, 1999, and fission neutrons were continuously emitted for about 20 hours. Materials of stainless steel or iron, and chemical reagents were collected at places between 2 m and 270 m from the criticality accident site on October 25 and 26, 1999, November 27, 1999 and February 11, 2000. Neutron-induced radionuclides, such as ⁵⁴Mn and ⁵⁸Co, in the materials exposed to fast neutrons from the accident were measured to estimate the neutron fluences and energy distributions. Highly sensitive γ-ray spectrometry with a well-type Ge detector was performed after radiochemical separation of Mn and Co from the materials. An instrumental neutron activation analysis was mainly applied for determinations of the target elements and chemical yields. The concentrations of ⁵⁴Mn and ⁵⁸Co in a mesh screen of stainless steel collected at a location 2.0 m from the accident site were determined. The total number of fission events was evaluated to be 2.5×10 ¹⁸ by Monte-Carlo calculations of neutron transfer by considering the observed values of ⁵⁴Mn and ⁵⁸Co. The results presented here are fundamental to estimate the neutron doses at various distances.

View full abstract

Specific activities (Bq/g-element) of residual neutron-induced radionuclides by the JCO criticality accident were measured for soil, concrete block and chemical reagent samples collected in the JCO campus. Induced radionuclides such as ²⁴Na, ⁴⁶Sc, ⁵⁴Mn, ⁵⁹Fe, ⁶⁰Co, ⁶⁵Zn, ⁸²Br, ¹²²Sb, ¹³⁴Cs and ¹⁴⁰La were detected in the samples, depending on the ground distance from the accident point and the sampling date. Apparent thermal, epi-thermal and fast neutron fluences, which reached the sample at each point, were roughly estimated from the specific activities and cross sections of the target nuclides taken from a literature. The present data are believed to be important as validation data for a three-dimensional neutron transport model calculation.

View full abstract

The validity of a method for individual dose reconstruction based on the anisotropic radiation distribution was confirmed by a comparison with the dose from Na-24 whole-body counting for seven persons at a neighboring company located on the western side of the JCO campus. The successful coincidence between D_i (Na-24) and D_i (present) also supports the validity of the second version of D(r) reported by the Head Office of Countermeasure. The present dose reconstruction for the 350 m zone at the western side showed an average value of 0.7 mSv and a maximal value of 3.1 mSv, as indoor-dose under the assumption of an effective transmittance of 0.4 for all of the houses. If all of the residents in 350 m zone were indoors during the accident, 83% of them might have received external doses of less than 1 mSv. The radiation exposure to the nearest residential area in the southern-west direction was significantly reduced with Φ(θ_i) between 0.4 and 0.2 by several buildings on the JCO campus. The present study on the public dose confirms that the official report on the public dose (the maximal value of 21 mSv for individual doses) from Head Office of Countermeasure is significantly overestimated due to their isotropic treatment of radiation from the source.

View full abstract

The NIRS staff interviewed the residents in the evacuated zone around the JCO facility in Tokai-mura on 19 and 20 November, 1999, to obtain the following parameters every 30 minutes starting from 10:35 A.M. on 30 September to 6:15 A.M. on 1 October: the distance from the precipitation tank, the type of the house, positions in the house, wall materials and their thickness in order to estimate individual doses due to the accident. The ambient dose equivalents were obtained based on monitoring data during the accident. In addition, computer calculations were conducted to evaluate the conversion factor from ambient dose equivalent to effective dose equivalent as well as the shielding effect of the house or factory to estimate the effective dose equivalent to the residents. The estimated individual doses based on the behavior survey were in the range from zero to 21mSv. The individual doses were reported to the residents during the second visit to each house and factory at the end of January, 2000.

View full abstract

²⁴Na in the human body, activated by neutrons emitted at the JCO criticality accident, was observed for 62 subjects, where 148 subjects were measured by the whole body counter of JNC Tokai Works. The 148 subjects, including JCO employees and the contractors, residents neighboring the site and emergency service officers, were measured by the whole-body counter. The neutron-energy spectrum around the facility was calculated using neutron transport codes (ANISN and MCNP), and the relation between an amount of activated sodium in human body and neutron dose was evaluated from the calculated neutron energy spectrum and theoretical neutron capture probability by the human body. The maximum ²⁴Na activity in the body was 7.7 kBq (83 Bq(²⁴Na)/g(²³Na)) and the relevant effective dose equivalent was 47 mSv.

View full abstract

Chromosome aberrations were studied in peripheral blood lymphocytes of 43 persons who were exposed to low-level radiation of mixed neutrons and γ-rays resulting from the JCO criticality accident. When the age-adjusted frequencies of dicentric and ring chromosomes were compared with the dose calibration curve established in vitro for ⁶⁰Co γ-rays as a reference radiation, a significant correlation was observed between the chromosomally estimated doses and the documented doses evaluated by physical means. The regression coefficient of the chromosomal doses against the documented doses, 1.47±0.33, indicates that the relative biological effectiveness of fission neutrons at low doses is considerably higher than that currently adopted in the radiation protection standard.

View full abstract

In the Tokai-mura criticality accident, three workers were heavily exposed. Biological materials, such as blood, urine, vomit and hair, were collected from the workers and analyzed for radioactivities, produced by the neutron irradiation. Activation products, such as ²⁴Na, ⁴²K and ⁸²Br, were found in these materials by gamma-ray spectrometry. The radionuclide of the highest activity observed in biological materials was ²⁴Na, e.g. the concentrations of this nuclide in the blood samples from the three patients at the accident time were 169, 92 and 23 Bq/ml, respectively. The concentrations of stable sodium in the same samples were determined by ICP-AES to obtain specific activities of ²⁴Na (concentration ratio between the produced ²⁴Na and stable ²³Na), which are essential for estimating the neutron fluences and radiation doses. The specific activities of ²⁴Na obtained for the three patients through the blood analysis were 8.2×10⁴, 4.3 ×10⁴ and 1.2×10⁴ Bq ²⁴Na/g²³Na. Based on these values, individual's neutron fluences were estimated to be 5.7×10 ¹¹, 3.0×10¹¹ and 0.85×10¹¹ cm^-2, respectively.

View full abstract

The measurement of β-emitters in biological samples (hair, urine and bone) from three patients in the JCO criticality accident was performed to assess the neutron dose to individuals. The result of the measurements of ³²P in hair and urine collected immediately after the accident showed that sufficient ³²P activities had been induced in the hair by fast neutrons and in the urine by thermal neutrons to know the severity of the exposure to the individuals and to the position. From the measurement of ³²P and ⁴⁵Ca in bone from various anatomical parts of two patients who died 82 and 210 days after the accident, it was suggested that the distribution of the induced β-emitters activities could prove the position and posture of the patients at the moment of exposure.

View full abstract

The doses for the overexposed patients were estimated by the measurement result of specific activity of ²⁴Na in blood. The present method is almost based on documents of the International Atomic Energy Agency (IAEA) and the Oak Ridge National Laboratory. The neutron energy spectrum obtained using the ANISN code (Multigroup One-Dimensional Discrete Ordinates Transport Code System with Anisotropic Scattering) was assumed. The values in ICRP Publication 74 were applied for the doses in each organ per unit neutron fluence. Gamma-ray dose was indirectly estimated based on (a) the result of environmental monitoring around the accident site and (b) a graph in IAEA manual, which gives the kerma ratio of neutrons and γ-rays as a function of the critical volume or the atomic ratio of hydrogen to ²³⁵U. The estimated neutron doses were 5.4 Gy for patient A, 2.9 Gy for patient B and 0.81 Gy for patient C. The estimated γ-ray doses were 8.5 or 13 Gy for patient A, 4.5 or 6.9 Gy for patient B, and 1.3 or 2.0 Gy for patient C.

View full abstract

A dose estimation by chromosome analysis was performed on the 3 severely exposed patients in the Tokai-mura criticality accident. Drastically reduced lymphocyte counts suggested that the whole-body dose of radiation which they had been exposed to was unprecedentedly high. Because the number of lymphocytes in the white blood cells in two patients was very low, we could not culture and harvest cells by the conventional method. To collect the number of lymphocytes necessary for chromosome preparation, we processed blood samples by a modified method, called the high-yield chromosome preparation method. With this technique, we could culture and harvest cells, and then make air-dried chromosome slides. We applied a new dose-estimation method involving an artificially induced prematurely condensed ring chromosome, the PCC-ring method, to estimate an unusually high dose with a short time. The estimated doses by the PCC-ring method were in fairly good accordance with those by the conventional dicentric and ring chromosome (Dic+R) method. The biologically estimated dose was comparable with that estimated by a physical method. As far as we know, the estimated dose of the most severely exposed patient in the present study is the highest recorded among that chromosome analyses have been able to estimate in humans.

View full abstract

A criticality accident occurred on September 30, 1999, at the uranium conversion plant in Tokai-mura (Tokai-village), Ibaraki Prefecture, Japan. When the criticality occurred, three workers saw a "blue-white glow," and a radiation monitor alarm was sounded. They were severely exposed to neutron and γ-ray irradiation, and subsequently developed acute radiation syndrome (ARS). One worker reported vomiting within minutes and loss of consciousness for 10-20 seconds. This worker also had diarrhea an hour after the exposure. The other worker started to vomit almost an hour after the exposure. The three workers, including their supervisor, who had no symptoms at the time, were brought to the National Mito Hospital by ambulance. Because of the detection of γ-rays from their body surface by preliminary surveys and decreased numbers of lymphocytes in peripheral blood, they were transferred to the National Institute of Radiological Sciences (NIRS), which has been designated as a hospital responsible for radiation emergencies. Dose estimations for the three workers were performed by prodromal symptoms, serial changes of lymphocyte numbers, chromosomal analysis, and ²⁴Na activity. The results obtained from these methods were fairly consistent. Most of the data, such as the dose rate of radiation, its distribution, and the quality needed to evaluate the average dose, were not available when the decision for hematopoitic stem cell transplantation had to be made. Therefore, prodromal symptoms may be important in making decisions for therapeutic strategies, such as stem-cell transplantation in heavily exposed victims.

View full abstract

Patient A who was exposed to a critical dose of radiation developed skin lesions throughout the body surface, gastrointestinal disorder with massive diarrhea and prominent bleeding, which caused severe loss in body fluids. Gastrointestinal bleeding due to the deteriorated intestinal mucosa was considered to be one of the major causes of death, although infection did not develop, possibly because of SDD and aseptic intensive care, until terminal stages. Patient A ultimately developed respiratory and renal failure in addition to skin exudate and gastrointestinal bleeding, and died of multiple organ failure on the 83rd day after exposure. The extreme unevenness of the dose distribution and the neutron versus γ-ray component made the clinical manifestation very complicated. Initially, the mean absorbed dose was calculated as 16-20 GyEq for Patient A, mainly based on neutron-activated ²⁴Na in the blood. However, a very recent calculation showed that the absorbed skin dose was highest at the upper-right abdomen reaching 61.8 Gy (27.0 as neutron plus 34.8 Gy as γ-ray). The dorsal side was calculated to have received one eighth of the value of the abdominal side, and much smaller neutron component. His absorbed-dose distribution throughout the body was very inhomogeneous because of the closeness of the standing point to the mixing tank. Despite prolonged survival because of intensive care with massive fluids and blood transfusion, peripheral blood stem-cell transplantation, cultured skin-cell grafts, and the administration of cytokines for marrow, the patient was not saved. Restoration of the bone marrow function, prevention of skin fibrosis, radiation lung damage, and repair of gastrointestinal mucosa, and final recovery of the patient were elusive. Abundant personnel and resources were also a prerequisite to allow for the comprehensive and collective intensive care. A further understanding of the effects of high-dose radiation as well as the basic and clinical development of regeneration medicine are important issues for the future.

View full abstract