information about Fukushima published in English in Japanese media info publiée en anglais dans la presse japonaise
15 Décembre 2012
December 15, 2012
The central government's nuclear regulator has released revised simulations on the spread of radioactive substances in the event of a severe nuclear accident to reflect a total of 75 mistakes, significantly altering the results for three plants.
The Nuclear Regulation Authority's secretariat released the results of its reexamination of the simulations Thursday, along with the results of the repeat simulations.
The original simulation for Hokkaido Electric Power Co.'s Tomari nuclear power plant in Tomari, Hokkaido, used rainfall data that was 10 times too large. There were 75 such mistakes found in the initial simulations conducted for all of the nation's 17 nuclear power plants, including Tokyo Electric Power Co.'s Fukushima No. 1 plant.
The nuclear regulatory agency, which serves as the secretariat for the NRA, said it issued strict oral warnings to three senior officials, including Hideka Morimoto, its second in command.
According to the agency, significant corrections were made to the simulations for three of the plants.
These were the Tomari plant and two others where errors had already been found--the Genkai nuclear power plant in Genkai, Saga Prefecture, and Sendai nuclear power plant in Satsumasendai, Kagoshima Prefecture, both operated by Kyushu Electric Power Co.
In the simulations for the Tomari plant, the Japan Nuclear Energy Safety Organization (JNES), an independent administrative organization tasked with the simulations by the NRA, entered the rainfall data in millimeters per hour, instead of the correct 0.1 millimeters per hour.
The repeat simulations showed a smaller area over which cumulative radioactive contamination could reach 100 millisieverts in seven days. Contamination was newly predicted to reach a point that would require the evacuation of residents from Kyowa, which is 15.2 kilometers from the plant, a change from the previous simulation, which indicated evacuations would be needed as far as Kutchan, which is 19.9 kilometers away.
For the Genkai and Sendai plants, Kyushu Electric gave JNES faulty data on wind directions that led to wind forecasts that were 180 degrees opposite of what they should have been.
The recalculations changed the farthest point of contamination from the Genkai plant from Karatsu, Saga Prefecture, which is 27.4 kilometers from the plant, to Sasebo, Nagasaki Prefecture, which is 29.1 kilometers away.
The farthest point of contamination for the Sendai plant was changed from Akune, Kagoshima Prefecture, which is 21.1 kilometers from the plant, to the sea off Ichikikushikino, Kagoshima Prefecture, which is 21.1 kilometers away.
Major changes were not made to the simulations for the other 14 nuclear power plants, but there were many minor mistakes about weather data and conversion codes.
The changes mean that some municipalities now find themselves either excluded from or included in areas where radiation contamination in a major accident could require evacuation.
The agency blamed JNES for the mistakes and said it lacked the capability to verify the results.
The agency said it sees the simulations as basic data on which to base the disaster-preparedness plans that are compiled upon requests from municipalities near nuclear power plants.
In August this year, the Nuclear and Industrial Safety Agency, one of predecessors of the NRA, changed its policy to require forecast data in all 16 compass directions, which was not initially planned.
Despite the enormous volume of data to be processed, the agency, after taking over from NISA, did not check the system before asking JNES to quickly complete the simulations.
The simulations produced maps predicting how radioactive substances would spread in the event of a nuclear accident for all of Japan's nuclear plants.
They indicated points in 16 directions where the accumulated radiation exposure levels could reach 100 millisieverts in seven days, based on evaluation guidelines from the International Atomic Energy Agency.
Although the simulations are to be used by local governments near nuclear plants to compile disaster plans, they do not take geographical data into account.
===
Trust of residents lost
By Yuki Inamura
Yomiuri Shimbun Staff Writer
People who live near nuclear power plants will likely be interested in the new simulation results, believing they contain important information that could affect their lives.
Even if only to be used for reference, the simulations were probably not seen by local residents as merely numerical data. With the repeated revisions, it is natural that trust has been damaged.
The NRA was established to better ensure safety by converting the old nuclear authorities, which were criticized for being mere rubber stamps for the industry, into a more trustworthy regulator.
However, NRA Chairman Shunichi Tanaka on Thursday talked about the simulation mistakes as if they were someone else's problem. "It was our first failure and we learned some good lessons by reflecting on what happened," he said.
The root of the problem may be in that the simulations were ordered by NISA and thus a defunct agency could be held responsible.
However, this incident seemed to show that the NRA has not yet settled into its role.
The recently established regulator has many important tasks, such as creating safety guidelines for nuclear power plants and conducting research on faults in and around the plants.
The NRA must show it can act in the public interest so it can regain the trust of the people.