Major problems have arisen with the cooling system of the vitrified high-level waste storage buildings at the Rokkasho nuclear fuel cycle facilities in Aomori Prefecture.
When complete, there will be a total of five storage buildings for vitrified high-level waste at the Rokkasho. Of these, only one, 'Building A', is currently operational. It holds vitrified high-level waste canisters that have been returned after reprocessing of spent fuel overseas. It commenced operations in April 1995 and has the capacity to hold 1,440 canisters. There are already 892 canisters in storage there. However, the total number of canisters to be returned after reprocessing in France and England is 2,200, so an application was submitted to construct a new building (Building B) to hold the excess. Aside from these two buildings, there are also three buildings for the high-level waste generated at the Rokkasho Reprocessing Plant itself.
The safety inspection for building B was completed in 2003. However, when the detailed design was checked, it became apparent that there was a possibility that the cooling capacity was inadequate. The owner of the facility, Japan Nuclear Fuel Ltd. (JNFL), set a benchmark of 500 degrees C for the canisters' maximum internal temperature , but the Nuclear and Industrial Safety Agency's (NISA) analysis indicates that the temperature could exceed 600 degrees C.
Design changed to reduce costs
Please refer to the following diagram when reading this account.
When the glass canisters arrive they are removed from the containers in which they were shipped. They are then placed vertically in containers in an underground storage pit. Each of these containers holds nine glass canisters. The facility is cooled by natural air flow. Air enters from the inlet shaft, moves along the bottom of the storage pit, goes around the outside of the containers and passes out of the top of the pit. It passes through a filter as it leaves the outlet shaft.
Buildings A and B use the same basic system, but there are three significant differences.
1. Whereas the cooling air outlet shaft in Building A is made of steel framed steel reinforced concrete, in the Building B design the steel reinforcement was removed to leave just steel framed concrete.
2. The steel plate on the floor of the storage pit was removed in the Building B design.
3. The positioning of the plates that the air flows past in the inlet and outlet shafts was changed and their orientation was changed from horizontal to vertical in the Building B design. These plates are to screen out radiation from the glass canisters.
The three buildings for high-level waste generated at the Rokkasho Reprocessing Plant all have the same design problems as Building B. Two of these have already been built and the third is currently undergoing safety inspections.
The changes made since Building A was approved - the reinforcement of the concrete, the floor plate and the plates in the inlet and outlet shafts - were all clearly cost cutting measures. When the Rokkasho Reprocessing Plant was approved in 1989, the construction costs were said to be 760 billion yen. However, in 1996 (construction began in 1993) there was a big reappraisal of the design and the costs. Due to rising costs, the refining process was reduced from two steps to one and the number of high level liquid waste tanks was reduced by three. At the same time changes were made to the high-level vitrified waste storage buildings. They look the same as Building A, but safety has been sacrificed. Despite these cost-cutting measures, the construction costs had already ballooned to 1.88 trillion yen. (The cost of construction has continued to rise since then. Now it stands at 2.24 trillion yen. The total cost of construction, operation and dismantling the plant will be 11 trillion yen)
The Nuclear Safety Commission's (NSC) safety inspection of the reprocessing plant's vitrified waste storage buildings had already been completed. However, when NISA carried out a detailed inspection of the design it discovered that the analysis of the flow of air to cool the canisters had not taken into account the changes to the plates in the inlet and outlet shafts. The problem was exposed when, on 14 January 2005, NISA requested JNFL to redo its analysis. It became clear from JNFL's own analysis that the internal temperature of the glass canisters would exceed 500 degrees C, reaching a maximum of 624 degrees C. This has the potential to create a huge safety problem for the glass canisters.
These glass canisters are made by mixing borosilicate glass, which is said to be very hard, with high-level radioactive liquid waste, pouring this mixture into stainless steel canisters and allowing it to harden by cooling. The purpose is to seal the radioactivity in a more manageable solid matrix, by mixing this glass with the radioactive wastes. In liquid form the high-level radioactive waste is very difficult to handle. However, the energy released by radioactive decay causes the temperature of the glass to rise. Even if it remains well below the melting point of the glass (1,150 degrees C), once the temperature reaches the so-called transition temperature (450-500 degrees C), the glass becomes rather similar to a liquid. Above 610 degrees C crystals of borosilicate acid begin to form and cracks may appear. As a result, the mechanical strength and resistance to erosion of the glass is reduced. Therefore, it is essential that the benchmark maximum temperature of 500 degrees C is not exceeded.
Safety inspection can't be trusted
JNFL says that it will change the design of these facilities, including those that have already been built. However the biggest problem is that all of these facilities had already passed the safety review and been officially approved before the fault was discovered. Furthermore, NISA's check didn't verify all the important values in the original analysis. It makes one wonder what other yet-to-be-discovered safety problems lie lurking in the Rokkasho Reprocessing Plant?
Connection with uranium trials
The problem was exposed on 14 January 2005, just over three weeks after uranium trials commenced. However, we have since learnt that the problem was already known to NISA when the uranium trials commenced on December 21st. On 16 November 2004 NISA commissioned the Japan Nuclear Energy Safety Organization (JNES) to carry out a cross-check of the heat analysis for Building B. JNES reported to NISA on December 17th, four days before the uranium trials commenced. One can speculate as to what led NISA to commission JNES to do a cross-check in the first place. Did they already suspect that there was a problem? Once it was confirmed that there were problems, clearly NISA should have called a halt to the uranium tests. However, as it turned out, the problem was concealed until after the uranium trials started. So once again, costs and schedules are prioritized over safety issues.