Letter to the Editor: The Coming Glut of Japanese Spent Fuel

Leonard Spector

Frank von Hippel’s article (“South Korean Reprocessing: An Unnecessary Threat to the Nonproliferation Regime,” January/February 2010) on the proliferation risks of South Korea’s plans for reprocessing spent fuel from its nuclear power program elegantly frames what is likely to become a major controversy as South Korea’s agreement for nuclear cooperation with the United States comes up for renewal in 2014.

Von Hippel argues that the South Korean approach, based on an unproven technology known as “pyroprocessing” and yet-to-be-designed fast reactors, is unlikely to succeed on a scale sufficient to alleviate South Korea’s spent fuel management problem. Moreover, he stresses, it could introduce new proliferation risks by creating stocks of material from which plutonium could be more easily extracted than from spent fuel.

To underscore his point, von Hippel highlights the great difficulties Japan has encountered in its own spent fuel reprocessing program, based on classic reprocessing technology that is well understood, and conventional reactors. The situation in Japan, however, is considerably worse than von Hippel describes, making his core point all the more powerful.

According to recent information, Japan plans to reprocess 32,000 metric tons of spent fuel (17,000 now in storage and 15,000 to be discharged in the future) over the next 40 years. But its reactors will discharge a total of 45,000 metric tons of spent fuel during this time, leaving it with 30,000 metric tons to be held at an interim storage facility to be built not far from the Rokkasho site.

At that point 40 years hence, with its current reprocessing plant at the end of its useful life, Japan would need to build another reprocessing plant to start working down this excess spent fuel; but during the operation of that plant, still more spent fuel will be discharged, probably at a greater rate than in the current 40-year period, as Japan’s nuclear sector grows. This implies that as far into the future as one can reasonably see, Japan will be storing tens of thousands of metric tons of spent fuel on an “interim” basis that in practical terms is likely to be perpetual because the Japanese reprocessing program never catches up with discharges.

Thus, although the Japanese spent fuel reprocessing program is presented as a solution to the country’s spent fuel management problem, it is only a partial solution at best. It is also an extremely costly one. Von Hippel notes that the Japanese have stated that the construction, operation, and decommissioning of the Rokkasho facility will amount to $100 billion. Assuming that the plant processes a total of 32,000 metric tons of spent fuel, the cost of the Japanese approach would be a hefty $3,125 per kilogram of heavy metal processed.

As von Hippel stresses, all of these challenges will confront the South Korean pyroprocessing endeavor, but will almost certainly be more severe, given the novelty of the technologies involved. As a result, almost certainly, the approach, like Japan’s, will be at most only a partial solution to Korea’s pending spent fuel glut.

Finally, we need to take von Hippel’s commentary one step further and articulate a strategy that works: strong international validation of 100- to 200-year interim storage of spent fuel as an appropriate spent fuel management option. Japan has tacitly accepted this option for half of its spent fuel. With the termination of the Yucca Mountain underground repository project, it appears that the United States also is going to be forced to do so for its entire spent fuel output, the largest in the world.

This is, in truth, the de facto international standard. But it needs to be formally recognized. The environmental community, in particular, needs to express confidence that this option is safe and can adequately protect the public from harm for generations to come.

Leonard Spector is deputy director of the James Martin Center for Nonproliferation Studies and heads the center’s Washington office.