Reprocessing Nuclear Waste: A Solution in Search of a Problem?
Recently, President Bush announced $250 million for a pilot nuclear reprocessing facility. The plan would use a process, called UREX+, that separates uranium, technetium, trans-uranic elements (TRUs) and possibly iodine, cesium, and strontium. The goal appears to be to reduce the volume, heat output, and future escaping toxicity of waste destined for geologic disposal.
Schemes like this face serious economic problems.
Today, the spot price of U3O8 (yellowcake) is around $37/lb. At this price, reprocessing does not come close to paying for itself from avoided fuel costs. It's cheaper to just fabricate new fuel elements from freshly mined uranium (especially when one considers the difficulty of fabricating MOX fuel elements, which have to deal with plutonium contamination of the fabrication line.)
But maybe uranium will get more expensive in the future, and reprocessing will make sense. How far in the future? Estimates of the cost of the Japanese amidoxime scheme for extracting uranium from seawater are in the range of $100 to $250/kg; the cost of extracting fissionables from spent fuel are at least three times this. The oceans contain 4.5 billion tons of uranium; roughly 150 tons/year of natural uranium would be required to feed the enrichment plant to fuel a 1 GW(e) powerplant on a once-through fuel cycle. This source will not run out soon, even if the world's nuclear capacity is greatly expanded.
So, why is Bush proposing this recycling research? My guess is he wants to close down the as-yet unused Yucca Mountain geological disposal facility, and needs to be able to present the nuclear waste problem as being solvable in the future. I personally have little doubt it is solvable, even without this program, but it does provide political cover.
I also think delaying geological disposal is the right thing to do. Spent fuel can be safely, securely and economically stored in above-ground facilities for decades or centuries. The storage scheme, generically called dry cask storage, involves placing cooled fuel elements in sealed armored steel containers filled with inert gas. The containers are further shielded with concrete or other materials; some double as transport containers for the spent fuel. These casks are rugged, resistant to diversion or attack, and require no active cooling systems.
At some point in the distant future, it might become economical to reprocess the accumulated spent fuel. At that point, I suggest it be done in space. Reprocessing inevitably leaks some material, contaminating the facility and, if you are not very careful, the surrounding land. If it were done on the moon (for example), this contamination would be unimportant. The cleanup cost of the Rocky Flats facility, where plutonium was handled, reached $7 billion, and that was only because some corners were cut (saving $29 billion). The volume of material handled at Rocky Flats was a small fraction of the volume that would flow through industrial reprocessing facilities.
Launch costs would have to come down a lot for this to be economical -- I'd guess a factor of 100 would be more than enough -- but it's a long term idea, so that's reasonable. PWR fuel elements would have to be armored against launch accidents (even on space elevators). However, the fuel elements from high temperature gas-cooled reactors could probably be launched without reentry armor, since they'd likely survive reentry largely unscathed, encased in rugged graphite/silicon carbide shells.
The proposed UREX+ process separates out technetium (which is important for long term radiotoxicity due to its solubility and the ease with which it escapes from the repository), so even if terrestrial reprocessing can be made to work, it's attractive to dispose of this element in space instead of, say, transmuting it in a reactor. The volume is much smaller than that of the unprocessed spent fuel, so this could be attractive even at today's launch costs.