GNEP and DUPIC in America: where security intersects with commerce

It’s pretty certain that the fast reactor–based back end of the U.S.-led Global Nuclear Energy Partnership (GNEP) is dead. What will happen with GNEP’s front-end objective, which is to manage global proliferation risks by offering comprehensive fuel services including spent fuel management? Here’s what I think should happen.

Regardless of decisions on the back-end, GNEP’s front-end should be preserved—and implemented strenuously and soon. The world needs U.S. leadership in this area, and GNEP is the best international anti-proliferation framework.

The program’s front-end could work perfectly well without a back end based on separative reprocessing and fast reactors, or, for that matter, on any kind of fuel recycling. But “spent fuel management” means removing spent fuel from your customer’s site. If this is a critical component of GNEP, and GNEP is the U.S.’s idea, then the U.S. must give credibility to GNEP by repatriating spent fuel. And this should happen sooner rather than later, if the U.S. wants to contain the spread of reprocessing technologies.

Some in the U.S. nuclear industry have sensed an opportunity in the general resistance to the fast-reactor approach, and have proposed an alternative back-end “interim” fuel cycle based on burning mixed-oxide (MOX) fuel in existing light water reactors.

The problem with this is proliferation. Opponents of the fast-reactor approach are as concerned about the spread of reprocessing technologies (MOX is one) as they were about the cost and technological uncertainties of the fast reactors themselves. MOX fuel offers only marginal advantages when it comes to reducing the attractiveness of material from the viewpoint of a would-be proliferator.

In any event, the MOX/LWR option is not the most proliferation-resistant way to re-use spent nuclear fuel. The award in that category goes to DUPIC (Direct Use of Pressurized Water Reactor Spent Fuel in CANDU). In DUPIC, spent fuel is only mechanically reconditioned. Fission products remain mixed with fissile elements, thereby preserving the same radiation barrier that protects regular spent fuel. Of all the reprocessing methods, DUPIC would be the least useful for a proliferator.

From a cost point of view, DUPIC is more expensive than the current once-through fuel cycle. Does this extra cost buy down proliferation risk? Yes, by engaging in the most proliferation-resistant method of reprocessing and showing the right example to the world. Yes again, by putting spent fuel to productive use in the U.S., thereby relieving pressure on spent-fuel storage in the U.S. And yes again, because the preceding benefit would give credibility to American offers of comprehensive international fuel services.

The extra cost has been estimated to be six to ten percent above the once-through cost. This is far below credible estimates for the cost of the fast-reactor fuel cycle, which started at 25 percent above once-through. Moreover, there are many CANDU reactors operating around the world, with excellent  in-service records. Their operating costs are well known, and competitive with those of light water reactors.

Canada would of course benefit in this scenario. CANDU sales in the U.S. would be a huge coup for state-owned AECL, as they would be for any reactor manufacturer. The Canadian government should support this, the same way it supports reactor sales to China and India.

Canadian diplomats are at this moment mounting an intense effort aimed at removing, or exempting Canada from, protectionist measures in the stimulus bill currently before the U.S. Senate. This diplomatic expertise could and should be marshaled for a similar effort in favour of DUPIC in America. It would benefit both countries, not to mention the world.