The US administration’s proposed Global Nuclear Energy Partnership (GNEP) has been something of a moving target. When first announced, GNEP was cast as a bold new approach toward the global nuclear economy, aimed at attacking the dangers of proliferation and significantly reducing the nuclear waste problem at the same time (see link below to GNEP: the right way forward?). The story presented to the US Congress is decidedly domestic, focusing almost entirely on a way to stretch the capacity of Yucca Mountain.

The objectives of GNEP are sound, going a long way toward fulfilling International Atomic Energy Agency director general Mohamed ElBaradei’s proposal for an internationalised nuclear fuel regime. But the administration’s actions toward that goal have been limited to an offer of 17t of highly enriched uranium to be downblended for reactor fuel, enough to keep one large reactor operating for 15-20 years, or a couple of months of US consumption.

At the same time, Congress is searching for politically plausible alternatives to Yucca Mountain. Last year’s Department of Energy (DoE) appropriation required the department to provide a report to Congress outlining its plan for recycling. This important document, the DoE Spent Nuclear Fuel Recycling Program Plan, was recently delivered. As directed by Congress, the focus is on the recycling technology, not on the larger GNEP. Even so, it is surprising how the report concentrates on management of domestic nuclear waste with barely a mention of the international responsibilities created by GNEP.

The current proposal is to take spent nuclear fuel from light water reactors, separate the fission products, the uranium, and the plutonium along with all other transuranics. The fission products will go eventually to a geologic repository, which by default must mean Yucca Mountain. The uranium is the huge bulk of the fuel by weight and volume but is only slightly radioactive and so presents a less challenging disposal job. The plutonium and other transuranics would be fabricated into new fuel for a fleet of yet-to-be-built fast neutron reactors. The fast neutron reactor fuel would be continually recycled until all the transuranics are consumed and rendered into fission products. Those fission products would then be disposed of in a geologic repository.

Debate in the Congress demonstrates that some of GNEP’s supporters believe that plutonium breeding and the consequent stretching of the world’s uranium supplies is one of its major assets. The DoE strongly downplays this, excluding it from the current technology plan and leaving it as a possible option for the long term.

There are two almost independent bases of support for GNEP in the Congress. Nuclear power supporters seem to believe it is an important advance in nuclear power technology and see it as a way of jump-starting America’s near moribund nuclear research and development community. Even nuclear skeptics see it as political cover that provides some protection from painful choices on the unpopular programme at Yucca Mountain.

Few are willing to argue that Yucca Mountain is the ideal long-term geologic waste repository. It was picked, in part – and there can be much debate about how large a part – because of political expediency. At the time of the decision, Nevada had a much smaller population and was politically weak. Since then, the population of Las Vegas has exploded and Senator Harry Reid of Nevada is the Minority Leader in Congress. Whatever the fate of Yucca Mountain, it continues to be an ugly political battle that no-one wants to repeat. Thus, a few look to recycling as an alternative to Yucca while most supporters of recycling look at it as a way to substantially increase the loading of Yucca while avoiding having to decide on a second repository site during any current politician’s lifetime.

The effects of reprocessing on waste storage are more complex than the DoE public information suggests. The GNEP website boasts that reprocessing will substantially reduce the volume of radioactive waste. While this is true, volume is easily the least relevant measure of the nuclear waste problem. For example, the great majority of the waste coming out of a nuclear reactor is uranium-238, which, with a five billion year half life, is not particularly radioactive. What is relevant is the total radioactivity, the half lives of the radioactive components, the geologically mobile radioactivity, and the heat generation.

The greatest source of radioactivity for recently-removed fuel rods is the fission products. These are unavoidable. All fuel cycles will produce about the same fission products for a given amount of electric energy produced. In some fuel cycles the fission products will come from uranium-235 and in others from plutonium-239, but there will be only small differences between the mixtures of fission products.

In fuel from the once through fuel cycle, longer-term radioactivity and the heat production after the first several decades are dominated by a few transuranic isotopes, such as plutonium and americium. With fully recycled fuel, these elements would, indeed, be substantially reduced. And it is heat production, not volume, that limits the amount of waste that can be put into Yucca Mountain. Since recycling would reduce substantially the heat production, it would allow a much tighter packing of fission products. However, the long-term exposure danger due to water transport to the surface comes primarily from fission products, particularly isotopes of iodine and technetium, not from the transuranics. Thus, removing the heat producers and packing more fission products in could actually increase the ultimate surface radiation exposure from fission products. Note, of course, that this does not increase total surface exposure, it simply concentrates the surface exposure at one site, which may be considered an advantage by someone living anywhere else.

Part of this proposal calls for the selected site to provide ‘process storage’. That is, fuel expected to be reprocessed at the site will be moved from the reactor site and stored at the reprocessing site until needed. Considering that the fuel will have to be packaged securely enough to be transported to the site and then stored potentially many years, the fuel will have to be in something equivalent to dry casks. Thus, even if no fuel is ever reprocessed, the plan will have produced an intermediate term interim above-ground storage site that will provide an alternative to Yucca.

One might hope that the DoE would be backing up its recycling proposal with a detailed economic analysis, but only the sketchiest assertions about costs have been made thus far. The report to Congress says that one of the goals of the R&D programme is to make recycling ‘economic’ but little else. The value of recovered fuel will not make recycling worthwhile for the many decades that cheap uranium supplies are expected to last. Moreover, fast neutron burner reactors will inevitably be more expensive than their thermal cousins. Cheaper electricity is not the justification for recycling. The cost of operating reprocessing facilities and fast neutron reactors will be reflected in their management, ownership, and control. Without economic incentives, commercial utilities will not choose to operate fast reactors. Various models are possible: government-owned plants, whether government or contractor operated; or commercial plants operating with large government subsidies to cover the difference between the cost of electricity and its market price.

Without a cost advantage, recycling is being sold primarily as waste management. The comparison is with a geologic repository and, with the ever-escalating costs estimates of Yucca Mountain, recycling should look attractive. It is important to keep in mind, however, that the amount spent on Yucca Mountain is far in excess of original estimates. Much of the increased cost came about because of changes mandated through various technical and environmental reviews. Reprocessing has not yet been subject to those sorts of reviews with their resulting cost increases. Tomorrow’s blueprint will almost always seem cheaper when compared to today’s real world engineering problems.

The political calculus is somewhat analogous to the economic analysis. To an extent, reprocessing is an attempt to escape the political pain of finding a site for a second geologic repository. But this simply trades the well-know political problems of Yucca Mountain for the thus far hypothetical, but most likely equally intense, local political resistance that can be expected from trying to site more than a dozen fast neutron reactors, a couple of reprocessing centres, and the transportation of spent fuel.

The political debate has not yet come to grips with the conundrum that the international responsibilities implied by the GNEP are in conflict with the waste reduction goals. Recycling can reduce nuclear waste but if the rest of the world is shipping additional waste to the USA, any gain could be easily wiped out – and bringing foreign radioactive waste into the country would be a political challenge.

GNEP was covered in considerable detail at the DoE FY2007 budget rollout that included a $250 million request for reprocessing technology development and demonstration planning. At the time of writing (mid-June) the Energy and Water Bill has been marked up in the House of Representatives and $100 million was cut from the request directly while another $30 million was diverted to other energy programmes in a second cut that left only $120 million. Members of the house expressed some skepticism about the DoE’s ability to efficiently manage the programme. Senate support is not clear at this moment but certain key Senators are strong supporters. There may be a wide gap between the Senate and House positions that will have to be resolved in a conference committee.


The other half of the GNEP proposal is its proliferation advantages. The anti-proliferation arguments contain two important logical fallacies. First, the overall GNEP proposal calls for current nuclear fuel producers to become the world’s exclusive nuclear fuel producers. That means that countries that cannot now enrich uranium would be denied that capability in the future because enrichment can be used to produce weapons grade uranium. In a reactor, nuclear fuel containing uranium-238 produces plutonium, which can also be used in nuclear weapons. Therefore, fuel producers would not only provide fuel for consumer nations but would take back the used fuel for reprocessing. The fallacy is not just that the fuel has to be taken back – it could be placed in an appropriate geological repository in the user country – but that it has to be taken back for reprocessing. Whether the fuel is reprocessed, put into a geological repository, or launched to the moon, once it is out of the hands of the non-nuclear weapons state, the anti-proliferation goal has been met. The GNEP proposal tries to make an essential logical connection between reprocessing and non-proliferation when, in fact, there is little or none.

Second, the GNEP proposal states that the envisioned separation technologies are ‘proliferation resistant’. (The DoE is very careful not to claim that anything is ‘proliferation proof’.) There have been various proposals for new separation techniques, for example, Urex, Urex+, UREX+1, and now Urex+1a. As the names imply, they are variations on a theme. In the longer term, other techniques, such as pyroprocessing might become available on an industrial scale. When GNEP proponents say that these techniques are ‘proliferation resistant’, they mean they are when compared to the Purex process. Purex was developed during the Manhattan project specifically to provide plutonium for the first atomic bombs. The claim is, then, that Urex variants are less proliferation prone than a process that was specifically designed for bomb manufacture, a very low hurdle indeed. But none of these processes is more proliferation resistant that what we are planning to do now, that is, disposal of sealed, intact fuel rods in a geologic repository.

Part of the alleged proliferation resistance comes about because some variations on Purex – for example Urex+ – intentionally leave radioactive contaminants in the plutonium to make them more difficult to steal and handle if stolen. Frank von Hippel and Jungmin Kang at Princeton University have calculated the radiation doses from Urex+ and pyroprocessed fuel and found them falling short of meeting the standards of ‘self protection’. Moreover, even if impurities are intentionally left in the plutonium, nothing prevents a thief from using a simplified version of the 60-year-old Purex technology to get pure plutonium out. Some approaches, such as leaving in chemically similar radioactive rare earth elements make self protection more robust but substantially increase the final fuel fabrication costs. Finally, as pointed out by Richard Garwin recently in Congressional testimony, spent fuel from a nuclear reactor is about 1% plutonium, while Urex+ fuel would be more than 90%, so a thief would need to steal only about 9kg of Urex+ fuel to get an 8kg critical mass of plutonium but would have to steal approximately 800kg of lethally radioactive spent fuel to get a critical mass.

The DoE proposal will restart plutonium reprocessing in the USA after a three-decade hiatus. Plutonium reprocessing was tried and abandoned in the country because it was uneconomic and increased the global availability of plutonium, which can be used in nuclear weapons. It has been US government policy to set an example against commercial reprocessing because of the nuclear weapons proliferation danger. The example has not proved persuasive in all cases, obviously. Some other countries, at great expense, continue reprocessing. The USA kept, of course, a programme for production of military plutonium, but even that has now ceased.

The track record on reprocessing is not good: the one commercial plant in the USA, in West Valley, New York, took six years to reprocess one year’s worth of reactor waste and was shut down as uneconomic, leaving behind a multi-billion dollar environmental cleanup bill; Japan has just opened a new reprocessing facility in Rokkasho that, at $20 billion, is about three times more expensive than originally budgeted; the Thorp facility in Britain was shut down last year after a huge leak; and the subsidised French programme continues to produce separated plutonium faster than commercial reactor operators are willing to accept it, resulting in ever-increasing stockpiles.

If plutonium and transuranic reprocessing and recycling are not the answer, then what is? Whether or not Yucca opens, there will be more waste that has to be handled somehow because Yucca’s capacity will be reached long before any recycling system is ready. Geologic storage will probably turn out to be cheaper, more proliferation resistant, and, as politically painful as a second repository would be, less painful than the recycling alternative. And there is no rush. We don’t have to force ourselves into early decisions about immature technologies. Even if reprocessing moves forward, waste will have to be placed in dry casks for transport. The consensus is that dry cast storage is stable for at least a century. Interim – but many decades long – storage of waste allows time for a research programme without a forced demonstration schedule, delays the capital costs, and allows time for the decay of important fission products.

Does this mean that GNEP dies? Not at all. The goals of GNEP can still be met. An international enriched fuel bank could be supplied by several nations across the political spectrum. If the bank were heavily subsidised so that enrichment is effectively sold at below cost, then any nation pursuing independent enrichment capability could be assumed to be up to no good. An international market in spent fuel disposal, operating under stringent international safety and containment standards, would more likely result in regional geologic repositories than regional burner reactor centres. But geologic storage under international observation in fuel supplier nations would solve the proliferation problem as well as recycling in the supplier nations. Plutonium recycling needs GNEP, but GNEP does not need plutonium reprocessing.

Author Info:

Ivan Oelrich is a member of the Federation of American Scientists, 1717 K St., NW, Suite 209, Washington, DC 20036, USA

Related Articles
Vision products
US launches Global Nuclear Energy Partnership
US GNEP programme dead, DOE confirms
GNEP: the right way forward?