by Allison Macfarlane and Adam Bernstein

Bulletin of the Atomic Scientists

May/June 1999

Macfarlane is a geologist and science fellow at the
Belfer Center for Science and International Affairs
at Harvard's Kennedy School of Government.

Bernstein, a physicist, is a post-doctoral fellow
at Sandia National Laboratory in Livermore, Ca.

Russia and the United States face an enormous and urgent post-Cold War task: converting their vast plutonium surpluses into a form that is relatively safe from theft by terrorists and from reuse [should either country later attempt to increase its nuclear weapons stockpile].

Each country has made 50 metric tons of weapons plutonium available for the conversion process -- generically referred to as "disposition" -- and both have informally agreed that two technologies may be suitable for the task.

One process would immobilize the plutonium in ceramic or glass, along with highly radioactive waste. The second would fabricate plutonium-bearing mixed oxide [MOX] fuel and burn it in civilian reactors. In either case, the plutonium would end up within a large, heavy, and highly radioactive waste form, making theft and recovery difficult and expensive.

While the United States plans to pursue both approaches, Russia is focused on only one -- burning excess military plutonium in reactors. Meanwhile, both countries seem to have overlooked the benefits that would come from also pursuing the immobilization process in Russia.

The unyielding Russian preference for burning surplus military plutonium is in tune with Russia's long-held commitment to a closed, plutonium-based nuclear fuel cycle in which plutonium is recovered from spent power reactor fuel and recycled into fresh fuel.

However, because of Russia's economic difficulties, it cannot afford to burn the excess military plutonium or to immobilize it. The United States will have to provide much of the money if the surplus is to be dealt with in a timely manner.

As the debate continues, both countries should consider not only the proliferation risks but the technical and economic feasibility of the options.

Little attention has been given to the fact that Russia is better positioned to immobilize its plutonium than to use it in reactors. A significant part of the infrastructure for immobilization already exists in Russia; in contrast, the infrastructure for the reactor option is poorly developed and will be expensive to construct and operate. Immobilization may be the faster and less expensive route to the disposition of plutonium in Russia.

Perhaps because of Russia's lack of interest in immobilization, close analysis of the cost and schedule needed for the two options has yet to be performed. We have tried to remedy this, in part, by outlining the difference between disposition methods. [We used 50 metric tons of surplus military plutonium as our basis.]

We compared available infrastructure, cost, and time to completion, using studies by the U.S. Energy Department, the Russian Ministry of Atomic Energy [Minatom], and independent analysts. Because all these studies have large uncertainties in their cost estimates, they served as rough guides rather than as precise predictions.

Both disposition options require an interim plutonium storage facility, a plutonium metal-to-oxide conversion plant, and a long-term storage facility once disposition is complete. Because these three facilities must be built regardless of which option is chosen, they were not included in the comparative analysis.

Based on preliminary cost and timing estimates, we estimate that the immobilization track would cost at least a billion dollars less than the MOX program, and it would be faster by five to ten years.


In our analysis, we used the preferred choice for immobilization in the United States, "the can-in-canister" design. First, plutonium is incorporated into a crystalline ceramic matrix and then loaded in small cans; second, the cans are placed within larger canisters that are then filled with molten radioactive waste-bearing glass that cools to form a lethal barrier. Can-in-canister requires a vitrification plant to make the high-level waste glass and a plant to immobilize the plutonium in cans.


Russia already has a large vitrification facility at the Mayak complex, built to immobilize high-level waste produced by the RT-1 plutonium reprocessing facility. The plant, in near-continuous operation since 1991, was shut down in 1997 for a standard replacement of the glass melter. It is expected to be back in operation in the next few years. Beyond security upgrades, little physical modification of the plant would be required to accommodate the plutonium-bearing cans.


Can-in-canister requires a new facility for immobilizing plutonium directly in glass or ceramic. To minimize transportation, it should be built near the existing vitrification plant at Mayak. Much of the plutonium-handling equipment and experience needed for construction of the plant already exists at Mayak. Russia already has a great deal of expertise relevant to the immobilization of radioactive materials in ceramic and glass waste forms.

While we were unable to find a Russian estimate of the total cost of immobilization, the Energy Department estimates the entire life-cycle cost of a can-in-canister program in the United States to be less than $2 billion. The figure could be much lower in Russia. The Energy Department estimates that in the United States, 50 metric tons of plutonium could be immobilized in 18 to 21 years. We believe a similar time scale should be possible in Russia.


To irradiate plutonium in reactors, a MOX fuel-fabrication plant would be needed, as well as enough reactors with an adequate capacity to accept 50 metric tons of plutonium in a reasonable time frame. The use of both breeder and light-water reactors has been suggested.


Although mixed-oxide fuel may be fabricated for either type of reactor, Russia has no full-scale MOX fabrication plant for either type of fuel. It does have pilot-scale plants at Dmitrovgrad and Ozersk [formerly Chelyabinsk-65] where a few tens of kilograms of plutonium can processed per year. At Mayak, a partially built plant designed to make MOX fuel for breeder reactors has a nominal capacity of about five to six metric tons of plutonium per year, which would be a suitable scale for disposition. The plant was about half complete when its funding was cut and construction halted in 1989. It would have to be substantially modified to produce fuel for light-water reactors.


Perhaps the most realistic option for a MOX fabrication facility would be to start over with a completely new plant -- or at the very least to import materials from a partially constructed plant in Hannau, Germany.

Overall, Russia's MOX experience is geared toward the design and use of MOX fuel in breeder reactors. Minatom officials acknowledge that Russia has little or no experience with fabricating or using MOX in light-water reactors. [There is extensive experience with MOX fabrication in Europe.]


Minatom's first preference for plutonium disposition is to use it in breeder reactors. However, Russia's single operational breeder, the BN-600 at Beloyarsk, was designed to use uranium-based fuel. Minatom has indicated that it would be difficult to convert the BN-600 to MOX. In the best of circumstances, moreover, the reactor could not handle more than five to 10 percent of the total plutonium surplus over its remaining lifetime. The BN-600 also has had serious safety problems, including a fire in 1994 caused by a sodium leak.

Because the BN-600's capacity is insufficient, Minatom hopes to construct a larger reactor, a BN-800, designed for MOX use. Such a reactor would take about 30 years to irradiate 50 metric tons of plutonium, not including the construction time for the reactor and MOX fuel plant.

But there are two problems with the plan. First, the cost of the reactor would be high. To construct and operate a BN-800 would take about $7 billion dollars, according to Russian estimates. Minatom argues, however, that because it already plans to build the reactor for energy generation, the marginal cost of plutonium disposition would be only a few hundred million dollars.

Western analysts are generally convinced that Russia cannot afford to build breeders. And the United States will not subsidize the construction of a breeder in Russia, because U.S. non-proliferation policy discourages their use on the grounds that they require the separation and recycling of plutonium suitable for weapons use. Given that, there is general agreement in the West that Russia will be unable to build breeders in time to affect the disposition program.


Less desirable to Minatom, but somewhat more realistic, would be the use of Russia's VVER-1000 series of light-water reactors. Although they are deficient by Western standards, the VVER-1000 series are the most modern of Russia's reactors. There are currently seven operating VVER-1000s at three sites.

The VVER-1000s would require additional or larger control rods to compensate for the increased reactivity of MOX fuel, as well as other safety measures. Beyond that, all the existing reactors will reach the end of their 30 year lifetimes within the next 25 years. In that time they could process only about half of the 50 metric tons.

Minatom plans to complete three more VVER-10002, but their construction has been repeatedly halted for lack of funds. According to Minatom, disposition of 50 tons of plutonium with the seven existing and three new light-water reactors would take about 32 years, including the construction time of the new reactors -- assuming that the existing reactors' lifetimes could be extended.

This estimate -- six years more than the estimate for immobilization -- rests on the assumption that a MOX plant could be made operational within three years of a go-ahead decision -- an assumption rightly characterized by Minatom as "optimistic."

The total lifetime costs for building and operating the new light-water reactors is far more than the cost of the MOX disposition alone -- some tens of billions of dollars. But Minatom argues that the marginal cost of disposition, once the reactors are built, would be about $3 billion, with remaining capital and operating expenditures recovered in whole or in part by electricity sales.

That assumption is optimistic. If the pace of past construction is an indication, Russia will have great difficulty building and operating the new reactors in a timely fashion. Moreover, Russian electricity consumers have difficulty paying their energy bills, just as nuclear reactor operators have difficulty paying their workers.

Minatom has also proposed that 11 Ukrainian VVER-1000 reactors be used along with Russia's own reactors. The Ukrainian reactors could greatly speed up the timetable, but the plan would involve additional negotiations between the United States, Ukraine, and Russia, and it would mean significantly more transportation of plutonium in the form of MOX fuel between Ukraine and Russia. Moreover, Ukraine, and is unlikely to become involved without some compensation, adding to the expense and uncertainty.

[Transportation is no small matter. Immobilization could take place at a single facility, probably at Mayak. In contrast, MOX fuel assemblies would have to be transported hundreds of miles to reactors at four or more widely separated locations over a period of decades, using a transport system of questionable reliability and security, adding to the cost of the whole process.]


Russia already has a good technical base from which to pursue immobilization. Further, the Unite States must consider the likelihood that it will be forced to provide more -- possibly much more -- than the marginal cost of the reactor program to avoid a slowdown in dealing with excess plutonium. The immobilization program, in contrast, is self-contained. Consequently, the United States, in parallel with its own dual-track program, should fund immobilization research and development in Russia as well as the MOX option.

That involves a financial risk. Minatom believes that plutonium is an energy resource to be used in reactors, not something to be wasted in immobilization. It may never accept immobilization. Nevertheless, we propose the following:

Without discontinuing its support of the MOX effort, the United States should guarantee funding for construction and operation of a small-scale immobilization experiment in Russia. This could be operated in parallel with the U.S. demonstration plant at the Savannah River Site. The feed for the experiment could be "waste" plutonium, material that is contaminated and therefore not suitable for fuel fabrication.

In the past, Russian officials have said they do not have significant amounts of waste plutonium in their military stockpile. If necessary, the United States should purchase a modest amount of surplus weapons plutonium from Russia for use in the demonstration facility.

As a follow-on, the United States should fund a joint U.S.-Russian project to design a larger-scale plutonium immobilization plant.

While US security interests might be better served if all surplus Russian plutonium stocks were immobilized, a useful near-term tactic to encourage immobilization would be to identify those parts of Russia's plutonium stockpile that are clearly unsuitable for fabrication into MOX fuel.

For example, even if Minatom's assertion is correct that none of Russia's military plutonium is too impure to be used as MOX fuel, the pit conversion and MOX fabrication processes would themselves generate plutonium-bearing wastes requiring immobilization.

Further, Minatom estimates that 100 tons of military plutonium ultimately will be declared surplus [in addition to the growing pile of about 30 metric tons of plutonium separated from civilian spent fuel]. If the estimate is correct, it will be virtually impossible for Russia to irradiate its plutonium surplus in reactors in any reasonable time frame. Some will have to be immobilized, if US and Russian disposition efforts are to proceed in parallel.

The United States should continue and expand the effort to identify and foster Russian expertise in immobilization. US efforts to support the immobilization option in Russia will fail in the long run if domestic expertise is not identified and supported. The Energy Department and its weapons laboratories, which provide funding to Russian researchers and which work closely with them on disposition, have not ignored the immobilization option in Russia. However the funding levels related to immobilization are low and represent just a small fraction of the total budget spent on disposition in Russia.

Contrary to common belief in the security community, immobilization is feasible in Russia. The United States should recognize this and encourage further study. Inevitably, US funding will play a key role in Russia's plutonium disposition program. By taking a closer look at immobilization, the United States can make sure that its money is well spent.

[ The Dangers of Encouraging Plutonium Use ]
[ Bomb Makers Speak Out Against Plutonium ]

[ Plutonium Sub-Directory ]
[ Short Directory ] [ COMPLETE DIRECTORY ]


Since March 27th 1996, there have been over
100,000 outside visitors to the CCNR web site, plus

(counter reset July 2nd 1998 at midnight)