Disposition of Fissile Materials: Status and Prospects
Dr. Edwin Lyman, a Senior Scientist in the UCS Global Security Program, prepared the following discussion paper for the Stanford Fissile Material Summer Study symposium in August of 2003.
Nearly ten years have elapsed since the National Academy of Sciences first called the accumulation of stockpiles of excess warhead plutonium in the United States and Russia "a clear and present danger." Yet the international community did not respond with the urgency or determination that this warning called for. To date, ground has not been broken in either the U.S. or Russia for any of the new facilities that are required for the program. The initial cost and schedule estimates for the MOX program have proven to be wildly optimistic. And there has been little progress on critical implementation issues, including the establishment of a multilateral financing plan for the Russian program, a liability regime to protect foreign contractors assisting the Russian program and completion of a bilateral agreement on monitoring and inspections of plutonium disposition facilities.
The ballooning costs and delays of the big-ticket items like MOX plant construction provide an opportunity for a thorough reconsideration of the current warhead plutonium disposition program before forging ahead. A fundamental requirement of the plutonium disposition program is that it must not exacerbate the risks posed by the warhead plutonium stockpiles in their present configuration; that is, the cure must not be worse than the disease. But it is far from clear that this standard is being met by the MOX program as currently conceived. Among the many issues worth examining today in this light are:
- What are the security benefits of plutonium disposition, and do they truly justify the costs and risks?
- Given funding constraints, can a Russian MOX disposition program be adequately protected against terrorist threats? Do terrorist groups pose a credible threat to the United States program?
- Is it wise to proceed with a Russian MOX program with insufficiently stringent monitoring and physical protection regimes, simply to get the program underway?
- Does it make more sense to minimize the required numbers of sites and transports involved in the Russian plutonium disposition program or to minimize the amount of time the program will take to complete?
- Should rejected technological alternatives to LWR MOX, such as immobilization with high-level waste or spent fuel, deep boreholes, sub-seabed disposal, or even dilution and burial in WIPP, be reconsidered with regard to relative cost, timeliness and feasibility?
- Should the United States make a serious attempt to purchase the Russian excess plutonium stockpile outright?
- Are there signs, as MOX opponents have warned, that the warhead plutonium disposition program is indeed breathing new life into the commercial plutonium industry, which has suffered stunning setbacks in the last few years?
- Is it reasonable to expect Russia to accept a liability regime in which it bears full responsibility for all third-party damages resulting from the plutonium disposition program? And would Russia have to bear responsibility for accidents at foreign plants where the MOX fuel was exported?
Growing public awareness of the warhead plutonium problem in the United States and Russia has also led to renewed attention to large stockpiles of civil plutonium around the world that are stored under inadequate security and have uncertain disposition paths. The technical and political difficulties encountered by commercial MOX programs around the world have left countries like the U.K., France, Japan and Germany with no realistic and timely means of disposing of the civil plutonium stocks that accumulated over several decades. A comprehensive plan for reducing the security threat posed by separated plutonium cannot ignore the civil plutonium issue. Since even France, with the most successful MOX program, cannot make a dent in its plutonium inventory, non-reactor alternatives for disposition of surplus plutonium need to be considered for civil stockpiles as well as military ones.
The technical difficulties encountered in plutonium disposition are largely avoided in the disposition of excess highly enriched uranium, as much of the feedstock can be downblended to produce commercially acceptable LEU fuel. (A significant portion of the US excess HEU stockpile has resulted from the reprocessing of irradiated HEU driver fuel, and as a result has contamination levels even after downblending that require restrictions on its use.) However, implementation of the U.S.-Russian HEU purchase agreement has repeatedly been jeopardized by efforts to minimize its impact on the commercial uranium fuel market and to protect the financial health of USEC. Although the Agreement is proceeding well at present, efforts to accelerate the rate of HEU downblending have gone nowhere. A modest budget proposal by NNSA to subsidize an increase the amount of Russian HEU downblended each year was zeroed out by a House Appropriations subcommittee, reflecting the view of some members that HEU disposition has to be able to pay for itself. The policy challenge here is to convince these opponents that the costs are modest in relation to the nonproliferation benefits.
Warhead Plutonium Disposition
Background
In September 2000, the United States and Russia signed a Plutonium Management and Disposition Agreement (PMDA), in which both parties agreed to dispose of 34 MT of weapons-grade plutonium, at a rate of at least 2 MT/year, beginning no later than December 31, 2007.[1] However, the September 2000 agreement really marked the beginning rather than the end of the negotiating process. The most difficult issues were kicked down the road, and as of this writing (August 2003), they have yet to be resolved. In addition, the PMDA has not entered into force, since the Russian Duma has not yet approved it.
Consistent with the "dual track" strategy of the Clinton administration, the United States committed in the PMDA to dispose of 25.6 MT of clean Pu metal and oxide by irradiating it in the form of MOX fuel; the remainder, consisting of impure metal and oxide, would be immobilized in high-level radioactive waste glass using the "can-in-canister" immobilization process. Russia, on the other hand, designated all 34 MT of pure metal and oxide to be disposed of via the MOX route, in both light-water and fast reactors.
While the final version of the PMDA reflects Minatom's well-known dislike of immobilization, the United States did its part to discourage further development of immobilization in Russia. Russia was originally willing to immobilize one MT of plutonium diluted in a large volume of radioactive sludge, but was rebuffed by the United States, which did not consider that material to be suitable for inclusion in the agreement. Russia then increased the amount of clean oxide to be made into MOX by one MT and eliminated the immobilization track altogether. This left the team of scientists working on immobilization in Russia with no potential application for their research. If the United States had really wanted to keep the immobilization option viable in Russia, it could have matched the one MT of plutonium in sludge with an additional one MT of TRU waste of comparable quality.
Status of the U.S. Disposition Program
In the nearly three years since the PDMA was signed, a number of events have caused significant disruptions to the implementation timeline put forth in the agreement. First, when the Bush Administration arrived in 2001, it almost immediately suspended work on can-in-canister immobilization while it conducted a review of the entire plutonium disposition program. While most observers understood this move to be a precursor to a cancellation of the immobilization program, the Administration denied this was the case for nearly a year, arguing that it intended to resume the program in time for a 2014 startup. In the meantime, the NNSA contractor in charge of the MOX fuel program, Duke Cogema Stone & Webster (DCS), submitted a design to the NRC for the US MOX fuel fabrication plant (MFFF) that assumed the feed material would be clean plutonium oxide derived from nuclear warhead pits, and began preparations to license the use of MOX fuel in four reactors operated by Duke Energy.
Citing cost constraints, NNSA officially pulled the plug on the immobilization program a year later, redirecting 6.5 MT of impure material to the MOX track that it had previously said was "unsuitable" for MOX use. Furthermore, an additional 2 MT of "very impure" material was stranded with no defined disposition path, leaving the United States with only 32 MT of plutonium that it was capable of disposing of under the PMDA. To make up for the disposition capacity (1.3 MT/yr) associated with the immobilization plant, NNSA pledged to locate two additional reactors willing to utilize MOX, bringing the total number to six. (None have yet been identified.)
As a result of the cancellation of the immobilization program, DCS was forced to redesign the PUREX-based aqueous purification cycle in the MFFF to account for the impure plutonium (or "alternate feedstock'), much of which was contaminated with chlorides that are incompatible with the stainless steel process equipment called for in the MFFF design. This delayed NRC's review process by more than six months.
The nuclear regulatory Comission (NRC) was also forced to delay issuance of its draft Environmental Impact Statement (EIS) for the MFFF by one year while it took into account the greater environmental impacts associated with the changes to the facility. Despite these substantial delays, the NRC did not push forward the date (September 30, 2003) when it anticipated granting DCS authorization to construct the MFFF. However, this compressed schedule is taking its toll on the quality and depth of NRC's safety review, and NRC said recently that it may grant the DCS construction authorization request prior to fully resolving its safety concerns with the facility, which would be a highly questionable action.
From another perspective, however, NNSA needed to cancel the immobilization program to keep the MOX program from falling even further behind schedule. This is because the design of the Pit Disassembly and Conversion Facility (PDCF), which will use a dry process to convert nuclear weapon pits and clean plutonium metal to relatively pure plutonium oxide feedstock for the MFFF, has experienced repeated technical problems, delays and cost overruns. This is apparent from the fact that even after cancellation of the immobilization program, NNSA's estimate of the cost of the U.S. disposition program was $3.8 billion, nearly twice as much as original estimates.
While according to the schedule in the PMDA, the PDCF would be operating by March 2006, in time to produce a sufficient amount of feedstock by the March 2007 startup date of the MFFF, in fact by 2002 the PDCF preliminary design was only 30 percent complete, and the projected date for PDCF startup had slipped to October 2009, nearly two years after the deadline for MFFF full-scale operation specified in the PMDA. Since only 0.57 MT of the US excess plutonium specified in the PMDA was already in the form of clean oxide, NNSA would not have been able to startup the MFFF on time at processing rate of 2 MT per year without redirecting the 6.5 MT of impure plutonium that was originally slated for immobilization to the MOX program. So NNSA essentially "robbed Peter to pay Paul."
Another management failure that will likely lead to delays in the U.S. program is related to the production and irradiation of weapons-grade MOX lead test assemblies (LTAs) or the MOX fuel qualification program. The NRC will require DCS to submit information about the irradiation behavior of MOX fuel assemblies in the mission reactors (Catawba and McGuire) as part of the license application for full-scale MOX loading. For the information to be useful, the LTAs must be as similar as possible as the mission fuel assemblies. However, the fabrication of the LTAs has turned out to be a significant problem, as the only viable facilities where it could be done are in France and Belgium. Since the LTAs must be fabricated with weapons-grade plutonium that has undergone aqueous polishing consistent with the MFFF process, DCS must figure out a way to supply this material to one of these facilities, which do not normally process weapons-grade plutonium, and to return the MOX LTAs to the United States, without causing an international incident via Greenpeace.
The difficulty in LTA fabrication has led to a significant delay in the fuel qualification program. DCS submitted a license amendment application to NRC for LTA loading in one of the Duke reactors in February 2003 after a delay of nearly two years, but was not able to specify the source of the LTAs. Assuming DCS is able to obtain the LTAs according to its current schedule, it will not load them into a reactor until spring 2005. Since the planned irradiation time is three years, the irradiated LTAs will not be available for post-irradiation examination until spring 2008. Since at least one year is required for cooling and non-destructive analysis (and considerably more time if other types of testing are required), this means the NRC will not get the test data before spring 2009. Assuming a six-month NRC review (and not taking into account the possibility of a public hearing), MOX loading could not take place until fall 2009, two years after the milestone date in the PMDA. And this will only occur if the NRC chooses not to require a more extensive set of tests on MOX fuel under accident conditions to verify whether MOX fuel exhibits worse behavior in loss-of-coolant accidents than does LEU fuel, as some experiments in Europe have suggested.
Status of the Russian Disposition Program
While poor management and planning in the United States have evidently resulted in delays in the disposition program, the Russian program is in even worse shape. Lack of progress in resolving a number of key international issues may set the Russian program even further back than it is in the United States; but since the United States has committed to ensuring that the two programs proceed in "rough parallel," it may have to delay implementation of key milestones to enable the Russian side to catch up, as NNSA itself admitted in its FY 2004 budget request.
From a safeguards and security perspective, the most optimal Russian MOX disposition plan would minimize the number of different sites at which plutonium in accessible forms would be stored and minimize the amount of transport of plutonium between sites. Such a scenario was identified in the 1995 NAS study on reactor-related options for plutonium disposition, in which all processing of bulk plutonium took place at a single existing nuclear weapons complex site—Mayak—and all reactor use of MOX at a single site—Balakovo, with four operating VVER-1000s. While the disposition rate would be limited by the capacity of these four reactors, the program would be much more manageable from a security standpoint.
This scenario should be compared with the default one today. Minatom has decided that it will build its facilities for pit conversion and MOX fabrication at Seversk, a distance of 1,900 km from the new fissile material storage facility at the Mayak plant that the United States has spent four hundred million dollars to build (but which may never be actually used unless a transparency agreement is reached for the facility). The Seversk site in turn is 3,300 km from the 4-unit Balakovo VVER-1000 station, 3,300 km from Beloyarsk, location of the BN-600 fast reactor, and 4,300 km from Kalinin, another VVER-1000 station within Russia. Since it will take until 2025 at a minimum to dispose of 38 MT of plutonium under this scenario, if a faster disposition rate is desired or if more plutonium becomes available for disposition under the PMDA, additional reactors outside of Russia would have to be brought into the program. This would require the shipment of MOX fuel for international export from Seversk to St. Petersburg or one of the Arctic seaports, introducing yet another transport link covering vast distances. This scenario involves hundreds of additional shipments of plutonium over far greater distances than the simple scenario described above, and is inconsistent with the goal of consolidating fissile material in Russia at as few sites as possible and minimizing the amount of transport afterward.
The cost of all this transportation will be considerable. Although the April 2003 Joint Study estimates the cost of transporting plutonium metal, oxide and MOX under the simplest scenario to be less than $20 million, the estimate excludes the cost of "provision of MPC&A in accordance with international standards." Given the world security situation after the September 11 attacks, and the growing threat of catastrophic terrorism within Russia by Chechen extremists, one may expect that providing protection against a realistic design basis threat will not come cheap.
The decision to site the conversion and MOX fabrication facilities at Seversk also means that large head-end storage capacity for plutonium will have to be incorporated into the design of these facilities, whereas if the facilities were sited at Mayak adjacent to the fissile material storage facility, much smaller head-end storage would have been required. The most recent Joint Working Group cost study, assuming that storage adequate to accommodate one year's worth of plutonium feedstock (3.5 MT) would be required at the pit conversion facility, estimated it would cost nearly $100 million to construct such storage—that is, 25 percent of the funds spent to build the Mayak storage facility for a facility that has less than 10% of the capacity. If plans for such storage proceed, the diminishing likelihood that the Mayak facility will ever be used will decrease even further.
Monitoring, Inspections and International Verification
Perhaps the greatest failure to date in the bilateral plutonium disposition program has been the inability of the United States and Russia to conclude an agreement on "monitoring and inspections" by December 2002, as was called for in the PMDA, and the procrastination of both sides in committing to an international verification protocol. The international community has been reluctant to pledge the billions of dollars that will be required to build a disposition infrastructure in Russia without a means to verify that the facilities are (1) being used as intended and (2) are operated under a stringent MPC&A regime, taking into account a realistic DBT.
The United States deserves a large share of the blame for the lack of progress in this area. While the Clinton administration expressed its intention to place the U.S. MFFF on the list of facilities eligible for IAEA safeguards, the Bush Administration has so far failed to do so, and may not actually take this step, which would subject the MFFF to a level of verification comparable to plutonium facilities in non-weapon states. One consequence of the Bush Administration's inaction is that the IAEA has not had any opportunity to review the design of the MFFF to facilitate verification later, even though the design is nearly final and is in the last stages of approval by NRC. This delay is in violation with the provision of the PMDA that requires "consultation with the IAEA at an early date." It is also inconsistent with the IAEA's practice of "design verification." And it does not project the kind of openness that could help to overcome Russia's considerable reluctance to commit to greater bilateral and international transparency for its excess fissile materials.
But neither the United States nor Russia appear to be very interested in ensuring a role for the IAEA in verification of the Russian MOX program. In the most recent report of the Joint Working Group, it is stated that "depending on the results of technical analyses later in 2003, it might be possible to integrate conversion/blending operations within the …[Russian] MOX fabrication facility, and thus avoid the costs of the separate facility."[2] The report makes no mention of the fact that a separate conversion/blending facility is an essential requirement for IAEA verification of the Russian MOX program, as Russia will not allow access to its warhead plutonium until it has been converted to oxide and blended with reactor-grade plutonium. Integration of this function into the MOX facility would require a classification curtain subdividing the MOX facility, making the implementation of IAEA verification (not to mention bilateral monitoring) much harder or even impossible.
The United States is also not setting a good example for Russia with regard to implementation of a domestic MPC&A regime. The NRC has permitted the proposed MFFF licensee, DCS, to submit its license in two parts, a Construction Authorization Request (CAR) and an Operating License Application (OLA). According to this scheme, detailed information about the MC&A and physical protection programs at the MFFF need not be submitted in the CAR, but only in the OLA. This means that the NRC will be able to approve construction of the MFFF without having to consider the detailed MC&A and physical protection arrangements for the facility. This approach is inconsistent with the well-established practice of integrating MPC&A design into facility design, and leaves open the possibility that expensive retrofits to the facility may be required later if certain design features need to be modified to facilitate MPC&A plan implementation. Further complicating this problem is that it is unclear whether the design basis threat (DBT) for plutonium theft and diversion that the MFFF is subject to is the now-obsolete, pre-September 11 DBT or the revised DBT that NRC recently issued for currently operating Category I fuel cycle facilities. These issues are at the heart of a public challenge to the MFFF license application that is now proceeding to a hearing.
Since Russia has agreed to accept the DCS design for its own MOX fuel fabrication plant, any design flaws that could affect adequate MPC&A implementation could be replicated in the Russian plant unless DCS is required to take MPC&A issues into account more thoroughly at the design stage.
Liability
In the PMDA, the two parties agreed to "continue negotiations on liability provisions to apply to all claims that may arise from activities undertaken pursuant to the Agreement…". Such an agreement is nowhere in sight, primarily because Russia opposes the US position that Russia should assume full liability for all damages that may occur as a result of plutonium disposition activities, even if they result from the "premeditated" actions of US contractors. This is similar to the liability agreement underlying the US Cooperative Threat Reduction (CTR) program, but broader than the liability provisions in the 1998 US-Russian Scientific & Technical Cooperation Agreement on plutonium disposition, which the United States allowed to expire in July 2003 for that very reason. Resolution of this issue is critical, because if a Chernobyl-type release were to occur at a reactor using Russian MOX fuel (whether in or outside of Russia), the third-party liability could dwarf the total cost of the disposition program.
Multilateral Financing
The G-7 nations have repeatedly missed target dates for concluding a multilateral agreement to finance the Russian disposition program, a prerequisite for the construction, modification or operation of any facilities under the PMDA. One news item attributed the reluctance of some countries to contribute to the failure to make progress on the monitoring and inspection agreements. Pledges remain about $200 million short of the $1 billion target, which is the capital cost of the program according to current estimates. Even if this amount is raised, the question of who will pay the facility operating costs (at least $1 billion) remains unresolved.
Disposition Alternatives
Given the delays, cost overruns and other problems being encountered in the U.S.-Russian MOX program discussed above, it is worthwhile to consider whether warhead plutonium disposition is the best use of the $6 billion that it is currently projected to cost, and if so, whether alternative technological approaches deserve reconsideration.
With regard to the first issue, one needs to take another look at the original motivation for disposition. If the goal is to demonstrate a commitment to bilateral arms reductions, then disposition has some symbolic value but in practical terms does little to curtail the breakout potential of the two states. If the goal is to reduce the threat that the plutonium will be diverted or stolen by terrorists, then it is not clear that the $6 billion is better spent on conferring an intrinsic radiation barrier to the warhead plutonium stockpile than on providing a highly robust and enduring MPC&A regime for plutonium storage.
If this analysis reconfirms the value of plutonium disposition, then alternatives to the status quo approach should be evaluated. It is obvious that the MOX approach using existing reactors is not the cheap and quick path to disposition that some imagined it to be ten years ago. But do any of the discarded alternatives hold out the promise of being cheaper, quicker or more secure?
Immobilization
One of the chief advantages of plutonium immobilization over MOX is that it does not require the use of reactors and thus avoids the security risks associated with transporting MOX to reactor sites.
The prospects for immobilization in the United States are dependent on the prospects for a successful high-level waste vitrification program. The U.S. HLW vitrification program suffered a major setback in 2000 when Westinghouse abandoned the only process it had under development for pretreating HLW prior to vitrification in the Defense Waste Processing Facility (DWPF), forcing DOE to search for an alternative strategy. DOE's preference was to dispose of as much cesium-137 at the Savannah River Site as possible in a concrete grout, thus avoiding the need for vitrification at all; however, a recent court decision will in all likelihood preclude that option, requiring DOE to seriously pursue a means of vitrifying the HLW for geologic disposal. If the DWPF is able to operate at full capacity, it could immobilize up to 5 MT of plutonium per year through the can-in-canister method; thus it would only take about seven years to complete the disposition of 34 MT once the DWPF began full-scale HLW vitrification, compared to 17 years for the current MOX scenario.
And DOE has not entirely abandoned R&D in plutonium immobilization technology. According to DOE Undersecretary Bob Card, DOE continues to "actively consider immobilization as one of the alternatives for a significant portion of our surplus impure Pu and [has] continued to be engaged in technical evaluations of it even after the SRS project was cancelled."[3]
The main obstacle to a resurrection of the immobilization option is, as always, Russian opposition. The challenge is to persuade Russia that immobilization is in its security interest, and that an accident at an immobilization facility is unlikely to result in the kinds of economic damages that a MOX reactor accident would. The United States is taking a hard line in almost every other negotiation with Russia; why has it caved in so readily on the MOX issue?
Dilution and Direct Disposal
Dilution and direct disposal of plutonium has the virtue of simplicity, but forces a near-term confrontation with local groups opposed to nuclear waste disposal. WIPP is the only operating geologic repository in the United States, and has the capacity to accept a considerable amount of plutonium in addition to the TRU waste that it was designed to accept. 34 MT of plutonium, packaged appropriately in 55-gallon drums, would take up approximately 35,000 cubic meters, only a fraction of WIPP's 175,600 cubic meter capacity. DOE currently allows the disposal of materials containing less than 10 percent plutonium in WIPP; however, for such a large quantity of plutonium, greater dilution may be desired.
Other direct disposal options, such as deep boreholes or under the deep seabed, merit reconsideration, although the environmental opposition to these approaches would likely be formidable. In the latter case, the involvement of the global commons and the possible violation of the London Convention on nuclear dumping present even greater obstacles.
[1] The total amount of plutonium that Russia committed to dispose of under the PMDA is actually 38 MT, because Russia wants to conceal the isotopic composition of its warhead plutonium by blending it with about 12% reactor-grade plutonium.
[2] Joint U.S.-Russian Working Group on Cost Analysis and Economics in Plutonium Disposition, "Scenarios and Costs in the Disposition of Weapon-Grade Plutonium Withdrawn from Russia's Nuclear Military Programs," April 2003, footnote 26, p. 13.
[3] Robert Card, DOE Undersecretary, personal communication, August 11, 2003.
