The Advanced Fuel Cycle Initiative and Global Nuclear Energy Partnership Programs
From the first introduction of nuclear power, the management of spent nuclear fuel, especially the highly radioactive components, has been a concern. Three main issues underlie this concern: the disposal of nuclear wastes, the reduction of opportunities for nuclear weapons proliferation, and the long-term supply of fissionable material for nuclear fuel. A central question in dealing with these issues is whether to close the nuclear fuel cycle by reprocessing the spent fuel and recycling its components or to employ a once-through fuel cycle, treating spent fuel as waste. Various nations have answered this question differently.
In 1976, the United States decided to suspend plans for reprocessing and recycling plutonium due to the potential risk of proliferation. Then in 1979, it changed its policy, deciding to defer reprocessing indefinitely and to pursue the once-through fuel cycle. Some countries, notably France, the United Kingdom, Germany, the Soviet Union, and Japan, continued to reprocess plutonium. In France, the recovered plutonium is now recycled once in the form of uranium-plutonium mixed oxide (MOX) fuel to produce power while the rest of the minor actinides, primarily neptunium (Np), americium (Am), and curium (Cu), and the fission products from the spent fuel are stored until a repository is available. Other isotopes such as krypton (Kr) and iodine (I) are released as effluent.
All nuclear fuel cycle options, including closed fuel cycles, require the capacity for permanent disposal of high-level wastes. The National Research Council (NRC) recommended in 1957 that deep geologic isolation would be a suitable approach for disposal (NRC, 1957). Other nations have adopted the same view. However, no nation yet has a fully functioning geologic disposal operation for high-level radioactive waste.
Since 2002, the United States has been conducting a program of spent fuel reprocessing research and development (R&D), in part to consider alternative spent fuel management options. This program is built on earlier work funded by DOE that was evaluated by the 1996 NRC report Nuclear Wastes: Technologies for Separations and Transmutation. The Advanced Fuel Cycle Initiative (AFCI) was the program under which DOE was carrying out its long-term direction to recycle nuclear fuel waste. In February 2006, 5 months before the committee’s first meeting, the United States announced a change in its nuclear energy programs. The FY 2007 budget request included work on recycling that would be done under a new effort, the Global Nuclear Energy Partnership (GNEP). This new effort would incorporate the AFCI as one of its activities. If the recycling R&D program leads to successful deployment, GNEP would eventually require the United States to be an active participant in the community of nations that recycle fuel, because part of the GNEP program has some nations recycling the nuclear wastes for other user nations. The presumption is that by having only a few supplier nations carry out the enrichment and recycling for many others, nuclear power could be made economically attractive to the user nations and, at the same time, the number of locations where enrichment and recycling are carried out would be minimized, reducing opportunities for diversion of fissionable material and misuse of fuel cycle facilities and technologies.
In this way, the AFCI/GNEP program under review by this committee is being conducted in the face of change and uncertainty in U.S. policies for the disposition of commercial spent fuel and high-level waste. One effect of this uncertainty is to make more difficult the acquisition of clear and complete program documentation. To develop the necessary information for its evaluation, the committee has drawn on interviews with individuals from DOE, the Nuclear Energy Institute (NEI), the Electric Power Research Institute (EPRI),
The committee also saw copies of slides presented at a GNEP panel session at the U.S. Nuclear Regulatory Commission (USNRC) on March 15, 2007, and GNEP-relevant presentations at the American Chemical Society annual meeting on March 27, 2007. The GNEP Technology Development Plan (TDP) was released on July 25, 2007, after the committee began its peer review stage. Because TDP said that the plans it described did “not necessarily reflect the views and decisions of the Department of Energy,” the committee could not accept it as DOE policy and had to use other references (e.g., reports of the Organisation for Economic Co-operation and Development (OECD) in evaluating the technical aspects of fuel recycling.
In the balance of this chapter, the committee first describes the AFCI program as it existed until 2006 and then describes and evaluates its successor, GNEP. The chapter concludes with the committee’s findings and recommendations.
Proliferation Concerns and Efficient Use of Nuclear Fuel: The AFCI Context
The United States rejected the idea of recycling spent nuclear fuel during the 1970s because the then-available methods all produced separated plutonium, which can be purified relatively easily into material to make a fission bomb. Similarly, the uranium enrichment process can be misused to generate enough highly enriched uranium to make nuclear weapons. The United States and other countries that are members of the International Atomic Energy Agency (IAEA) have worked to reduce proliferation risks and to rectify the shortcomings identified by the International Nuclear Fuel Cycle Evaluation (IAEA, 1980).
Since the time of that decision not to recycle, other recycling processes have been under development that do not yield separated plutonium. In the United States, processes were worked on, beginning in 2002, under the AFCI, which itself had grown out of the Accelerator Transmutation of Waste program, initiated in 1999. This effort was under the direction of DOE’s Office of Nuclear Energy (NE). The AFCI program was created with the following objectives (DOE, 2005; 2006c, p. 3):
AFCI technology development focuses on reducing the long-term environmental burden of nuclear waste, improving proliferation resistance, and enhancing the use of nuclear fuel resources. The program has one major objective associated with each of these three considerations. The AFCI Program also has a fourth “system management” objective that emphasizes safe and economic nuclear materials management, integrating all of the above considerations.
It is of particular importance to note that the AFCI was to provide an alternative to building the multiple repositories that might be needed for the once-through fuel cycle and to support a growing role for nuclear energy. The published DOE GNEP strategy does not consider the possibility of Yucca Mountain being rejected or of it being accepted and its capacity significantly increased for the storage of more spent fuel. AFCI was to inform the Secretary of Energy about the need for a second repository as early as January 1, 2007, and no later than January 1, 2010, because according to the Nuclear Waste Policy Act, the Secretary is required to report to Congress on that schedule.
To meet its objectives, AFCI examined four fuel cycle strategies (DOE, 2006c, p. 11):
The current U.S. strategy is once-through—all the components of spent fuel are kept together and sent to a geologic repository for disposal.
The second strategy is recycling in thermal reactors only. Uranium in spent fuel and depleted uranium would be disposed of as low-level waste. Transuranic elements, such as plutonium and neptunium, would be recycled several times, deferring the need for a second geologic repository. However, eventually transuranic elements would accumulate and would require geologic disposal. Long-lived fission products would also go to geologic disposal. Short-lived fission products would be first stored and ultimately disposed of as low-level waste. This strategy would use existing types of nuclear power plants, which are all thermal reactors.
The third strategy is sustained recycle with a symbiotic mix of thermal and fast reactors, recycling transuranic elements from spent fuel repeatedly until destroyed. The introduction of fast reactors makes this strategy sustainable from the repository standpoint; the accumulation of transuranic elements during repeated recycle passes is controlled and limited by fast reactors serving as transuranic element burners. Essentially no transuranic elements would go to geologic disposal, only processing losses. Uranium and fission products would be disposed of as with thermal recycling. This strategy requires a significant, but minority, fraction of nuclear power plants to be fast reactors, which are being researched by the Generation IV Nuclear Energy Systems initiative.
The fourth strategy is sustained recycle with fast reactors, recycling both uranium and transuranic elements repeatedly until all energy is extracted. Phasing out thermal reactors in favor of fast reactors means that all types of uranium ultimately serve as fuel; thus this strategy is sustainable both in terms of repository constraints and in terms of uranium ore resources. Essentially no uranium or transuranic elements would be wasted, only processing losses. As with other recycle strategies, long-lived fission products would tend to
go to geologic disposal; short-lived fission products would be stored and ultimately disposed of as low-level waste after sufficient decay. This strategy would use Generation IV fast reactors.
AFCI envisioned that for all fuel cycles, long-lived fission products and residual transuranics would go to geologic disposal. For the last three fuel cycles, short-lived fission products would be managed separately to allow decay heat levels to drop before disposal as waste, either into a high-level waste geologic repository after several decades of interim storage or as low-level waste after approximately 300 years’ storage. Large inventories of transuranics would reside in the fuel cycle. Depending on the future evolution and use of nuclear energy, particularly if nuclear energy is replaced in the longer term with other energy sources, most of these transuranics could also require geologic disposal when the fast reactors are decommissioned.
The newer recycling processes would, if adopted, impact security in a number of ways. To help protect against the threat of concealed diversion of fissionable material, keeping other materials mixed with plutonium increases the effectiveness of safeguards containment and surveillance measures but may complicate material accounting. Avoiding the separation of pure plutonium is beneficial because it may increase the mass, bulk, and radioactivity of the material and can shift the handling of the material into less accessibe locations, such as hot cells. At the same time, the radioactivity of the plutonium plus actinides is not significantly higher than that of just plutonium itself. Moreover, separation of plutonium plus actinides does not preclude its use in weapons. Although weapons made from the unseparated material may be less powerful than those made from material meant to be put into weapons, the effects would still be devastating.
The programs that would eventually become AFCI received funding of $68.7 million in FY 2001, $77.2 million in FY 2002, and $57.3 million in FY 2003. In FY 2004, AFCI officially came into existence and was funded at $65.8 million in FY 2004, $66.4 million in FY 2005, and $78.4 million in FY 2006 (see Table 1-1). Beginning in FY 2007, DOE requested that the AFCI program be subsumed in a larger program, GNEP, described below, and requested $243 million for the AFCI account.
OVERALL PROGRAM DESCRIPTION
The goals of DOE’s GNEP program appear to consist of what DOE terms “objectives” and “criteria.” In its GNEP Strategic Plan (DOE, 2007, pp. 1-10 and 2-10), DOE says that in order to
enable the expansion of nuclear energy for peaceful purposes and make a major contribution to global development into the 21st century, the United States seeks to pursue and accelerate cooperation to:
Expand nuclear power to help meet growing energy demand in an environmentally sustainable manner.
Develop, demonstrate, and deploy advanced technologies for recycling spent nuclear fuel that do not separate plutonium, with the goal over time of ceasing separation of plutonium and eventually eliminating excess stocks of civilian plutonium and drawing down existing stocks of civilian spent fuel. Such advanced fuel cycle technologies would substantially reduce nuclear waste, simplify its disposition, and help to ensure the need for only one geologic repository in the United States through the end of this century.
Develop, demonstrate, and deploy advanced reactors that consume transuranic elements from recycled spent fuel.
Establish supply arrangements among nations to provide reliable fuel services worldwide for generating nuclear energy, by providing nuclear fuel and taking back spent fuel for recycling, without spreading enrichment and reprocessing technologies.
Develop, demonstrate, and deploy advanced, proliferation resistant nuclear power reactors appropriate for the power grids of developing countries and regions.
In cooperation with the IAEA, develop enhanced nuclear safeguards to effectively and efficiently monitor nuclear materials and facilities, to ensure commercial nuclear energy systems are used only for peaceful purposes.
The charge to the committee concerns the technical, scientific, economic, and management aspects of the GNEP program. Therefore, it has focused primarily on the second and third objectives. Though the fifth objective is also within the committee’s purview, DOE appears to be in only the early stages of formulating a plan for this work, so the committee has not attempted to evaluate it.
Questions of international collaboration lie outside the charge of this study. It is worth noting that the committee learned of efforts to establish discussions with other countries, notably to initiate collaboration with the Russian Global Nuclear Infrastructure (GNI) (WNN, 2007). It is unclear how well the GNI goals fit with those of GNEP. In addition, the committee learned from some of its outside expert consultants about the challenges surrounding the international aspects of bringing GNEP to reality, and there are some aspects of international interactions that do have a direct bearing on the response to the charge. These will be addressed in a later section.
DOE’s strategic plan for GNEP contains the following criteria:
Proliferation/safeguards risk. “The risk of non-peaceful use of the civilian nuclear fuel cycle comes from two principal sources: (1) a nation wanting to advance toward the capability to build nuclear weapons in a shorter period of time and (2) a terrorist group wanting to divert nuclear materials to quickly fabricate and explode an improvised nuclear device or a dirty bomb. GNEP aims to address both of these issues by providing incentives to forego enrichment
and reprocessing facilities, and by eliminating over time excess stockpiles of civil plutonium.” (DOE, 2007, p. 2-10)
Proliferation prevention. “Preventing the spread of commercial nuclear technology does not by itself prevent the spread of weapons capability…. The plutonium contained in spent fuel discharged from a light water reactor is not considered ‘weapons grade.’ However, plutonium separated from spent nuclear fuel could be fashioned into a weapon and achieve a nuclear yield of some magnitude…. While safeguarding bulk-handling facilities will continue to pose significant technical challenges, advances have been made in developing processes that are easier to safeguard, allow improved materials accountability, are more resistant to terrorist threat, and offer the possibility of placing a much reduced burden on our waste disposal facilities. However, there is no technology ‘silver bullet’ that can be built into an enrichment plant or reprocessing plant that can prevent a country from diverting these commercial fuel cycle facilities to non-peaceful use…. GNEP seeks to develop advanced fuel cycle technology for civil purposes, centered in existing fuel cycle states that would allow them to provide fuel services more cheaply and reliably than the other states could provide indigenously.” (DOE, 2007, p. 3-10)
Terrorist threat reduction. “In the most general terms, GNEP seeks to eliminate over time excess stocks of separated plutonium and reduce stocks of spent fuel worldwide, thereby strengthening nuclear security worldwide. In more specific terms, a key objective with respect to any GNEP recycling facility is to deny access to fissile nuclear materials of critical mass that could be readily made into a nuclear device. Supportive policies can be implemented in this regard: (1) minimize transportation; keep fissile materials inside one integrated facility from the time used fuel enters until recycled material leaves; (2) maintain a mixture of fissile material with non-fissile material in a ratio that is not easily useable as a weapon; (3) use advanced safeguards and security techniques; and (4) maintain a goal of minimizing the buildup of, and eventually eliminating, stockpiles of separated civilian plutonium or its near equivalent.” (DOE, 2007, p 3-10)
Reduced repository burden. “Commercial spent nuclear fuel can either be disposed of directly into a repository (e.g., Yucca Mountain in the U.S.) or reprocessed/recycled and the byproduct high level waste sent to a repository…. The full benefit envisioned for the separations process in GNEP anticipates substantial repository benefits (by separating out all the actinides) and a reduction in liquid process waste. The most significant repository benefits can be achieved by removing the very long-lived minor actinides and recycling them as part of the fuel for fast reactors. To obtain a repository capacity increase ranging from one to two orders of magnitude and allow Yucca Mountain to satisfy our repository needs for the remainder of the 21st century it will be necessary to remove and fission through recycle the very long-lived minor actinides. Further repository benefit can be achieved by removing the fission products cesium and strontium from the high level waste stream and allowing them to decay separately. These elements have a relatively short half life and after decay could be disposed of as low level waste. Additionally, removing the technetium and fixing it in a matrix with the cladding hulls could reduce the possibility of this fission product migrating away from the repository area. DOE has been conducting work on processes to achieve all of these additional advanced partitioning objectives as well as work on how to recycle and consume these materials in a fast spectrum reactor. To date these efforts have been carried out as part of the Advanced Fuel Cycle Initiative, and it is proposed to continue this work as part of the broader GNEP initiative. Similar work is being carried out in Japan, France, and Russia with promising results.” (DOE, 2007, p. 4-10)
Assured fuel supply. “The U.S. seeks to encourage the world’s leading nuclear exporters to create a safe, orderly system that spreads nuclear energy without proliferation. States that refrain from enrichment and reprocessing would have reliable access at reasonable cost to fuel for civil nuclear power reactors…. The implication for the U.S. is that if we are going to participate in assuring access to nuclear fuel and, in the longer term, spent fuel services to these countries as they enter the nuclear arena, the U.S. must have the capability to provide the needed fuel cycle services—capability that we do not currently possess. Our fuel cycle technology should also build our ability, and those of our partners, to establish and sustain ‘cradle to grave’ fuel service or leasing arrangements over time and at a scale commensurate with the anticipated expansion of nuclear energy by helping in a major way to solve the nuclear waste challenge. (DOE, 2007, pp. 4-10 and 5-10)
Capability and leverage. “The GNEP vision has been well received by the international nuclear community, particularly among the leading fuel cycle states. Sustaining and building on that enthusiasm depends on the U.S. ability to get back in the commercial nuclear business and assume an active role. Participating fully in that business is essential in order to shape the rules that apply to it…. We have a vision of a future world that can universally enjoy the benefits of safe, economical, emission-free energy; and we have programs and plans to put the U.S. back in the nuclear energy game in a leadership role. Access to our market is itself a form of leverage.” (DOE, 2007, p. 5-10)
Three facilities are key components of the GNEP program as currently planned: (1) a nuclear fuel recycling center or centralized fuel treatment center (CFTC), (2) an advanced sodium-cooled burner reactor (ABR), which is a fast-neutron reactor, and (3) an advanced fuel cycle facility (AFCF). At the CFTC, spent fuel would be separated into specific waste streams, some of which would go to the ABR (the CFTC is sized to fuel many ABRs, as discussed later in this report) as transmutation fuel and others of which would go to a repository or long-term storage or be disposed of as low-level
waste. Initially the transuranics and much of the uranium would go to the AFCF, which would turn those streams into transmutation fuel in the form of lead test assemblies, send its waste to a repository, and accept spent fuel from LWRs as well as partially transmuted fuel from the ABR. Subsequently, once the lead fuel designs were qualified, fuel fabrication would be located at the CFTC to minimize the transport of materials. The ABR would need, in addition to the fuel from the CFTC and the AFCF, some start-up fuel, whether uranium or plutonium. A principal function of the ABR would be to fission transuranic elements, while a secondary function would be to produce electricity.
The DOE has proposed that the CFTC be able to handle 2,000 to 3,000 metric tonnes (MT) per year of spent fuel. (Note that the current U.S. fleet of 104 operating reactors produces only 2,000 MT/yr of spent fuel, and 56,000 MT is already in storage.) At the time of the writing of this report, the latest information the committee had was that the baseline process was UREX+1a, although some other comparable separation technology, most notably pyroprocessing, may be adopted at a later stage. The ABR thermal power is planned to be 500 to 2,000 MWth. Both facilities should be capable of being licensed by the USNRC, although it is not clear if licensing is part of the GNEP plan. The locations of GNEP facilities have not been determined, although various expressions of interest and environmental impacts are being assessed.
GNEP as currently proposed has DOE as the leader for the AFCF and private companies as leaders for the CFTC and ABR. The strategic plan states that “a GNEP goal is to develop and implement fuel cycle facilities in a way that will not require a large amount of government construction and operating funding to sustain it” (DOE, 2007, p. 6-10). According to DOE, industry has filed expressions of interest (EOIs) that show a potential willingness to invest large sums of private funds to build and operate GNEP fuel cycle facilities. Because the EOI responses include proprietary information, the committee was not allowed to review them. The plan does recognize, however, that federal support for R&D and incentives is needed to ensure that the long-term goals are met. The strategic plan does not elaborate on the character or scale of the federal incentives, nor does it say how reprocessing and recycling costs, including potential subsidies for fast reactors, would (presumably) be passed on to nuclear electricity consumers in the form of fees or other charges to recover private investors’ initial investments.
Since the federal government is funded in FY 2007 through a Continuing Resolution (CR), the complete redirection of AFCI into GNEP is proceeding at a slower pace than had been planned. The FY 2007 CR appropriation agreement funds AFCI/GNEP at the level of $167.5 million, including the authority to redirect other programmatic funds to this initiative. The administration has requested $395 million for FY 2008. A decision by the Secretary of Energy on the future of GNEP—whether to conduct more R&D or proceed to commercial scale—is scheduled for June 2008.
ANALYSIS AND EVALUATION OF THE PROPOSED GNEP PROGRAM
The results of the committee’s evaluation of the technical, scientific, economic, and management aspects of the GNEP program are presented in this section. The evaluation looked at the technical and scientific options available for accomplishing some of the GNEP goals, particularly minimizing the burden on domestic nuclear waste repositories.
Reducing the Nuclear Waste Repository Burden
Under the Nuclear Waste Policy Act of 1982 (NWPA), Congress mandated that high-level nuclear waste be put into a geologic repository to be managed by DOE. The 1987 Nuclear Waste Policy Amendments Act directed DOE to evaluate only the Yucca Mountain site in Nevada for its suitability as a geologic repository. Disposal was to begin in 1998 but was delayed for several reasons, including strong opposition by the state of Nevada, technical issues associated with the site, the rewriting of EPA standards as the result of lawsuits and congressional action, insufficient appropriations from the Nuclear Waste Fund, and differences of opinion between the two political parties. The site was approved by the President and Congress in July 2002, though final approval rests with the USNRC, which grants construction and waste acceptance licenses. Program delays have continued for several reasons, including design changes, inadequate quality assurance, and management problems relating to the Yucca Mountain site. DOE is now scheduled to submit a license application in June 2008. If DOE keeps to that schedule, the USNRC’s review of the license application should be completed by 2012 if the USNRC meets certain reporting requirements. Spent fuel could then be accepted starting in 2017, but even DOE has little expectation of meeting that schedule. Meanwhile, spent fuel waste continues to be stored at reactor sites.
The total volume of nuclear waste from conventional LWRs is large enough to require serious attention. The thermal and radiation characteristics of the waste are the main concerns in designing the repository and determining its capacity. The NWPA established a capacity limit of 70,000 MT of waste for Yucca Mountain, 63,000 MT of which is designated for commercial spent fuel and the remainder for defense wastes. By the end of 2006, about 56,000 MT of spent fuel had been generated by U.S. nuclear power plants, and that inventory is growing at approximately 2,000 MT/yr. If all operating reactors receive 20-year license extensions, the total amount of waste from the current U.S. fleet could exceed 120,000 MT.
Although a statutory limit has been placed on the repository capacity, there is a wide range of opinion about
. the technical limits on the capacity. The technical limits of Yucca Mountain capacity are determined by the total area available with suitable geologic characteristics and by two criteria related to the management of heat from the decay of spent fuel. Significant uncertainty surrounds both the area available that is suitable for repository use and the maximum achievable areal loading. The draft Environmental Impact Statement identified 4,200 acres that possess four characteristics required for use as repository space: 200 meters of overburden, consistency of elevation and dip with the upper block, distance from the saturated zone, and favorable excavation characteristics (CRWMS, 1999). At the current design loading of 60 MT per acre, 4,200 acres would be large enough to store 252,000 MT of spent fuel. Larger areal loading might be possible for fuel with greater burn-up (the extent depending on radiation dose calculations), a trend already under way in the nuclear industry. A study by EPRI likewise suggested that with revised repository design, areal loading could increase by a factor of 2 or 3 (Kessler, 2006), although the study did not take into account limits imposed by geologic considerations.
Areal loading could increase much more if advanced fuel cycle technology, such as that envisioned by GNEP, is used. According to Wigeland and others (2006) the repository’s capacity could be increased by reducing the amounts of short-lived cesium (Cs) and strontium (Sr) fission products as well as by lowering the amount of transuranics (TRUs) (Pu, Np, Am, and Cm) in the wastes reaching the repository. For example, the repository’s areal capacity could be increased by a factor of 4.4 if the fractions of Pu, Np, Am, and Cm in the waste were decreased to 10 percent of their original values and by a factor of 10 if the fractions of Cs and Sr were also decreased 10-fold. Decay heat from Cs and Sr can be reduced 10-fold by a combination of interim storage and repository ventilation for 100 years; with approximately 300 years of storage, radiation levels drop sufficiently that disposal as low-level waste might be possible. It must be noted that removing Cs and Sr brings up a new siting issue: where and how to store such wastes for several decades to hundreds of years.
Considerations other than areal loading may dominate the Yucca Mountain decision, however. Detailed characterization would be required to determine what fraction of its space also meets other geological constraints (including spacing from fault and fracture zones) required for repository use. The USNRC must also consider other criteria, including public health and safety, in deciding whether to grant a license for the Yucca Mountain repository. It is difficult to predict when, if ever, any of these options for the use of Yucca Mountain might become reality. Significant uncertainty surrounds the maximum technical capacity of the Yucca Mountain site. Geologic studies may limit this capacity significantly. These and other issues will be considered by the USNRC at an uncertain date in the future, and it may or may not ever grant a license for the repository. If it does, the available evidence suggests that the capacity of Yucca Mountain exceeds the current statutory limit of 70,000 MT. If its opening is delayed, spent fuel can be stored using dry-cask storage. Spent nuclear fuel that has spent 5 years cooling in on-site water pools can be put into passively cooled casks, each holding approximately 10 MT of waste. There is general agreement and approval by the USNRC that such a scheme would provide safe, secure storage for at least 100 years.
As noted earlier, one goal of the GNEP program is to reduce the burden on the repository by reducing the volume of waste it must handle. Given the uncertainties discussed above, however, it is difficult to judge precisely when the technical need for additional repository capacity will arise. Therefore, the committee concludes that the need for an accelerated program to deploy commercial-scale reprocessing and fast reactors to reduce the nuclear waste repository burden has not been established. In particular, the near-term need for deployment of advanced fuel cycle infrastructure to avoid a second repository is far from clear.
But even if a second repository were to be required in the near term, the committee does not believe that GNEP would provide short-term answers. As the later discussion will show, however, the committee considers the DOE-preferred option—the GNEP program—also to be a very long-term effort, measured in decades, and very expensive, measured in tens of billions of dollars (or more). Its approval and survival will depend heavily on its broad technical and societal support, steady and continued funding, and effective management. GNEP will need positive actions from several successive presidential administrations and Congresses. With respect to management, GNEP needs a partnership and a business plan agreed on by industry, DOE, and participating foreign countries. To sustain such support, there needs to be more clear evidence that GNEP is preferable to the other options for expanding deep geologic disposal capabilities.
In the committee’s view, the GNEP concept rests on a set of technologies that present very challenging development and engineering issues. Moreover, it is not clear that all of the relevant options had been evaluated before arriving at the program’s preferred choices. Below, the committee discusses these issues, which relate to recycling methods, advanced fuel development, and fast neutron reactors.
DOE is currently examining two methods for recycling nuclear fuel that do not isolate plutonium: UREX+ (in effect, a collection of methods) and pyroprocessing. The various separation steps of the UREX+la process were demonstrated at Argonne National Laboratory and reportedly achieved better than 99.999 percent extraction efficiency for U. This test used irradiated fuel from the Cooper nuclear station
power plant in Nebraska. The committee understands that a full, integrated project using all the steps has not yet been carried out. A preconceptual design for an AFCF has been completed. In addition, the AFCI has been developing pyrochemical (or pyroprocessing) methods for the treatment of both legacy LWRs and future advanced reactor fuels. While the UREX+ processes work with oxide fuels, pyroprocessing deals with metallic fuels or oxide fuels, with an additional processing step to reduce the oxide to metal. With oxides, “the pyrochemical reduction (PYROX) process is being developed for treatment of Generation IV oxide fuels. High-capacity reduction experiments and improvements in cell design have been completed.” (DOE, 2006c, p. 39)
Significant technical problems remain to be solved before either process can be considered to have been successfully demonstrated. One of GNEP’s most important goals is showing that TRUs can be consumed, a satisfactory alternative to requiring a means to store them. Special attention must be given to the radiation level of recycled fast reactor fuel and the constraints it will impose during shipment and handling by plant operators. As noted elsewhere, however, it is very unclear whether UREX+ will be able to deal with the high decay heat of fast reactor fuel. Pyroprocessing may better satisfy those needs because it is more suited for remote handling and it can be carried out in much smaller facilities, which could be co-located with fast reactors. It might be best to accelerate the development of pyroprocessing so that it can deal with both water reactor and fast reactor spent fuel.
Beyond these two processes, however, an OECD report (OECD, 2006, p. 11) explains that “given the wide range and flexibility of advanced fuel cycles under development … strategic choices will be based on the priorities of policy makers which reflect continuing specific criteria such as characteristics of available waste repositories, access to uranium resources, size of the nuclear power program, and social and economic considerations.” The committee has seen no evidence that GNEP has explored those options. Indeed, potential GNEP partners are considering other fuel cycles; these cycles need to be assessed for various projected scenarios of growth in nuclear power production. If the G in GNEP is to be taken seriously, the selection of technologies and their allocation among the partners must surely be the result of common agreement.
Advanced Fuels Development
TRU fuels are central and problematic in GNEP technology because “no [reactor] concept can be considered seriously if the appropriate fuels are not defined and proven, i.e., characterized, fabricated, irradiated, and reprocessed” (OECD, 2002, p. 298). A presentation by Frank Goldner2 and a report by the Nuclear Energy Agency (OECD, 2002) both provide an excellent accounting of how difficult TRU fuel development will be. Goldner pointed out that for oxide fuels, the effect of group TRU on the fabrication process is unknown, as is the effect of lanthanides, and a large-scale fabrication amenable to hot-cell operations must be developed. For metal fuel, he noted that large-scale fabrication without loss of Am must be demonstrated, that fuel-clad interactions at high burn-up must be investigated, and that the effect of lanthanides on fuel cladding chemical interactions must be addressed.
These technical challenges are compounded by the need to repeatedly refabricate the fuel. Although GNEP documents do not specify the number of expected fuel recycles, other sources illustrate the scope of the issue. For example, one report (OECD, 2002, p. 41) says that an
actinide (or TRU) burner requires a fuel cycle which allows the fuel to be recycled many times…. For a maximum burn-up of 25% and recycle intervals of 6 years, it takes 96 years … to achieve a hundred fold waste mass reduction.
On page 21 of the same report, it is noted that
[because] transmutation systems involve unusual fuels with high decay heat and neutron emission … a significant effort is required to demonstrate the manufacturability, burn-up behavior, and ability of reprocessing of these fuels. In order to reprocess via pyrochemical methods they would have to tolerate from ten to more than twenty times higher decay heat than those encountered in the pyrochemical reprocessing of fast reactor fuels.
In their presentations to the committee, DOE personnel confirm that no TRU fuel fabrication has been achieved with prototypic materials obtained from actual separation processes and using prototypic fabrication processes suitable for remote operations. DOE reports that it has fabricated mixed actinide fuel successfully and that test fuel pins have been manufactured to permit placement in test reactors. In-reactor testing is in progress at the Phenix fast reactor in France. LWR mixed-oxide fuel pellets containing Np and Pu have been irradiated in the advanced test reactor (ATR). DOE has fabricated and tested inert mixed fuels using magnesium and zirconium oxides, MgO-ZrO2, as well as microdispersion pellets of MgO-ZrO2-PuO2.3 DOE is also working on advanced fuels: tristructural isotropic (TRISO), a multilayer micropellet form, for gas-cooled reactors; nitride; sphere-pac; and dispersion fuels. DOE fabricated a variety of test samples of candidate matrix materials and shipped them to the Phenix reactor for irradiation.
For these reasons, the committee regards the development and qualification of advanced reactor fuels as a major technical challenge. Because of the time required to test the fuel
Frank Goldner, DOE, “GNEP transmutation fuel development,” Presentation to the 2007 Regulatory Information Conference on March 15, 2007.
Three members of the committee feel very strongly that the thermal recycling of inert matrix fuel should have priority over GNEP multirecycling in sodium fast reactors; their rationale is summarized in Appendix B. The other committee members believe the concept deserves consideration but are not willing to sponsor it because it may be premature.
through repeated refabrication cycles, achieving a qualified fuel will take many years.
For its GNEP program, DOE has selected for first consideration the sodium-cooled fast reactor (SFR). Other reactor concepts identified by the Generation IV Technology Roadmap (Chapter 3) and possible GNEP candidates are (1) the lead-cooled fast reactor (LFR), which would also encompass reactors using alloys of lead with other elements; (2) the gas-cooled fast reactor (GFR); (3) the supercritical water-cooled reactor (SCWR); (4) the very-high-temperature reactor (VHTR); and (5) the molten salt reactor (MSR). In its analysis, DOE notes that the SFR and VHTR are the most extensively studied reactors. Because the SFR can perform transmutation effectively and is relatively mature, the GNEP plan has proposed it as the baseline case and presumably the first fast reactor to be built for the overall GNEP program.
SFRs have some important characteristics that make them attractive for development and deployment, including flexibility with respect to mission (e.g., electricity production, breeding of fissile material, or transmutation), high efficiency, and some safety advantages over LWRs, even as they have their own vulnerabilities. Of course reactor safety is a complex issue, and other safety advantages belong to LWRs. The choice of the SFR over other fast reactor options and thermal recycle options (inert matrix fuels for LWRs and deep-burn fuels for VHTRs) should be considered in light of the history of SFRs. There is indeed a several-decade-long history of experience with these reactors dating to the experimental breeder reactor (EBR I), although fewer than 20 have supplied electricity. Accidents involving sodium can be serious, even disastrous, and there have been notable accidents with sodium-cooled reactors. A year after the MONJU reactor went on line in 1994 in Japan, it suffered a sodium leak and has remained closed ever since. The French Superphenix, a 1,200-MWe fast sodium reactor, the largest ever built, had many sodium leaks; it was closed for 2 years in the 1990s and finally shut down altogether in 1998. This plant operated at full capacity for only 174 days. There is no definite announced date for its restart. The outlook for the sodium-cooled fast reactor Fast Flux Test Facility in Hanford, Washington, and the integral fast reactor (IFR) at Idaho Falls is much better. In particular, the IFR demonstrated very high metallic fuel burn-up, is inherently safe, and introduced the important step of electrorefining to pyroprocessing (Hannum, 1997).
Other fast reactors have their own vulnerabilities. Lead-cooled reactors have been used to power Russian submarines, but lead-cooled reactors, especially those using lead-bismuth alloy because of its very low melting point, have suffered from corrosion. Whether some other noncorrosive alloy could be developed is a particularly interesting challenge for research in nuclear science and engineering and illustrates the kind of open problems that the GNEP program faces. Thermal reactor recycle options have lower risk in their reactor technologies but still face substantial transmutation fuel development issues.
The capital costs of sodium-cooled fast reactors have been estimated to be 10 to 50 percent greater than those for LWRs (Bunn et al., 2003). Fast reactors have never been deployed on a commercial scale in the United States, and research has been funded at a low level for a decade or more. This of course must be seen in light of the complicated (and discouraging) history of MONJU and Superphenix, discussed previously. Very little is said in published GNEP documents about the status of safeguards and security, management, and resources. The diffuseness of what brief discussions there are implies that much work lies ahead. The overall portrayal of the state of development of fast reactors, even of the somewhat-more-studied SFRs, suggests that the judicious course of action now would be to study and develop the prototype designs, at most at the engineering scale and presumably with as many options as possible for reactor types and designs kept open at this stage. This suggested direction is inconsistent with the GNEP Strategic Plan (DOE, 2007).
The Generation IV program developed criteria for evaluating reactor technologies (Table 3-1), but to the committee’s knowledge, these evaluation criteria were not applied in selecting the SFR. The lack of analogous selection criteria for GNEP represents an important program deficiency because it means the program lacks a basis for choosing among technology options.
GNEP Program Design and Scheduling
The GNEP program emphasizes accelerated schedules. Specifically, the Strategic Plan proposes to proceed to build commercial-scale facilities and “to define a technology roadmap … that obviates the need to build engineering scale facilities” (DOE, 2007, p. 7-10). The reasoning behind the accelerated schedule was not clear from the material available to the committee. Indeed, several factors militate against a schedule-driven program design.
Most important is the long-term nature of GNEP and the current state of knowledge about its component parts. For example, the CFTC is expected to be very large—2,000 to 3,000 MT/yr of spent fuel—larger even than the brand-new Japanese reprocessing facility in Rokkashomora, an 800-MT/yr facility. The Strategic Plan indicates that the first construction would be at this large commercial scale, skipping the engineering-scale facility step. However, the Mission Need Statement suggests that the demonstration objectives for the transmutation fuels and separation technologies will require an engineering-scale facility. Moreover, other considerations—technology readiness, fuel cycle plant costs, waste volume and radiotoxicity, vulnerability to diversion or theft, and degree of support by industry, Congress, the U.S. public, and other nations—are at least as important as
schedule. They should be assessed and, wherever possible, quantified. If the proposed commitment to UREX+ at a commercial scale turns out to be the course taken by GNEP, then its technology roadmap and business plan (called for in the Strategic Plan) would have to make clear how a facility at that scale, designed for production with one technology, can also serve as a modular test bed for other commercial-scale separation technologies.
The second issue is whether commercial fast reactors would be available to consume the TRUs separated from the spent fuel of LWRs. That is very doubtful, because with present procedures, it will take a very long time to have fast reactors licensed, operating competitively, and accepted commercially as power producers. To make the GNEP closed fuel cycle a reality, fast reactors would have to account for a significant fraction of new construction in the coming decades, a scenario the committee views as completely implausible. These timing, cost, and deployment rate issues need to be addressed.
Third, the Strategic Plan does not discuss whether the demonstration facilities are to be reviewed and approved by the USNRC, although this is implied in the request for EOIs. A position on this issue, reviewed by the USNRC, would be needed before any decisions can be made about GNEP at the Secretarial level.
DOE claims that GNEP is being implemented to save the United States nearly a decade in time and a substantial amount of money. In view of the technical challenges involved, the committee believes that the opposite will likely be true. For example, going ahead with smaller engineering facilities such as a 100- to 200-MT/yr separation facility and a 50- to 100-MWe advanced burner test reactor (ABTR) could save time and money in the long run, for a number of reasons:
The engineering facilities might not require USNRC licensing and public hearings. This could save about 3 years for the CFTC and 3-5 years for the ABTR because the commercial fast reactor is anticipated to run into increased opposition.
The engineering facilities construction schedule could be shortened by 1 or 2 years because they are smaller.
The engineering facilities could cost only about one tenth as much as the full-scale facilities, and the possibility of structuring an acceptable government–industry partnership could be enhanced considerably owing to the smaller cash flow.
Engineering facilities can be modified much faster and much more cheaply than large-scale facilities. Also, they would be more appropriate for evaluating other recycling options, while large facilities would have to be more dedicated to production. Separation of spent fuel from LWRs, with appropriate treatment and storage of fission products and high-level wastes as well as recycling of fast reactor fuel, can be demonstrated much sooner.
The timing issue between CFTC and ABTR can be resolved by sizing the engineering separation facilities at AFCF so that they can handle the needs of ABTR.
The handling, storage, and packaging of fission products will be a much smaller effort for engineering facilities, and the resolution of any remaining problems will not be as difficult at a slower production rate.
The time by which commercial-scale reprocessing will be needed depends on variables that cannot now be predicted with any reasonable accuracy. In particular, the actual future deployment rate of nuclear reactors and the actual capacity of the repository would be key variables. Engineering-scale facilities allow sufficient time to pass to reduce some of the uncertainties.
The committee concludes that the case presented by the promoters of GNEP for an accelerated schedule for commercial construction is unwise. In general, it believes that the schedule should be guided by technical progress in the R&D program. If and when technical progress justifies construction of a major facility, it is the very strong view of this committee that an engineering-scale facility would be by far the safest, most effective, least risky course. And, as discussed in Chapter 6, the committee believes that DOE should commit to the construction of a major demonstration or facility only when there is a clear economic, national security, or environmental policy reason for doing so.
DOE has not yet completed a cost analysis of the alternative pathways of research, development, and deployment (RD&D) that could be pursued to achieve the goals of GNEP. Documents reviewed by the committee indicate that the only costs that have been estimated so far are those for a single path and a single scale, with no allowance for contingencies or uncertainties. While there are large uncertainties in any such effort, it appears to the committee that the costs of alternative pathways must be projected to enable regular updating and revision as more is learned and to evolve an RD&D strategy and the tactics for carrying it out. At what stage, for example, do the next-phase costs justify a decision to continue or to drop work on a process that has just emerged, apparently successful as gauged by scientific criteria, from the first laboratory level? Even at the outset, the full complement of alternative methods should be examined for several projected scenarios of growth in nuclear power production. The amounts of spent fuel, uranium needs, and the shipments of spent fuel or high-level waste to repositories should be determined as well as their volumes, radiotoxicity, and vulnerability to diversion or theft. Costs, benefits, and cash flow, including the fees that would be charged to nuclear electricity consumers, should be estimated as a function of the dates for initial deployment of commercial fast reactors, their capital costs, and their growth rate. The GNEP Strategic
Plan implies that these analyses will be part of a business plan to be provided to the Secretary of Energy in June 2008. The committee does not find it credible that such analyses, with uncertainties, can be accomplished by that time. Even implementing an effort to develop such a plan, which would imply that a credible decision can be made by June 2008, is a matter of concern to the committee.
Furthermore, it seems likely that the GNEP fuel cycle will be more costly to operate than some other options. GNEP objectives are satisfied only with transitional or sustained recycles that require partial or full participation by fast reactors. Fast reactors complicate the selection of advanced fuel cycles since their estimated capital costs are currently expected to be 10-50 percent higher than those of LWRs, according to a Harvard study (Bunn et al., 2003, p. 68). Similarly, a preliminary predecisional economic evaluation (Crozat, 2007, p. 8) shows that the cost of nuclear electricity for an SFR would be $71/MWh compared to $56/MWh for an LWR. If that difference is reasonably accurate, producers of nuclear electricity will balk at adopting fast reactors or subsidizing them through an increase in the Nuclear Waste Fund fee, which is only $1/MWh, for thermal reactors.
Finally, a thorough economic analysis should consider several questions not apparent in the work made available to the committee. For example, closed fuel cycle cost analyses seem to have been carried out without considering temporal coordination of the components of GNEP. DOE apparently fails to recognize the crucial importance of the timing of the required separation and fast reactor facilities as well as of the time required to develop qualified fuel and its recycling in fast reactors. For a number of reasons, fuel cycle costs would rise if the separation facilities are ready but the fast reactor requires many more years to be deployed. One reason is that the TRUs separated from spent fuel would have to be stored in the interim. Moreover, the GNEP program would suffer long delays from time spent qualifying new fuels with each successive recycle. The committee is concerned that the plan to move rapidly to recycling and fast reactors has no economic basis.
One international aspect of the GNEP plan falls within the purview of this study. Because the United States has far less experience with fast reactors and recycling than other nations that are potential partners in the program, it is very important to make the program a truly cooperative one, to allow American scientists and engineers to learn from the previous work of their counterparts, and to shape the research and engineering program to be as efficient a win-win program as possible for all the participating nations. For this reason it would be very desirable as GNEP goes forward to enhance the international collaboration that was initiated with the Generation IV Technology Roadmap. One example is the area of waste separation and fuel preparation. While the proposed GNEP plan names UREX+1A as the most favored and presumably first method it wishes to pursue, other nations appear to favor other methods with which they have more experience. If GNEP is to really be an international collaboration, it is crucial that all the participating nations share the knowledge and experience each accumulates as new technologies evolve.
FINDINGS AND RECOMMENDATIONS
The committee concludes that the rationale for the GNEP program, as expressed through the stated goals, objectives, and criteria, has been unpersuasive. The program is premised on an accelerated deployment strategy that will create significant technical and financial risks by prematurely narrowing the technical options. Moreover, there has been insufficient external input, including independent, thorough peer review of GNEP.
In light of the foregoing, the committee finds as follows:
Finding 4-1. Domestic waste management, security, and fuel supply needs are not adequate to justify early deployment of commercial-scale reprocessing and fast reactor facilities.
Finding 4-2. The state of knowledge surrounding the technologies required for achieving the goals of GNEP is still at an early stage, at best a stage where one can justify beginning to work at an engineering scale. However it seems to the committee that DOE has given more weight to schedule than to conservative economics and technology. To carry out or even initiate efforts on a scale larger than the engineering scale in the next decade would be inconsistent with safe economic and technical practice.
Finding 4-3. The cost of the GNEP program is acknowledged by DOE not to be commercially competitive under present circumstances. There is no economic justification for proceeding with this program at anything approaching commercial scale. Continued research and development are the appropriate level of activity, given the current state of knowledge.
Finding 4-4. Several fuel cycles could potentially form the basis for a recycling system. However none of the cycles proposed, including UREX+ and the sodium fast reactor, is sufficiently reliable and well understood to justify commercial-scale construction at this time.
Finding 4-5. The qualification of multiply-recycled transuranic fuel is far from reaching a stage of demonstrated reliability.
In short, all committee members agree that the GNEP program should not go forward as is and that it should be replaced by a less aggressive research program. Nonetheless,
the committee believes that a research program similar to the original AFCI is worth pursuing,4 for three reasons: to extend uranium resources (when and if this need arises), to greatly reduce the long-lived, high-level actinides in nuclear waste, and to improve the waste forms for disposal of high-level nuclear waste. It may be that the international aspects of GNEP will provide technical benefits to all the participants, and there may even be some benefit in regard to inhibiting proliferation and improving physical protection as well. Such a program should be paced by national needs, taking into account economics, technological readiness, national security, energy security, and other considerations. The committee envisions such a program in the following way:
Recommendation 4-1. DOE should develop and publish detailed technical and economic analyses to explain and describe UREX+1a and fast reactor recycle as well as a range of alternatives. An independent peer review group, as recommended in Chapter 6, should review these analyses. DOE should pursue the development of other separation processes until a fully fact-based comparison can be made and a decision taken on which process or processes could be carried to engineering scale.
Recommendation 4-2. DOE should devote more effort to the qualification of recycled fuel, as it poses a major technical challenge. A fast neutron test facility is needed for fast-spectrum fuel qualification; the committee recommends carrying this out using existing facilities in collaboration with international partners. Parallel development of nonfertile LWR fuel and deep-burn TRISO fuel should be pursued to reduce program risk.
Recommendation 4-3. DOE should compare the technical and financial risks with the potential benefits. Such an analysis should undergo an independent, intensive peer review, as recommended in Chapter 6. Moreover, DOE should identify program benchmarks and report regularly on its attempts to meet them.
Recommendation 4-4. DOE should bring together other appropriate divisions of DOE and other appropriate federal agencies, representatives from industry, and representatives from other nations well before any decisions are made on the technology, in order to create and exploit shared perceptions of the roles of the participants, of the states of the various technologies, and of the commitments and schedules of each of those participants. A research, development, and deployment program can succeed only if all of those participants see themselves as its co-owners and creators.
Recommendation 4-5. DOE should defer the Secretarial decision, now scheduled for 2008, which the committee believes is not credible. Moreover, if it makes this decision in the future, DOE should target construction of new technologies at most at an engineering scale. DOE should commission an independent peer review of the state of knowledge as a prerequisite to any Secretarial decision on future research programs.
In summary, the committee concludes that without first demonstrating relevant technologies at an engineering scale, there are unacceptably high financial and technical risks to commercial-scale construction of a separations facility and a fast burner reactor.
Bunn, M., S. Fetter, J.P. Holdren, and B. van der Zwaan. 2003. The Economics of Reprocessing vs. Direct Disposal of Spent Nuclear Fuel. Harvard University. December.
Crozat, M.P. 2007. Evaluating the Economics of GNEP Deployment. January 8. Distributed to the committee by DOE.
Civilian Radioactive Waste Management System (CRWMS). 1999. Engineering File—Subsurface Repository. BCA000000-01717-5705-00005 REV 02 DCN 01. Las Vegas, Nev.: CRWMS Management and Operating Contractor (M&O). ACC: MOL.19990621.0157; MOL.19990615.0230.
Department of Energy (DOE). 2005. AFCI, Comparison Report, FY 2005. May.
DOE. 2006a. Report to Congress, Advanced Fuel Cycle Initiative: Status Report for FY 2005. February.
DOE. 2006b. Mission Need Statement for GNEP Technology Demonstration Projects. March.
DOE. 2006c. AFCI, Comparison Report, FY 2006 Update. July.
DOE. 2006d. GNEP Implementation Strategy. November.
DOE. 2007. Global Nuclear Energy Partnership Strategic Plan. GNEP-167312. January.
Hannum, W.H., ed. 1997. Special Issue of Progress in Nuclear Energy 31 (1).
International Atomic Energy Agency (IAEA). 1980. International Nuclear Fuel Cycle Evaluation.
Kessler, J. 2006. Program on Technology Innovation: Room at the Mountain, Analysis of the Maximum Disposal Capacity for Commercial Spent Nuclear Fuel in a Yucca Mountain Repository. Electric Power Research Institute Report 1013523. May.
National Research Council (NRC). 1957. The Disposal of Radioactive Waste on Land, Washington, D.C.: National Academy Press.
Organisation for Economic Co-operation and Development (OECD), Nuclear Energy Agency. 2002. Accelerator-Driven Systems (ADS) and Fast Reactors (FR) in Advanced Nuclear Fuel Cycle.
OECD, Nuclear Energy Agency. 2006. Advanced Nuclear Fuel Cycles and Radioactive Waste Management, NEA 5990.
Wigeland, R.A., T.H. Bauer, T.H. Fanning, and E.E. Morris. 2006. Separations and transmutation criteria to improve utilization of a geological repository, Nuclear Technology 154 (April): 95-106.
World Nuclear News (WNN). July 3, 2007.
The dissenting view of two members of the committee is presented in Appendix A.