Proliferation Resistance Assessment of the Integral Fast Reactor

The Integral Fast Reactor (IFR) concept includes a sodium-cooled fast reactor collocated with an integrated pyroprocess fuel recycling facility. The pyrochemical processes and the inert atmosphere of the heavily shielded fuel cycle facility provide inherent proliferation-resistant features for this advanced technology. The reactor can be designed to operate with a number of different conversion factors, so that it could be used for excess plutonium consumption or as a breeder if needed for rapid expansion of energy supply. The system contains a large quantity of plutonium and minor actinides, which at all times remain in extremely hostile environments and in chemical and physical forms that would require additional processing to extract weapons-suitable material. The aqueous processing equipment and facilities to accomplish such separation would not be available on site. Transportation would not be required in the reference deployment scenario. Nevertheless, the proliferation-resistance of some parts of the system could be considerably strengthened by advanced safeguards technologies. In spite of its inherent features, international deployment of the system would probably be limited to stable countries with a strong existing nuclear infrastructure. INTRODUCTION Assessing the proliferation resistance of Argonne National Laboratory’s Integral Fast Reactor (IFR) concept has been a relatively popular pastime activity for the past 16 years. [1,2,3,4,5] This particular assessment is based on the unpublished work that went into preparing a presentation for the Nuclear Energy Research Advisory Committee’s (NERAC) Special Committee on Technical Opportunities for Proliferation-resistant Systems (TOPS) [6]. Speculation on the proliferation resistance of the concept endures because the technology continues to develop and mature, the assessment tools improve, and the possibility of applying elements of IFR technology to national problems continues to be raised. As originally conceived [7], the Integral Fast Reactor comprised a fast-spectrum, sodium-cooled, metal-fuelled reactor and a collocated fuel recycling facility that employed pyroprocessing and fully remotized metal fuel casting and assembly. No transportation of nuclear materials would be required other than the initial shipment of fuel for startup. The system would be self-sustaining, i.e., producing as much plutonium as was consumed and lost to incidental waste streams. Because of the unique fuel cycle, plutonium would remain in a highly radioactive matrix at all times in facilities that were literally inaccessible to humans at all times. In designing the system during the postInternational Nuclear Fuel Cycle Evaluation (INFCE) [8] era, robust proliferation resistance was a requirement. The fundamental assumption was that nuclear fuel recycle would be required and therefore the best approach to plutonium management was to avoid producing, storing or using it in any form that could be easily stolen or concealed, or that could be used without further refinement to fabricate a nuclear explosive. Furthermore, it was important that the process equipment and facilities could not be easily modified to produce a weapons-suitable product. The reactor concept was not tied to a particular size, but rather was envisioned to work for any size, from small modular reactors to greater-than 1000 MWe systems. During the early development, industrial participants in the Advanced Liquid Metal Reactor (ALMR) program [9] adopted the concept. Smaller designs were favored, with General Electric advocating its rail-shippable PRISM system and Rockwell International favoring its barge-shippable SAFR system. The common link was the U-Pu-Zr fuel—a radical departure from the traditional mixed-oxide (MOX) line of breeder reactor development—comprised of depleted uranium, nominally 20% plutonium and 10-wt% zirconium. The alloy components were melted in a ~1450 C furnace, electromagnetically stirred to a homogeneous mixture and simultaneously injection cast into multiple fuel-pin molds. The pins were removed from the molds, cut to length and placed in stainless-steel cladding that contained sufficient metallic sodium to provide a thermal bond in the gap between the cladding and the pin. Demonstration of remote fabrication of the fuel was not accomplished prior to cancellation of the IFR program by the Clinton Administration in 1994. However, all the processes were operated remotely in a glovebox environment and a complete set of equipment was fabricated and qualified for hot cell operation. Extensive test data were obtained on glovebox-fabricated metal fuels with irradiations in the Experimental Breeder Reactor-II (EBR-II) and the Fast Flux Test Facility (FFTF), and transient overpower tests in the TREAT facility. EBR-II operated for years with a U-Zr metal core and ternary experiments, while FFTF irradiated several whole metal fuel assemblies within its MOX core. The fuel demonstrated burnup to 20% without failure; transient testing indicated a major safety advantage over MOX. In the IFR concept, the fuel would be recycled on site using a technique that has at various times been known as pyroprocessing, electrometallurgical treatment or dry reprocessing. Completely different from aqueous reprocessing that has been industrialized as PUREX, pyroprocessing uses a molten salt in the separations process. Various mixtures of chloride or fluoride salts have been used, but all must operate in high temperature (450 C and up) and in a dry argon atmosphere. Other differences include much higher concentrations and volumes of plutonium due to reduced criticality limitations and very poor (<10) separation factors for plutonium relative to other actinides and some rare earths. Sodium and zirconium are compatible with the process, whereas they are not with conventional PUREX processes. Also, no minimum fuel cooling time is required, since there are no organic solvents to be destroyed by intense radiation. Spent fuel is chopped into short segments, arranged in a mesh basket and lowered into a molten salt electrorefiner. [A LiCl-KCl eutectic operating at 500 C has been the primary line of development.] The fuel basket becomes the anode for the electrorefining cell, with UCl3 or CdCl used as an oxidant for the sodium and the active fission products. A small potential [<1 volt] is applied, which results in oxidation of uranium, transuranics and most fission products at the anode and reduction of uranium at the cathode. There is some carryover of zirconium and noble metal fission products, but uranium decontamination factors of 100 have been demonstrated. When the ratio of plutonium to uranium becomes sufficiently high, the transuranics can be removed. This step has not been demonstrated with irradiated fuel due to policy restrictions during the 1990’s, but small-scale tests are now being planned. Various electrorefining, electrolysis and electrochemical techniques have been proposed, but with development arrested, no well-defined flowsheet has emerged. This makes nonproliferation analysis more problematic, but only marginally so, since each technique would collect a witch’s brew of transuranics in rough proportion to their relative concentration in the salt, uranium (~50%), and 1% or more of rare earth fission products. Throughout the 1990’s, it was assumed that the material would be electrorefined into a liquid cadmium cathode, but that approach now seems to be losing favor because of the difficulty in scaling to industrial proportions. The salient characteristics of the transuranic product are intense heat, radiation and neutron emission. Both the transuranic and the uranium products contain a high fraction (~20 wt.%) of adhering salt when removed from the electrorefining operation. A distillation furnace is used to separate the more volatile salt and subsequently to consolidate the metal into an ingot. These ingots from the uranium stream and the transuranic stream are broken into smaller pieces and used in proper proportion as charge to the fuel casting furnace, along with recycle scrap from previous castings. In the classic IFR concept, the actinides are quantitatively removed from the salt prior to its disposition into the waste stream. However, during the treatment of the EBR-II fuel, the plutonium has been intentionally directed to the waste stream. The salt must be discarded when its heat generation rate reaches the design limit for the process vessels or when the composition of the salt reaches a point where it is no longer molten at a sufficient margin below the prescribed operating temperature for the process. The salt, containing the bulk of the fission products, is mixed with zeolite particles at ~500 C to occlude the fission product chlorides in the zeolite structural cages. Twenty-five weight-percent glass frit is added as a binder and the resulting mixture is baked at about 900 C to transform the zeolite into sodalite, a rugged natural mineral found in some areas that would be considered potentially suitable geologic repositories. Very large monolithic waste forms—up to 50 tonnes— can be produced in this manner. The metal cladding hulls, containing some zirconium, noble metal fission products, and a small fraction of uranium (plutonium is preferentially oxidized) are collected and mixed with additional zirconium if necessary to be melted at a favorable eutectic ratio (15% Zr). The metallic waste ingot produced in this way is also considered to be high-level waste and has proven to be extraordinarily corrosion resistant. Significantly for geologic disposal, the ingot contains most of the technetium, which is insoluble in metal form. PREVIOUS PROLIFERATION ASSESSMENTS As one of the cornerstones of the IFR concept, affirmation of proliferation resistance has been necessary to the continued, albeit erratic, development of the fuel cycle technology. The envisioned system relies on active plutonium management, maintaining a large inventory of plutonium, but only as much as needed to maintain a nuclear island with power plants and a fuel cycle facility. The material would remain continuously in a sequence of highly radioactive matrices within inaccessible facilities. The cost of safeguarding the material would be compensated by the sale of electricity. Three independent assessments were critical in gaining authorization to advance the technology to the next level: Bengelsdorf (1986) when development was just getting started, Wymer (1992) as the technology was being readied for the demonstration phase, and the 1999 assessment by the Department of Energy’s Office of Arms Control and Nonproliferation when a negative report would have killed the little remaining development of the technology in the U.S. The early assessments acknowledged that the technology appeared to be interesting, but until it was developed and demonstrated there would be gaps in the analysis. One firm conclusion was that containment and surveillance would have to play a larger role than in established fuel cycles because of the difficulty in confirming the composition of the spent fuel. However, since the fuel cycle facility would contain few portals and no pipes, containment would be a natural advantage. The more recent Department of Energy report was positive in most respects, but expressed concern about general technology transfer (hot cell operations, metal melting and casting, etc.)