2

Merits and Viability of Existing Nuclear Fuel Cycles for U.S. Light Water Reactors

This chapter reviews the deployment and status of existing commercial nuclear reactors and associated fuel cycles with emphasis on the predominant light water reactors (LWRs) and the once-through and monorecycle fuel cycles. The chapter responds to the part of the first charge of the statement of task, which calls for an evaluation of the merits and assessment of the viability of different nuclear fuel cycles, including fuel cycles that may use reprocessing, for existing reactor technology options. The evaluation and assessment of fuel cycles for advanced reactor technology options are described in Chapters 3 and 4.

In this chapter, the committee provides the summary and findings up front (Section 2.1); describes the development of nuclear energy generation for civilian applications with emphasis on the United States, which pioneered the technology (Section 2.2); reviews the global status of the commercially established nuclear fuel cycle facilities (Section 2.3); and summarizes the not-yet-available facilities that would be needed to complete the existing full cycle options (Section 2.4). In Section 2.5, the committee uses a comparison of the situation in France—one of the two countries that currently operate commercial-scale reprocessing facilities—and in the United States—a country without such facilities—to illustrate the differing development of nuclear fuel cycle policies in these countries, which have the largest global share of nuclear power. Finally, in Section 2.6, the committee provides insights about the merits and viability of the fuel cycle options for LWR technologies.

2.1 CHAPTER 2 SUMMARY AND FINDINGS

At one time, the United States was a global leader in the development of LWR nuclear power technologies and in developing and testing some of the first non-LWR technologies. However, the lack of a long-term policy commitment and research and development (R&D) investment to nuclear energy has resulted in a partial decline in U.S. technical leadership and expertise. In contrast, China, France, and Russia, among other countries, have set clear national goals with regard to the role of nuclear energy in an overall national energy-security context and have supported the industry and nuclear R&D community for decades, in order to develop advanced LWR nuclear power technologies. These multidecade investments have led to the development of more technically complex technologies; even then, it is still unclear that these technologies will be available by 2050. Some limited progress is apparent, such as Russia’s BN-600 and BN-800 fast reactors and France’s La Hague reprocessing facility for LWR spent fuel, but there are many more instances where development of alternatives to the LWR-based once-through fuel cycle has been halted or delayed for decades for various reasons, including cost proliferation and concerns.

Monorecycling of plutonium in LWRs in France and other countries was born out of the need to mitigate or minimize the buildup in plutonium inventory in the absence of a demand for fueling fast reactors, while simultaneously maintaining an industrial ability to improve the know-how of reprocessing and recycling technology. The detailed information available on the nuclear program in France provides a benchmark that indicates the level of commitment needed to develop monorecycling. Notably, even with substantial investment, France is still decades away from achieving more advanced fuel cycles, putting into perspective the claims of advanced reactor developers that certain progress could take place on a shorter time frame. In the United States, industrial experience with reprocessing was short-lived and rather unsuccessful, as measured by the experience at the West Valley plant (shut down in 1972) and the Morris facility (determined to be inoperable in 1972). The amount of fuel reprocessed at West Valley over a period of 6 years was less than 1 percent of the amount of spent fuel that had been generated at that time by the U.S. fleet of LWRs. Therefore, the issue of plutonium storage or reuse is not a current issue for the U.S. commercial sector.

The default option for the existing U.S. spent fuel inventory remains the once-through nuclear fuel cycle, which is still not being fully implemented because of the political impasse over the Yucca Mountain geologic repository site in Nevada. No incentives presently exist for undertaking monorecycling in the United States, largely because of the high costs involved and the decreasing contribution by LWRs to the generation of electricity due to plant shutdowns; substantial challenges, based on past experience, for successful licensing and construction of spent fuel reprocessing and mixed oxide fuel fabrication installations; security and environmental concerns; and the abundance of natural uranium and uranium enrichment at relatively low costs for the foreseeable future.

Finding 1: Substantial, sustained investments to 2050 and beyond are required to develop technically complex advanced nuclear technologies and fuel cycle facilities and to enable potential commercial success. Notably, France has had a consistent vision on nuclear energy’s role in its energy security for more than five decades, and as a result, the majority of France’s electricity comes from nuclear power; however, after delaying development of fast reactors, it has yet to close the fuel cycle and is still decades from doing so. No matter what fuel cycle option the United States chooses—whether direct geologic disposal or a closed fuel cycle using advanced technologies—long-term vision and significant and sustained financial commitment will be required to execute it.

Finding 2: Continued use of the once-through fuel cycle for the existing U.S. light water reactor (LWR) fleet has several merits: (1) lower cost compared with any fuel cycle that involves reprocessing and recycling; (2) a reliable international market for nuclear fuel services from multiple suppliers (although that could be disrupted by international crises, such as war)¹; (3) compatibility with the projected available uranium resources; (4) well-understood proliferation resistance of the entire fuel cycle; and (5) theft resistance of spent nuclear fuel. However, the once-through cycle remains incomplete in the United States because there is still no progress toward establishing an operating geologic repository for the spent fuel from nuclear power plants. Pursuing the monorecycling fuel cycle with existing LWRs in the United States would add cost to nuclear power generation but produce no significant benefits, given the projected abundant supply of natural uranium and uranium enrichment at relatively low cost for the foreseeable future.

2.2 DEVELOPMENT OF THE CURRENT GENERATION OF NUCLEAR POWER PLANTS AND SUPPORTING FUEL CYCLES

In the United States, starting in the late 1940s and continuing through the 1950s, the U.S. Atomic Energy Commission (AEC), along with utility and industry partners, spearheaded development and construction of the first U.S. nuclear power reactors. In particular, the world’s first full-scale nuclear power plant devoted exclusively to

___________________

¹ The October 2020 agreement between the U.S. Department of Commerce and Rosatom (the Russian state nuclear energy corporation) allows Russia to continue to export enriched uranium to the United States, but it reduces the proportions from approximately 20 percent of U.S. demand to no higher than 15 percent from 2028 to 2040. The amendment also limits the natural uranium and uranium conversion services from Russia to an amount equivalent to no more than 5 percent of U.S. enrichment demand from 2026 to 2040.

peacetime uses for generating electricity was a demonstration pressurized water reactor,² a type of LWR, located at Shippingport, Pennsylvania, that reached criticality in December 1957 (DOE-NE, 2002). The 1960s and 1970s brought further development and construction of commercial LWRs for electricity generation, taking advantage of the technology and facilities developed for defense purposes, especially the capability, developed during the Manhattan Project, to enrich uranium. Additionally, the U.S. Navy’s Naval Nuclear Power Program led developments in reactor technologies, including the pressurized water reactor that was a model for Shippingport. These nuclear power plants were promoted as relatively inexpensive and emission-free sources of electricity. As a result, nuclear power was seen at the time as becoming the major power source for electricity generation of the future. However, because the natural abundance of uranium ore was perceived as being potentially scarce, the AEC was concerned about the availability of sufficient material for both nuclear weapons and nuclear power programs, and sought an alternative to long-term sole reliance on uranium-fuel–based LWR technology.

As a result of this concern, at the outset of commercial nuclear power development, the AEC sought to develop a type of reactor that would have the potential of generating more fissile material than it consumed, hence the term breeder reactor. A plutonium breeder reactor requires a fast neutron spectrum, so thermal-spectrum LWRs cannot be used.³ The main plutonium isotope, plutonium-239, is created by neutron capture in uranium-238, which is 138 times more abundant than uranium-235. To accumulate the initial plutonium for the first core of a breeder reactor, it is necessary to first recover the plutonium present at a low concentration (less than ~1 weight-percent) in the spent fuel discharged from LWRs. Therefore, from the beginning, a significant amount of R&D and demonstration worldwide has been focused on developing and maturing the technologies associated with (1) breeder reactor design, fueling, and operation, and (2) chemical recovery of the plutonium from irradiated LWR fuel. Both technologies—breeder reactors and reprocessing—turned out to be technically complex and economically challenging. After several decades of development, no country has yet operated an economic and reliable commercial breeder reactor.

In the late 1940s, the AEC authorized construction of the Experimental Breeder Reactor-I (EBR-I) in Idaho to investigate a technology that could breed fissile material (plutonium⁴) from fertile uranium-238. On December 20, 1951, EBR-I generated the first electricity from nuclear energy, although it was intended to be used for experiments and not as an electricity-generating reactor. On November 29, 1955, this reactor experienced a partial fuel meltdown during a coolant flow test. Despite this accident, EBR-I proved the principle of plutonium breeding in a fast reactor. The reactor was repaired and continued to be used for experimental purposes. It was deactivated in 1964.

Next, the Experimental Breeder Reactor-II (EBR-II), a 20-MWe (megawatts electric) unit, supplied electricity intermittently from 1964 to its closing in 1994. “The original emphasis in the design and operation of EBR-II was to demonstrate a complete breeder-reactor power plant with on-site reprocessing of solid metallic fuel” (Westfall, 2004). The original breeder cycle testing was conducted until 1969; subsequently, EBR-II was used for testing concepts for the proposed integral fast reactor.⁵

In 1970, the Clinch River Breeder Reactor project was authorized. The project’s conception was to lead to development of liquid-metal breeder reactor technology for commercial electric-power generation in the United

___________________

² In addition to electricity generation, the Shippingport reactor was used to demonstrate “seed-and-blanket” technologies, in which neutrons from the reactor core “seed” were used to breed fissile material in a “blanket” of fertile material such as uranium-238 and thorium surrounding the core. One demonstrated configuration tested for using the reactor as a thermal breeder, but the reactor did not actually breed. As explained in the main text and next footnote, breeder reactors have almost always used fast neutrons. Breeding is theoretically possible in thermal molten salt reactors operating on the thorium/uranium-233 cycle if protactinium-233 is extracted (Nagy et al., 2008).

³ Fast, or highly energetic, neutrons result, on average, in more neutrons produced per fission than slow, or thermal energy, neutrons. Having excess neutrons provides more neutrons for breeding fissile material from fertile material in breeder reactors. To make sure that neutrons are not slowed down, breeder reactors use metallic coolants, because metals such as lead and lead-bismuth eutectic do not slow down colliding neutrons appreciably, even after a neutron experiences several collisions. In comparison, water used as coolant in LWRs has a significant neutron slowing down, or moderating, effect even after just a few collisions between neutrons and hydrogen atoms in water.

⁴ Plutonium consists of several isotopes, the main ones being Pu-238, Pu-239, Pu-240, Pu-241, Pu-242, and Pu-244. Both Pu-239 and Pu-241 have favorable characteristics for fuel breeding technology when they interact (i.e., fission) with high-energy neutrons. The isotopic makeup of plutonium is a function of the composition of the starting material makeup and irradiation conditions.

⁵ The integral fast reactor would breed more fuel than EBR-II and would be distinguished by a nuclear fuel cycle that performs reprocessing via electrorefining (see Chapter 4) at the reactor site.

States and was intended to be a 350-MWe prototype and demonstration for this type of reactors. Continued escalation in the cost of the project (from an initial cost estimate of ~$400 million to more than $3 billion) and concerns about nuclear weapons proliferation eventually resulted in its termination in 1983 (Breeder Reactor Corporation, 1985).

In addition to the programs in the United States, several test, experimental, or prototype breeder reactors were built in other countries (MIT, 2018), including a commercial-scale reactor, France’s Superphénix, which was connected to the grid in 1986 and permanently shut down in 1998.⁶ Some of these reactors experienced operating difficulties, including some accidents that resulted in shutdowns (IAEA, 2004b). Today, only Russia operates two industrial-size fast reactors: the BN-600 and BN-800. Russia’s experience with operating commercial-size fast reactors is described in Appendix H.

As noted above, recovery of plutonium from reprocessing spent LWR uranium oxide fuel is required to support the development and fueling of breeder reactors. To that end, several fuel reprocessing facilities have been built and operated, most notably in France, Russia, the United Kingdom, and the United States. The West Valley Demonstration Project (New York) was the first, and to date the only, commercial LWR fuel reprocessing plant in the United States.⁷ Starting in 1966 with a potential capacity to reprocess 300 MT (metric tons) of spent LWR fuel annually, the facility eventually reprocessed about 640 MT in 6 years before shutting down in 1972 in response to new regulatory and environmental requirements. Required plant modifications were predicted to increase the reprocessing costs by an order of magnitude, and U.S. utilities chose not to enter into additional contracts at the increased price, given better prospects for natural uranium resources and delays in the development of fast reactor technology. As a result, the owner abandoned the plant, and West Valley became a nuclear waste cleanup site with an estimated remediation cost of $4.5 billion (GAO, 1977; von Hippel, 2007).

In 1974, India tested a nuclear explosive device using plutonium produced with technology obtained from Western supplies; this triggered a major shift away from reprocessing in U.S. nonproliferation policy in 1976.⁸ This new policy was intended to discourage further adoption of technologies applicable for potential weapons production. It also effectively emphasized and encouraged global implementation of the once-through fuel cycle concept. The simplicity, favorable economics, and nonproliferation characteristics of the once-through fuel cycle facilitated its acceptance with the nuclear power industry in the United States and in many other countries.

As the commercial nuclear industry was developing in the United States, the Navy’s Naval Nuclear Power Program selected thermal LWR technology to power its nuclear fleet. This selection and the program that developed were accompanied by a corresponding buildup of a supporting industrial base and associated expertise among the industry and regulators. The commercial nuclear industry, leveraging these economies of scale, also preferentially deployed LWRs in its power plants. The adoption of the once-through fuel cycle and LWR technology by the U.S. nuclear industry has resulted in low-enriched uranium oxide fuels becoming the single de facto standard nuclear fuel, thereby simplifying the fuel cycle. An efficient infrastructure emerged to support the use of a single fuel type, and considerable research has been performed to further improve the economics of the uranium oxide once-through fuel cycle. On the back end, understanding of spent fuel performance in storage and disposal has focused on uranium oxide fuels. The industry has taken advantage of the learning curve with uranium oxide fuels to establish a mature supply chain and well-understood waste behavior to reduce costs and improve performance.

From the late 1970s to today, the relatively low cost of uranium, the abundance of natural uranium, and the deployment of more economical methods for enriching uranium have all contributed to the viability of the once-through fuel cycle for LWRs, not only in the United States but also in most nuclear power–producing countries. Reprocessing of LWR fuel has continued in France and Russia with various degrees of commercial success, as illustrated by the ebb and flow of contractual agreements between reprocessing plant owners and utilities. A sum-

___________________

⁶ The Superphénix plant started its operation 4 months before the Chernobyl accident, and the project was eventually terminated by governmental decision in 1997. Despite its complicated “political life,” Superphénix provided a wealth of experience on construction and operation of an industrial-size prototype.

⁷ Two additional commercial reprocessing facilities were constructed but never operated: General Electric’s Midwest Fuel Recovery Plant at Morris, Illinois, which was completed but declared inoperable in 1970, and Allied General Nuclear Services, which began construction of a 1,500-tons/year facility in Barnwell, South Carolina, but was canceled in 1977.

⁸ President G. Ford in 1976 first began the policy of “indefinite deferral of reprocessing.”

mary of reprocessing activities in relevant countries is provided in Appendix H. For more information about the history of U.S. nuclear reactor and fuel cycle development, see Vine (2011).

2.3 COMMERCIAL NUCLEAR FUEL CYCLE OPERATIONS SUPPORTING LIGHT WATER REACTORS

The nuclear fuel cycle consists of the front end, which involves the preparation of the fuel; the service period, in which the fuel is used during reactor operations; and the back end, during which the spent fuel and nuclear waste is managed, stored, and disposed.

Figure 2.1 shows the once-through fuel cycle operations that are currently supporting the operation of most LWRs worldwide, as well as the steps for the monorecycle option being implemented in a few countries.⁹ Transport of fuel or materials created by nuclear fission is shown by arrows between different activities or facilities. Transportation, an important element of any fuel cycle activity, is covered in greater detail in Chapter 5. Not shown in Figure 2.1 is the anticipated final disposal of spent fuel or high-level wastes in a geologic repository, because

___________________

⁹Figure 2.1 is an oversimplification of the once-through fuel cycles that have been implemented in national programs. See Krahn et al. (2014) for a more recent assessment of the contemporary nuclear fuel cycle in the United States.

there are presently no such operating facilities for commercial spent nuclear fuel anywhere in the world, although repositories in Sweden and Finland may begin operations later this decade.

2.3.1 Front-End Operations: A Global Enterprise

The front-end technologies and facilities required to sustain LWRs include uranium mining and milling, conversion and enrichment, deconversion, and fuel fabrication. These technologies are well established and comprise a global network of nuclear fuel services. However, this global nature of the nuclear fuel supply chain can make the United States vulnerable to foreign disruptions, including war, as evidenced by the 2022 Russian invasion of Ukraine.

Uranium mining and milling: Since the 1970s, prospecting has discovered that uranium is significantly more abundant than previously assessed (NEA-OECD, 2006a). While 20 countries have uranium mines, in recent years, more than 50 percent of production of uranium comes from nine mines in four countries: Kazakhstan, Canada, Namibia, and Australia (WNA, 2021a). Using only the open, once-through fuel cycle, the world has sufficient uranium resources to meet the demands of nuclear power for the foreseeable future. In particular, according to the Nuclear Energy Agency’s assessment, as of the end of 2019, the global recoverable uranium resources, based on a price of $130 per kilogram uranium, are sufficient for over 135 years, considering global requirements (NEA and IAEA, 2020).

The United States, once having a substantial domestic uranium mining industry, currently contributes only about 1 percent of the global uranium market because of the relatively low grade of uranium ores in most U.S. mines. Increasingly, since the late 1980s, U.S. commercial nuclear power production has relied on foreign-origin uranium for reactor fuel. In 2020, for example, the U.S. nuclear power sector purchased about 70 percent of its uranium from just four countries: Canada (22 percent), Kazakhstan (22 percent), Russia (16 percent), and Australia (11 percent) (EIA, 2021a).

Uranium conversion and enrichment: Chapter 1 introduced the fundamentals of conversion and enrichment. Here, the focus is on the status of global capacities for conversion and enrichment.

The four primary uranium conversion companies are Russia’s Rosatom at about 38 percent, China’s National Nuclear Corporation and Canada’s Cameco at about 25 percent each, and France’s Orano at about 8 percent (WNA, 2022b). As mentioned in Chapter 1, there is currently no operating conversion facility in the United States, although the Honeywell Metropolis Works facility plans to restart operations in 2023 (WNN, 2021d).

While 13 countries have some enrichment capabilities, approximately 90 percent of the enriched uranium is produced in seven countries by four companies: Tenex (Russia), Urenco (Germany, the Netherlands, the United Kingdom, and the United States), Orano (France), and China National Nuclear Corporation (WNA, 2020a). The only operating enrichment plant in the United States is owned by Urenco. Located in Eunice, New Mexico, this plant supplies almost one-third of U.S. commercial nuclear power industry requirements for uranium-235 enriched to less than 5 percent (WNA, 2021b).

Deconversion and fuel fabrication: “The enriched uranium is transported to a fuel fabrication plant where it is converted to uranium dioxide powder” (WNA, 2021c). This powder is then pressed and sintered into fuel pellets, which “are subsequently inserted into thin tubes known as fuel rods, which are then grouped together to form fuel assemblies” (WNA, 2021c). The number of fuel rods and exact dimensions of a fuel assembly vary and depend on the specific design of the reactor lattice (WNA, 2021c).

Domestic fuel fabrication is the only step of the fuel supply chain that is currently sufficient to meet the needs of the U.S. commercial power industry for uranium oxide fuel with less than 5 percent enrichment. Currently, three fuel fabrication plants are licensed by the U.S. Nuclear Regulatory Commission to produce low-enriched uranium fuel that is sold worldwide: Global Nuclear Fuel-Americas in Wilmington, North Carolina; Westinghouse Columbia Fuel Fabrication Facility in Columbia, South Carolina; and Framatome, Inc., in Richland, Washington.

2.3.2 Back-End Operations: Managed Storage

Assemblies loaded in an LWR typically generate energy for 4–6 years before the fuel is no longer suitable for power production. The assemblies are then discharged from the reactor and placed in an underwater storage and cooling facility, commonly referred to as a “spent fuel pool,” located at the reactor site.

During the 1960s and 1970s, when the nuclear power plants were first deployed at large commercial scale, it was anticipated that spent fuel would be removed from the spent fuel pool after a few years of cooling and shipped to a reprocessing plant in order to extract the plutonium to fuel an expanding fleet of fast reactors. While the prospects for fast reactors fueled with plutonium dimmed progressively in the 1970s and 1980s, countries that developed reprocessing programs (with government support) with the intent of separating plutonium for fast reactors began to market the technology for management of spent fuel and monorecycling in LWRs instead. Commercial reprocessing services for LWR fuel became available in the United States (West Valley) in 1966 and in France (La Hague) in 1976, although as discussed above, the West Valley plant was shut down in the 1970s. In 1979, the Thermal Oxide Reprocessing Plant (THORP) for uranium oxide fuel began construction at Sellafield in the United Kingdom and started operations in 1994.¹⁰

Several Western European (France, Belgium, Germany, Italy, the Netherlands, Spain, and Switzerland) and Japanese utilities took advantage of the commercial reprocessing services available in France and the United Kingdom for an extended period of time. However, many of these utilities no longer use commercial reprocessing services, for a variety of reasons, including nuclear power phase-out policies and economic penalties deriving from the higher expense of reprocessing relative to spent fuel storage, the higher fabrication cost of plutonium-based mixed oxide fuels compared with uranium fuels, and the lack of prospects for plutonium reuse in fast reactors. Presently, France and Russia continue to reprocess, and Japan intends to do so in a facility located at Rokkasho. Between La Hague (France)¹¹ and Sellafield (United Kingdom),¹² ~50,000 MT of spent LWR fuel have been reprocessed.

Today, most of the approximately 30 countries with nuclear power programs neither reprocess spent nuclear fuel nor use mixed oxide fuels. Only France and Russia currently operate commercial-scale reprocessing facilities. China has one operating small-scale facility for reprocessing civilian nuclear fuel and one under construction. The countries that currently reprocess or have explored this technology in the past represent a large fraction of global nuclear energy generation. About 10 percent of the world’s reactors are licensed to use mixed oxide fuel, but mixed oxide makes up only about 5 percent of the world’s new nuclear fuel (WNA, 2017a). See Appendix H for a summary of reprocessing and recycling programs in China, India, Japan, Russia, and the United Kingdom, and Section 2.5 for a comparison of reprocessing programs and policies in France and the United States.

2.3.2.1 Management of Spent Fuel in the Once-Through Cycle

Spent fuel generated by a once-through fuel cycle is managed in interim storage until a final disposal site becomes available. In the United States, continued delays have required utilities to assume the responsibility for storing spent fuel for much longer than anticipated, including beyond the lifetime of the plants. Initially, utilities typically opted for reracking the spent fuel pools to accommodate larger numbers of assemblies. By the mid-1980s, it became apparent that space limitations would eventually require a significant fraction of the spent fuel inventory to be transferred out of spent fuel pools into dedicated interim storage systems.

Some countries built at-reactor systems for dry storage under inert conditions, while others built consolidated interim storage facilities, both wet (Sweden) and dry (Germany).¹³ A variety of both cask- and canister-based dry technologies have been employed.

___________________

¹⁰ The United Kingdom’s earlier reprocessing facility was for Magnox fuels, which is not relevant to this section.

¹¹ Orano La Hague reprocesses ~1,100 tons of spent fuel per year from EDF, the world’s leading nuclear operator, totaling ~40,000 tons of spent fuel reprocessed to date (Orano, n.d.).

¹² Reprocessing operations ended in November 2018 at THORP after 24 years of operation because of “a significant downturn in demand. The plant reprocessed 9,331 tons of spent fuel” (WNN, 2018).

¹³ In the United States, two consolidated interim storage facilities located in Texas (Interim Storage Partners) and New Mexico (Holtec) are under active development; the Interim Storage Partners’ facility received a 40-year license in September 2021, while the Holtec license application is presently under review.

See Chapter 5 for a discussion about the laws and regulations on spent fuel storage and the impact on disposal in a permanent repository.

2.3.2.2 Management of Fissile Products and Waste Streams Resulting from Spent LWR Fuel Reprocessing

When reprocessing is implemented, spent LWR fuel is first transported from the spent fuel pools located at the reactor sites to the pool(s) located at the reprocessing facility. In France, for example, spent uranium oxide fuel is shipped to the La Hague facility within 2 years following discharge from the reactor and is stored in the spent fuel pools located at La Hague for a minimum of about 5 years prior to reprocessing.

The main product streams from reprocessing operations typically consist of reprocessed uranium, reactor-grade plutonium, vitrified high-level waste¹⁴ (primarily fission products, any residual plutonium, and minor actinides); fuel assembly hardware waste (cladding, grids, thimble tubes, nozzles, etc.); and low-level waste, including effluents. A main drawback of this approach is an accumulation of reactor-grade plutonium stockpiles (see Table 2.1). Storage of plutonium is expensive because of physical protection requirements. In addition, reuse of the plutonium after extended storage would likely require chemical separation of the americium-241 (half-life of 432 years) that builds up in the stored plutonium as the result of the decay of plutonium-241 (half-life of 14 years). Loss of fissile plutonium-241 and buildup of americium-241 degrade the reactivity of the fuel in thermal and fast reactors, and increase worker dose during mixed oxide fuel fabrication.¹⁵

To mitigate the buildup of the plutonium inventory, reactor-grade plutonium can be substituted for enriched uranium, and fuel rods containing a mixture of depleted or natural uranium oxide and plutonium oxide (i.e., mixed oxide fuel) can be fabricated and used in LWRs. Reprocessed uranium and separated plutonium have been recycled in several European and Japanese LWRs. Recycling of both plutonium and reprocessed uranium can result in natural uranium savings of up to ~20 percent. Recycled plutonium alone helps reduce up to 12 percent of the amount of natural uranium and enrichment work required for producing new fuel (NEA-OECD, 2021). This approach, which reduces the amount of natural uranium and enrichment required for producing new fuel, decreases the economic penalty associated with reprocessing. However, even if the cost of reprocessing is not included, “MOX [mixed oxide] fuel is more expensive than uranium fuel” (Bunn, 2021).

Plutonium inventories, as shown in Table 2.1, may result from technical and institutional factors that prevent keeping plutonium separation and reuse in LWRs in balance, or more generally from the fact that the market for plutonium for fueling fast reactors has not materialized.

Reprocessing: Commercial reprocessing of spent LWR fuel relies on the plutonium and uranium extraction (PUREX) process, an aqueous process for treating spent LWR fuel that was developed at Oak Ridge National Laboratory (Long, 1967). PUREX has more than 50 years of operational experience at commercial scale, during which time significant improvements in plant operations and safety have been incorporated (Poinssot, 2021). Presently, La Hague, where about three-quarters of all LWR fuel reprocessed worldwide has been treated to date, is the largest PUREX-based reprocessing facility for LWR fuel. For more information about the PUREX process and the French reprocessing program, see Chapter 4.

Reprocessed uranium recycle and storage: Recycling of reprocessed uranium prior to fuel fabrication has previously been implemented in several countries, notably in Belgium, France, Germany, Japan, the Netherlands, Sweden, and Switzerland (IAEA, 2007a, 2009). Presently, reprocessed uranium is not intended to be recycled, except in France, primarily for economic reasons, but that is subject to change if natural uranium costs increase

___________________

¹⁴ According to a U.S. Nuclear Waste Technical Review Board fact sheet, “HLW [high-level (radioactive) waste] is vitrified by mixing it with a combination of silica sand and other glass-forming chemicals, heating the mixture to very high temperatures [approximately 1,150°C (2,100°F)] until it melts, and pouring the molten material into stainless steel canisters where it cools to form a glass” (U.S. Nuclear Waste Technical Review Board, 2017).

¹⁵ Safety issues can also arise from storage of plutonium over extended periods; as an example, helium accumulation in the United Kingdom’s packages of plutonium dioxide produced at THORP could lead to higher internal pressures and embrittlement of plutonium dioxide ceramics, such as mixed oxide fuel (Hyatt, 2020).

TABLE 2.1 National Holdings of Civil Separated Plutonium as of 2021

^a Values as of December 31, 2016 (from China’s most recent information circular as of March 2022).

^b Value as of March 2014 (IPFM, 2014). Swedish separated Pu is held in the United Kingdom.

SOURCES: IAEA (2021c); IPFM (2021)

(EPRI, 2010a). Reprocessed uranium contains two additional uranium isotopes (uranium-232 and -236) that complicate reuse in enrichment and reactors.¹⁶ Reprocessed uranium (U_rep) in LWRs can be recycled “via direct enrichment using centrifuge technology. In this case, U_rep experiences essentially the same process as natural uranium feedstock, albeit with some purification as well. U_rep is converted into UF₆ and then sent as soon as possible to the enrichment plant to avoid the buildup of ²³²U daughters” (EPRI, 2010a). Reenrichment of reprocessed uranium must be performed in a dedicated cascade to avoid cross contamination by uranium-232.¹⁷ In the absence of a dedicated cascade, reenrichment of reprocessed uranium can be obtained by downblending, using some of the existing stockpiles of excess defense-related enriched uranium; this approach avoids the buildup of uranium-236 and its associated neutronic penalty (EPRI, 2010a).

Because reprocessed uranium may require overenrichment in uranium-235 to compensate for the neutron-absorbing uranium-236, and because natural uranium resources are available at reasonably low cost, fuel economics favor the use of fresh rather than recycled reprocessed uranium in existing LWRs. As a result, some countries have chosen not to recycle reprocessed uranium and are storing it until its use becomes economically competitive.¹⁸ Countries that have exercised the fuel cycle option to reenrich and recycle reprocessed uranium in an LWR have successfully managed the radiological aspects of uranium-232 associated with handling reprocessed uranium during conversion, reenrichment, fuel fabrication for recycle, transportation, and fuel loading during reactor operations. Spent reprocessed uranium oxide is not presently scheduled for further reprocessing¹⁹ and will require storage until a decision is made to either reprocess or dispose of it in a geologic repository, when one becomes available. If a decision is made to not recycle reprocessed uranium and its use is no longer considered economically advantageous, the stored inventory of reprocessed uranium could be disposed of as low-level or Greater-than-Class-C (GTCC) waste (IAEA, 2007a). (See Appendix D for waste classifications.)

___________________

¹⁶ Buildup of uranium-232 decay products over time causes both contamination and radiological hazards, resulting from the intense beta and gamma radiation of its short-lived daughter products (lead-212, bismuth-212, and especially thallium-208). Uranium-236 is a parasitic neutron absorber that requires an increase in enrichment of uranium-235 of ~0.5 percent.

¹⁷ Such a facility is available in Russia and planned in France.

¹⁸ Most of the reprocessed uranium remains in storage, though its conversion and reenrichment (in the United Kingdom, Russia, and the Netherlands) has been demonstrated, along with its reuse in fresh fuel. In Belgium, France, Germany, and Switzerland, over 8,000 MT of reprocessed uranium have been recycled into nuclear power plants (WNA, 2020b).

¹⁹ Compared with spent uranium oxide, spent reprocessed uranium oxide contains ~4 times more uranium-232 and ~3 times more uranium-236, and accordingly presents a greater radiological hazard and has a more consequential neutron absorber-to-fissile content ratio.

Plutonium recycle and storage: Plutonium can be recycled in an LWR by blending it with natural or depleted uranium to form a mixed uranium-plutonium oxide fuel, or LWR-MOX. Fabrication of MOX fuel was pioneered for breeder reactor fuel and adapted for LWR fuel as an alternative to low-enriched uranium oxide. To use MOX, existing LWRs need to make some modifications, such as adding more control rods and amending their safety analyses and licenses (EPRI, 2009a). The core loading of LWR-MOX (with typically 8–10 percent plutonium) is limited to less than 50 percent, though most LWRs operate with a core loading of around one-third.²⁰ Notably, utilities in the United States are not using MOX fuel, and during the recent development of a potential MOX facility at the Savannah River Site, use of this fuel by utilities was conditional on the U.S. government providing financial support.

Irradiation of MOX fuel to the burnup targets of interest in LWRs typically limits the fuel to monorecycling because LWR-MOX fuel is significantly different from conventional uranium oxide fuel from a neutronic point of view. The fissile quality of the plutonium (i.e., the relative fraction of fissile plutonium [plutonium-239 and -241] to total plutonium) is lower in the plutonium recovered from spent LWR-MOX. Box 2.1 illustrates the impact of monorecycling on the buildup of plutonium, americium, and curium isotopes in the spent fuel. It would then be necessary to increase the plutonium content for a second recycling in order to compensate for the decrease in fissile quality. Deterioration of the safety parameters²¹ beyond plutonium concentrations of ~12 percent becomes a safety barrier to multirecycling of plutonium in present-day LWRs.²² As a result, enriched uranium has to be added to the fuel for such a potential option.

From the perspective of spent fuel assembly volume, each spent MOX assembly is obtained by reprocessing seven spent uranium oxide assemblies, thereby significantly reducing the number of spent fuel assemblies to be stored. Therefore, it is also a more compact approach for interim storage of the Pu itself. However, a fair comparison of the impacts of waste storage requires accounting for all the waste streams generated, including high-level waste and reprocessed uranium. In addition, because of waste generated during reprocessing and MOX fuel fabrication, low-level waste volumes, including GTCC wastes (similar to intermediate-level waste), increase compared with uranium oxide storage.

Moreover, in addition to the handling, storage, and disposal technologies developed for spent uranium oxide (which are applicable to spent MOX), spent MOX fuel management must take into account decay heat, potential criticality safety, and radiation source terms. These additional considerations are required because the decay heat generation of spent LWR-MOX decreases more slowly than that of spent uranium oxide (see Figure 2.2). Therefore, wet storage of spent LWR-MOX assemblies is preferred over dry storage. Additionally, the fissile inventory and minor actinide content are significantly larger for spent LWR-MOX than for spent uranium oxide.

As discussed above, monorecycling of the plutonium recovered from spent LWR fuel by reprocessing has been implemented to avoid the economic penalties associated with storage of the extracted plutonium. Otherwise, there is little technical rationale for LWR spent fuel reprocessing and monorecycling, given the modest uranium resource savings and larger economic penalties compared with direct geologic disposal. However, some countries justify LWR spent fuel reprocessing as a means of separating plutonium for use in future fast breeder reactors, as well as establishing and maintaining the technological know-how for developing and operating reprocessing facilities. In any case, monorecycling would appear to be an interim strategy at best. Alternatively, countries that plan to pursue fast breeder reactors could defer LWR spent fuel reprocessing until there is a realistic prospect for use of the resulting separated plutonium to avoid the cost and security liabilities of long-term plutonium storage.

High- and intermediate-level waste storage and low-level waste disposal: Vitrified fission products/minor actinides (i.e., high-level waste [HLW]) and compacted fuel/process hardware (i.e., intermediate-level waste [ILW]²³) are stored in containers in a dry environment. While estimates vary, Bunn et al. (2003) cites British Nuclear Fuels Limited (BNFL)

___________________

²⁰ Some LWRs designed for 100 percent MOX are under construction (EPRs, Ōma Nuclear Power Station in Japan), and others are capable of operating with 100 percent MOX (Palo Verde Station in the United States).

²¹ These safety parameters are control rod reactivity worth, boron reactivity worth, moderator void, and temperature coefficients.

²²Present-day LWRs refers to LWRs with conventional lattices. For example, a conventional 17×17 pressurized water reactor (PWR) lattice, such as used in many of today’s PWRs, is characterized by a moderator (water)-to-fuel ratio of 2:1. Multirecycling of plutonium will typically require a higher moderator (water)-to-fuel ratio, and therefore will require different fuel assembly geometries and reactor internals.

²³ The United States does not have an ILW category, but it is similar to Greater than Class C waste. ILW is used here to provide an international context to the estimate of LLW generated by nuclear power plants.

and Compagnie générale des matières nucléaires (COGEMA) estimates, which indicate approximately 0.12 m³/tHM (cubic meters per ton of heavy metal) of HLW. For ILW, BNFL’s estimate is 0.8 m³/tHM and COGEMA’s is 0.35 m³/tHM. Bunn et al. (2003) also state an estimate for low-level waste (LLW) of 2.8 m³/tHM, which is substantially greater in volume than HLW and ILW combined. LLW can be disposed of in near-surface disposal sites, but doing so would require additional costs. As to the costs, when BNFL was reprocessing, it had permission from the UK government to negotiate substitution agreements with customers, in which a customer would agree to take back a larger amount of HLW in exchange for the reprocessing company handling disposal of the LLW. Shipping LLW back to the customer could be problematic given its large volumes. Bunn et al. (2003) estimate that LLW handling and disposal could add several tens of dollars per kilogram of spent fuel, which is on the order of 1 percent of the cost of reprocessing (Bunn, 2021).

Atmospheric and gaseous effluents: The release and dilution of some gaseous and liquid radionuclide (especially helium-3, krypton-85, and iodine-129) effluents have raised concerns about their impacts on human health and the environment. The French industry is working on tailored ceramics to confine specific radionuclides, such as iodine (CEA, 2009). According to a European Commission study in 2000, the radioactive discharges from La Hague and Sellafield were significantly greater than those from other nuclear power plants in Europe, but significantly smaller than those from the fertilizer (use of phosphogypsum²⁴ during the production of phosphoric acid) and petroleum²⁵ (pumping of oil and gas from the continental shelf in the North Sea) industries (Betti et al., 2004).

___________________

²⁴ Radionuclides contained in phosphogypsum are mainly uranium, thorium, and radon, and their decay products.

²⁵ Radionuclides contained in petroleum are radon and its decay products.

2.3.3 Managed Storage

Independently of the option chosen for the LWR fuel cycle (once-through or monorecycling), the back-end endpoint of all commercial activities is, at the present time, storage. Spent fuel is stored as spent uranium oxide fuel, spent mixed oxide fuel, or spent reprocessed uranium oxide fuel, in either wet or dry environments, and either onsite with the reactor or at away-from-reactor sites. Interim storage of these spent fuel assemblies will be required until either a geologic repository is available for direct disposal or a decision is made to reprocess them. However, in the latter case, even if the separated fissile materials are used as feed material for a more advanced fuel cycle involving multirecycling, there will be spent fuel with low-residual-fissile content, which will not be worth reprocessing and will therefore require long-term management and disposal.

Product and waste streams from reprocessing, including plutonium, reprocessed uranium, vitrified high-level wastes, and low-level wastes (including GTCC waste), are in storage either at the reprocessing plant or in utilities’ facilities. Product materials, which are recovered with the intention of recycling in LWRs or in more advanced fuel cycles, may also require long-term management if they cannot be readily utilized as fuel.

Given that presently neither geologic repositories for spent fuel nor commercial-scale advanced reactors are operating around the world,²⁶ back-end operations and implementation decisions have largely been driven by the need to ensure safe storage of the material containing fissile elements, minor actinides, and fission products. For this reason, back-end operations have been referred to as “managed storage.”

2.4 COMPLETING THE LWR FUEL CYCLE

Interim storage, the final storage step in the management of spent fuel prior to final disposal, is presently the de facto endpoint for all nuclear power operations worldwide, but it does not presuppose or preclude any other endpoint for the spent fuel. Figure 2.3 depicts the back end of a fuel cycle, where it is assumed that interim storage is no longer the ultimate activity. Two additional options, referred to as “geologic repository” and “fast reactors” with recycling of plutonium and minor actinides, are shown in the figure.

2.4.1 Geologic Repository

As further discussed in Chapter 5, all fuel cycle options require a geologic repository. There is broad agreement in the scientific community that deep geologic disposal, if well-designed and executed, constitutes a safe option

___________________

²⁶ The two current exceptions are the Russian BN-600 and -800 reactors in Beloyarsk, and the Chinese HTR-PM reactor at Shidao Bay. Starting in 2020, the BN-800 reactor is being fueled with reprocessed uranium-plutonium fuel; the other two reactors (BN-600 and HTR-PM) use uranium-235; see Appendix H for more details.

for most forms of spent fuel and HLW generated by nuclear power plants and the nuclear fuel cycle (NEA-OECD, 2020). “The safety case for an HLW repository requires extensive R&D (regarding site suitability and waste packaging, for example) because the final selection of a site and disposal steps” (EPRI, 2010b) (emplacement, monitoring, retrievability, etc.) are expected to be challenged by both technical and societal issues. Although there is no geologic repository for commercial HLW in operation at this time,²⁷ Finland, Sweden, Canada, and France have made substantial progress toward licensing and could be operating repositories before the end of this decade. In Finland, operations at the Onkalo repository facility are expected to begin in 2024 or 2025, according to Posiva, the company behind Onkalo (The Economist, 2022). Sweden is just a few years behind, with its own repository at Forsmark. In France and Canada, repository facilities could begin operation in ~2035 (Bure, France)²⁸ and in 2040–2045 (Canada, site to be selected by 2024) (Dalton, 2022; WNN, 2022a).

2.4.2 Fast Reactor Technology

Fast reactors can potentially enable nearly full recovery of the energy contained in natural uranium resources by converting the fertile uranium-238 that makes up more than 99 percent of natural uranium into fissile plutonium-239. Plutonium recovered by reprocessing spent LWR uranium oxide fuel can be used to provide the initial

___________________

²⁷ In the United States, the Waste Isolation Pilot Plant is an operational geologic repository for defense-generated transuranic wastes.

²⁸ Estimate as of 2021.

plutonium inventory for fast reactor fueling, because, as discussed in Chapter 4, fissions induced in the fissile plutonium isotopes (-239 and -241) create neutrons in sufficiently large numbers to both maintain power production and produce more fissile material (plutonium-239) than is consumed. In an ideal breeder reactor system, no additional fissile material other than uranium-238 is needed from external sources. However, realizing such a system in practice will be challenging and will require multiple cycles of reprocessing of spent fast reactor fuel and fuel refabrication, as illustrated in Figure 2.3 (see Chapter 4 for additional details).

Historically, the most common design has been the sodium-cooled fast reactor. The largest operating fast reactor is presently the Russian BN-800 (2100 MWth [megawatts thermal]), which began operating in 2015 with a hybrid core consisting of both highly enriched uranium and mixed oxide fuel. See Chapters 3–5 for more details on sodium-cooled fast reactor technology.

Reprocessing of fast reactor mixed oxide fuel: Fast reactor spent fuels will have higher fissile concentrations than LWR fuels and potentially higher burnups. Technologies for reprocessing LWR spent fuels, such as PUREX, can generally be used for reprocessing fast reactor spent fuels, but some modifications are required to address criticality concerns and higher radioactivity levels.²⁹ In addition, pyrochemical methods are being developed in a few countries as integral parts of the refueling/waste management system for fast reactors, especially for nonoxide fuels. These methods would permit the treatment of different types of radioactive fuels with high plutonium content. In the longer term, fuel cycle applications specifically related to advanced reactor concepts may favor the use of pyrochemical processes (IAEA, 2008). These technologies are covered in Chapter 4.

Fast reactor fuel fabrication: Fuel fabrication will require more extensively shielded facilities compared with the glove box facilities in use for fabricating today’s LWR and LWR-MOX fuels. Fuel fabrication is covered in greater detail in Chapter 4.

2.5 NUCLEAR FUEL CYCLE POLICIES: A COMPARISON BETWEEN THE UNITED STATES AND FRANCE

Given the history of the development and deployment of LWR technology and its fuel cycles, it is worth comparing how different energy markets and policy contexts result in different nuclear strategies. For this purpose, the United States and France are compared; while they have the two largest installed nuclear power capacities in the world, they have chosen different approaches to the nuclear fuel cycle.

2.5.1 The United States

The United States operates the largest nuclear fleet of power reactors in the world. At the end of 2021, 93 reactors—of a dozen different models of the pressurized and boiling water reactor types—were operating with a combined generation capacity of about 95,464 MWe (EIA, 2022). Reactor power uprates and high-capacity usage rates have contributed to nuclear power plants maintaining a consistent portion of about 20 percent of total annual U.S. electricity generation from 1990 through 2019.

___________________

²⁹ During reprocessing, the driving safety concerns are keeping the aqueous and organic solutions containing the radioactive, heat-generating nuclides at low-enough temperatures, and especially maintaining subcriticality of the aqueous and organic solutions containing fissile materials. Maintaining subcriticality involves some creativity with component geometry (e.g., annular “cylinders” instead of regular cylinders to maximize neutron leakage, and subsequent capture of the neutrons by neutron absorbers located on the inside and outside peripheries of the annular cylinder). A plant optimized to handle spent LWR fuel with a “low” fissile content could be reconfigured to reprocess other fuels. For example, La Hague has reprocessed spent LWR-MOX and fast reactor fuel, which contain high levels of fissile nuclides, using “dilution,” in which a spent LWR-MOX assembly is reprocessed in a string of spent uranium oxide (UOX) assemblies with a UOX-to-MOX ratio on the order of 10 to allow the plant to stay within its safety boundaries. This approach relies on having an abundance of spent UOX fuel to reprocess with the other fuel. A facility dedicated to the reprocessing of spent MOX (LWR-MOX, enriched uranium MOX, fast reactor MOX) would require substantially greater redesign of equipment (pulse columns, centrifugal contactors, piping and container geometry, use of neutron absorbers, etc.) compared with a facility for reprocessing only spent UOX fuel.

U.S. policies on reprocessing and recycling technologies: As mentioned in Section 2.2, abandonment of the commercial reprocessing of spent nuclear fuel in the United States in the mid-1970s, due to proliferation and economic concerns, set the path of U.S. policy and its nuclear industry into a once-through fuel cycle. While the government ban on reprocessing was lifted in the early 1980s, freeing utilities to pursue reprocessing as long as they paid for it, U.S. commercial reprocessing was not economically competitive and never resumed. In 1993, to try to dissuade additional countries from separating plutonium from spent fuel, the Clinton administration reaffirmed the 1970s policy by stating that “the United States does not encourage the civil use of plutonium and does not itself engage in plutonium reprocessing” (White House, 1993) but committed to meeting its nuclear energy cooperative agreements with Japan and countries in Western Europe that reprocess or use reprocessed plutonium (Andrews, 2008).

In the early 2000s, however, the George W. Bush administration undertook another policy shift and invested in R&D for advanced reprocessing methods that would not completely separate plutonium from other elements, with the view that this might be more resistant to proliferation. The U.S. Department of Energy (DOE) provided research funding for a new program called the Advanced Fuel Cycle Initiative (AFCI) “to develop fuel cycle technologies for Generation IV reactors including reprocessing and using fast neutron reactors to transmute long-lived components of wastes” (WNA, 2021b). The Energy Policy Act of 2005 (Public Law 109-58 § 953) codified an objective of AFCI to “evaluate proliferation-resistant fuel recycling and transmutation technologies that minimize environmental and public health and safety impacts as an alternative to aqueous reprocessing technologies.”

Moreover, in 2006, the George W. Bush administration launched the Global Nuclear Energy Partnership (GNEP), which proposed that the United States work with other developed nuclear power states that have existing fuel cycle capabilities to “develop proliferation-resistant recycling technologies and provide nuclear fuel to developing countries that promised not to engage in enrichment and reprocessing activities” (WNA, 2021b). When the Obama administration began in 2009, it canceled these aspects of GNEP “because it [was] no longer pursuing domestic commercial reprocessing, which was the primary focus of the prior Administration’s domestic GNEP program” (DOE, 2009a).

In 2008, in response to DOE’s GNEP program and industry interest in potential commercial-scale reprocessing and recycling, the U.S. Nuclear Regulatory Commission (U.S. NRC) started down the path of a possible rulemaking for updating the licensing basis for such facilities (U.S. NRC, 2011, 2013b). From 2013 to 2016, the U.S. NRC staff focused its analysis “on assessing the quantitative risk associated with reprocessing facility accidents” (U.S. NRC, 2021a). In 2016, the U.S. NRC staff observed that “industry interest in building and operating a commercial spent fuel reprocessing facility had declined” and consequently suspended its work on this activity (U.S. NRC, 2021a). In March 2020, the U.S. NRC staff sought public comment on whether it should resume the rulemaking or terminate it.

In May 2020, the American Nuclear Society (ANS) and the Nuclear Energy Institute responded by requesting that the U.S. NRC renew its rulemaking activity because of the renewed interest among several vendors in advanced reactor technologies that may benefit from reprocessing and recycling to achieve their full potential. Also, the ANS stated that “the lack of an efficient, technically robust, and technology inclusive regulatory foundation for reprocessing and recycling is a barrier to innovation in advanced reactor designs” (U.S. NRC, 2021b). In July 2021, the U.S. NRC filed notice in the Federal Register that it was formally suspending its efforts on potential rulemaking in this area. The U.S. NRC concluded that “in addition to using fresh fuel obtained from enrichment and fabrication, some advanced reactor designs have the capability to eventually source their fuel from the spent fuel of other reactors, but there was limited interest in pursuing reprocessing activities in the near future (within 10 to 20 years),” and that it was not worth the cost to the U.S. NRC to continue rulemaking in light of this limited interest (U.S. NRC, 2021b).

One outcome of the decisions since the 1970s by the U.S. government and utilities to not pursue commercial reprocessing is that, unlike France and several other countries, the United States did not generate a large stockpile of commercial plutonium, with its attendant security and safety risks. Given that the United States has experienced significant challenges over the past 30 years in seeking to dispose of its stockpile of roughly 50 MT of surplus military plutonium (see below), the addition of a comparably sized commercial stockpile would have greatly increased the disposal burden.

MOX fuel fabrication facility experience: The United States attempted to develop a capability for producing MOX fuel with separated military plutonium that the United States determined to be excess to defense needs. The MOX Fuel Fabrication Facility (MFFF), originally estimated in 2002 to cost $1 billion, was to be the central element of the National Nuclear Security Administration’s (NNSA’s) Plutonium Disposition Program and was initiated to “dispose” of at least 34 MT of excess weapons plutonium by using it in MOX fuel for commercial nuclear power reactors (Holt and Nikitin, 2017). The effort began in 2001 with Duke, COGEMA, Stone & Webster submitting a request to the U.S. NRC to construct the MFFF at the Savannah River site in South Carolina. The U.S. NRC granted a construction authorization in March 2005 (effective for 10 years) with ground breaking commencing approximately 2 years later. In 2014, the U.S. NRC was requested by the applicant (now CB&I Areva MOX Services, LLC) to extend the construction authorization. The request was granted, and the construction authorization was extended from the original March 2015 expiration date to March 2025 (U.S. NRC, 2020g). In October 2018, DOE issued a notice of termination of the construction contract for MFFF to CB&I Areva MOX Services, noting that the MFFF was “significantly behind schedule,” and “would cost significantly more than previously projected” (Perry, 2019). NNSA is presently pursuing a dilute-and-dispose approach to dispositioning surplus weapons plutonium that would not use the plutonium as commercial reactor fuel (NASEM, 2020).

Economically competitive electricity markets and their effect on nuclear power in the United States: In the early 2000s, the Massachusetts Institute of Technology (MIT) published two major reports examining potential changes for the future of nuclear energy in the United States. The 2003 MIT report The Future of Nuclear Power observed that “if in the future carbon dioxide emissions carry a significant ‘price,’ nuclear energy could be an important—indeed vital—option for generating electricity,” and advised that “the nuclear option should be retained, precisely because it is an important carbon free source of power that can potentially make a significant contribution to future electricity supply” (MIT, 2003). The update of this MIT study, published in May 2009, concluded:

The past decade has provided confirmation of MIT’s “sober warning.” From 2013 through April 2021, 12 U.S. reactors were permanently shut down, and through the mid-2020s, seven more are slated for closure (Holt, 2021b). Increasing awareness of, and concerns about, climate change among the U.S. public and political leaders could have benefited social and political acceptance of nuclear power as a low-carbon alternative for baseload generation of electricity. However, the economic competitiveness of nuclear power in the form of LWRs has continued to degrade. Low-cost natural gas, production-credit incentives to renewables, and the lack of any carbon fee have limited the economic and political incentives for pursuing construction of new nuclear plants with large capital costs and potentially large financial risks. In addition, the combination of intermittent renewable and flexible electricity plants are progressively displacing future installed capacity and production of baseload generation technologies, which include LWRs. Moreover, about half of the U.S. nuclear power plants are in deregulated electricity markets and face daily economic competition from more cost-effective power sources. As a result, the U.S. Energy Information Administration projects that nuclear power’s share of electricity generation will be reduced to about 11 percent by 2050 (EIA, 2021b). However, unanticipated market disruptions can quickly change the economic picture for nuclear energy, and the energy system more generally, as evidenced by the 2022 Russian invasion of Ukraine. These issues will be addressed in greater detail in the parallel National Academies study, Laying the Foundation for New and Advanced Nuclear Reactors, which is expected to be released in early 2023.

Past U.S. Department of Energy proposed justification for reprocessing and advanced fuel cycles: Around 1990, a new justification for reprocessing—the benefit of waste disposal—was promulgated for the U.S. nuclear

power development program. At the request of the U.S. Secretary of Energy, Admiral James Watkins, several organizations, including the Electric Power Research Institute (EPRI) and the National Academies addressed the question “Would the benefits to radioactive waste disposal justify processing of existing spent LWR fuel and deploying liquid metal fast reactors to consume the separated transuranic elements?”

Results from these studies tended to reach similar conclusions:

• According to EPRI, adoption of a process-before-disposal policy for current spent fuel would accrue minimal benefits. “The policy would likely incur a large cost penalty, encounter major institutional difficulties, multiply licensing difficulties, and amplify political and public opposition to the nuclear power program as a whole” (EPRI, 1991).
• According to the 1996 National Research Council report on separation and transmutation (S&T) (National Research Council, 1996):
  • A sustained and carefully focused program of R&D over the next decade “should focus on the factors that strongly influence fuel-cycle economics, especially the costs of reprocessing spent LWR fuel, minimization of long-lived radionuclides to secondary wastes in the reprocessing cycle, and on the need to reduce the possible increase in proliferation risks that could result from the commercial use of plutonium in recycled fuels.”
  • “The current policy of using the once-through fuel cycle for commercial reactors, with disposal of the spent fuel as HLW, should be continued.”
  • “None of the S&T system concepts reviewed eliminates the need for a geologic repository. DOE should continue efforts to develop a geologic repository for spent LWR fuel.”

Based on congressional allocations and administration directions, DOE and its predecessor agencies have, for decades, conducted research on advanced reactor technologies, which differ significantly from existing commercial nuclear plants with respect to both the nature of the design and power capacity per unit. Since the mid-1980s, because of inconsistent or opposing perspectives on the role of nuclear energy for the United States, project outcomes have too often been limited to publication of comprehensive reports repetitively assessing these technologies, and rarely have proceeded to the construction of facilities where new concepts could be tested and evaluated at engineering scale.

DOE represents the U.S. government in the Generation-IV International Forum (GIF), which provides a framework for international cooperation in R&D for the next generation of nuclear energy systems, as further detailed in Chapter 3. This R&D collaboration has been ongoing for two decades and is expected to aid national and international progress toward the realization of such systems. Chapter 3 provides details on DOE’s current advanced nuclear energy and associated fuel cycles programs.

Geologic repository: For the once-through fuel cycle, the major gap is the lack of a permanent geologic repository. As detailed in Chapter 5, the United States presently lacks a strategy for licensing and operating a geologic repository for commercial spent fuel.

Bottom line: U.S. energy policy has been in flux because of short-term oscillations in political support and leadership. Whatever choices are made about advanced nuclear fuel cycles, U.S. utilities will continue to operate LWRs as long as they remain cost-competitive over the next few decades. Seven more LWR units are expected to shut down by the end of 2025. Two new reactors, Vogtle 3 and 4, are scheduled to be in service in the first quarter of 2023 and by the fourth quarter of 2023, respectively, based on information from August 2022 (NEI, 2022).

In the Annual Energy Outlook 2021, the U.S. Energy Information Administration notes that “renewable energy incentives and falling technology costs support robust competition with natural gas as coal and nuclear power decrease in the electricity mix” (EIA, 2021b). Renewable electric generation is projected to meet an increasing share of additional demand. The report continues, “As the share of natural gas-fired generation remains relatively flat, and as the contribution from the coal and nuclear fleets drops by half, the renewables’ share of the electricity generation mix more than doubles from 2020 to 2050” (EIA, 2021b).

2.5.2 France

France relies on nuclear generation for 70–75 percent of its electricity supply—the highest of any nation. This position results from implementing a deliberate national energy policy to prioritize energy independence and security, prompted by the political and economic consequences of the Arab oil embargo of 1973 (Giraud, 1983). In less than a couple of decades, electricity generation in France shifted from about 75 percent dependence on oil to about 75 percent dependence on nuclear energy. Electricité de France, the main electricity generation and distribution company, owns and operates the country’s 56 power reactors. Thus, the French nuclear power system has centralized control in one large utility company.

The low generation costs of this nuclear fleet may be due in large part to an early commitment to standardization on just a few LWR designs (900 MWe, 1300 MWe, and 1450 MWe, all pressurized water reactors). As a result, the average cost of electricity in France is significantly lower than that in the European Union as a whole; for example, in neighboring Spain and Germany, prices are 46 percent and 79 percent higher, respectively, than in France (Selectra, 2021). Furthermore, France is the world’s largest net exporter of electricity (WNA, 2022c). Standardization, however, could have the disadvantage that an equipment failure in one reactor could result in the shutdown of all reactors in that design class, thereby risking shutting down a large fraction of the nuclear power fleet (Ramana and Saikawa, 2011). In 2016, the French fleet manifested this type of flaw when about 20 reactors were shut down to address a carbon segregation problem in the steam generators’ lower plates (Les Échos, 2016).

In 2015, the French Nuclear Assembly passed the Energy Transition for Green Growth law, which endorsed France’s long-standing policy on energy security but sought to increase the share of renewable energies, such as solar and wind, and decrease the share of nuclear energy to 50 percent by 2025. In part, this law recognized that many of the reactors are nearing their nominal 40-year end of operational life. In November 2017, Environment Minister Nicolas Hulot noted that this goal was unrealistic, postponing the reduction to 2030 or 2035 and seeking to extend the operational life of the reactors. In November 2021, French President Emmanuel Macron announced plans to begin construction of new nuclear reactors in order to maintain the share of nuclear power at 50 percent of the electricity mix. In January 2022, the minister for ecological transition stated a target date of 2035–2037 for the new reactors to be commissioned (WNA, 2022c).

The relative stability of French energy policy results from general agreement within the French government on this issue and the absence of major policy shifts in the administration after each election, even considering the implications of the 2015 Energy Transition for Green Growth law. The size and integration of its nuclear industry, owned and controlled in large part by the French government, has allowed for a significant influence on the global nuclear industry and has had technology export benefits. Because nuclear power generates the vast majority of French electricity, France emits much lower amounts of carbon dioxide than other industrialized nations. Thus, nuclear power contributes significantly to French industrial, energy, and environmental policies.

Past experience and current policy on developing nuclear fuel cycles for multiple recycling of fissionable materials: Ensuring sustainable supplies of uranium has always been a top priority of French nuclear fuel cycle strategy. France has little domestic natural uranium. In addition, as mentioned earlier in this chapter, during the early decades of nuclear power, global resources of natural uranium were perceived as scarce, a view that drove France to devote R&D to fast reactors and reprocessing capabilities. However, development of fast reactor technology fell behind the deployment of LWR reprocessing technology. The French experience with fast reactors involved four main projects:

• Rapsodie (40 MWth): 1967–1983
• Phénix (250 MWe): 1973–2003
• Superphénix (1240 MWe): 1985–1997
• Phénix (170 MWe, after safety reevaluation): 2003–2009

According to the Commissariat à l’énergie atomique (CEA), these projects demonstrated the “excellent use of the uranium resource as well as the capability of these reactors to recycle the plutonium without any limitation in the number of recycling operations.” At the same time, Phénix and Superphénix also showed how several material

selections could be unsuitable, and they shed light on sodium fast reactor safety functions, fuel handling, sodium leaks, operations, maintenance, and dismantling (Patel, 2019).

Given the lagging development of breeder reactors, the French program changed its near-term emphasis from the deployment of a closed fuel cycle with multirecycling to reliance on a fuel cycle in which recovered plutonium is recycled once in the form of LWR-MOX fuel (in the French 900-MWe LWR fleet) and the resulting spent LWR-MOX fuel is held in interim storage. This approach, however, has proven unable to prevent the accumulation of a large and growing stockpile of plutonium, totaling nearly 80 MT as of the end of 2020, in the form of both plutonium oxide and unrecycled MOX scrap (IAEA, 2021f). The growing inventory of plutonium-rich scrap generated by production problems at the MELOX MOX fuel fabrication plant has led to an urgent need to increase storage capacity for this material at La Hague (Johnstone, 2022). Faced with limited options for addressing this stockpile, France’s long-term goal remains committed to closing the fuel cycle by recycling the plutonium in a fleet of cost-competitive fast reactors.

To progress toward this long-term goal, in 2006 the French government announced that the CEA would build a fourth-generation (Gen IV) 600-MWth fast reactor prototype, referred to as “ASTRID,”³⁰ by 2020. Initial connection of a first commercial-scale fast reactor to the French electrical grid was expected in the 2040–2070 time frame. However, in August 2019, after completing a detailed design, the plan for construction of ASTRID was shelved until at least the second half of the 21st century because of the sustained abundance and availability of uranium ores at low prices. In the short term, to stabilize the recovered plutonium stockpile and spent LWR-MOX fuel inventories created by the present policy, the CEA mission now includes an assessment, to be completed by around 2040, of the feasibility of multirecycling plutonium in the existing fleet of LWRs. Advanced European Pressurized Water Reactor (EPR) fuel designs are being investigated to enable plutonium multirecycling and stabilization of all spent fuel inventories. Two fuel assembly design concepts are being studied: (1) CORAIL, where the fuel assembly design contains both low-enriched uranium and MOX rods, and (2) MIX (also called MOXEUS), where the fuel assembly design consists of fuel rods containing a mixture of plutonium and enriched uranium oxides. Neutronic simulations indicate that introducing MIX and CORAIL in EPRs by the middle of the century could lead to a fast stabilization of spent MOX fuel and plutonium inventories (Martin et al., 2018). In addition, use of MOX assemblies, presently limited to the 900-MWe reactors, is scheduled to be extended to the 1,300-MWe reactors by the end of the decade.

Geologic repository: ANDRA, the French National Agency for Waste Management, has responsibility for nuclear waste management. In 1991, the “Bataille”³¹ law was enacted, which defined three areas of waste management research: (1) separation and transmutation³² of long-lived radioactive materials in the spent fuel (including minor actinides and some fission products); (2) final disposal of the waste in a geologic repository; and (3) interim storage for up to 300 years. In addition, the Bataille law required a 15-year period after which (in 2006) the French Parliament would create follow-on legislation to specify the path for the French nuclear waste management program.

The 2006 enactment of clarifying legislation included the following:

• Directed ANDRA to receive approval for and operate a permanent geologic³³ repository by 2015³⁴ and 2025,³⁵ respectively.
• Delayed a decision on the implementation of transmutation, possibly until at least 2040, corresponding to a possible time frame for building a new reprocessing facility with updated technology to replace La Hague. For the new facility, three main reprocessing R&D lines were to be maintained, not necessarily

___________________

³⁰ ASTRID = Advanced Sodium Technological Reactor for Industrial Demonstration.

³¹ The law was named after Christian Bataille, a member of the French National Assembly.

³² “Transmutation for waste management purposes is to convert a long-lived radionuclide that is potentially troublesome at a waste disposal site to a shorter-lived or stable nuclide by exposing the troublesome nuclide to a high flux of neutrons for a sustained time” (National Research Council, 1996).

³³ From 1991 to 2006, ANDRA opened an underground research laboratory in a clay formation in Bure (Meuse), France.

³⁴ Approval from the French government of the Bure site (CIGEO project, Centre Industriel de Stockage GEOlogique) was obtained in 2013.

³⁵ In 2021, ANDRA was preparing the license application. If authorization is granted, industrial disposal operations will start around 2035.

separate from each other: evolution of the PUREX process to coextract uranium and plutonium and possibly neptunium (COEX^TM process); selective separation of minor actinides (DIAMEX-SANEX process); and group extraction of all actinides for homogeneous recycling in fast reactors (GANEX). (COEX, DIAMEX-SANEX, and GANEX are described in technical detail in Chapter 4.)

• Did not take any action on long-term interim storage because of the lack of experience of the French industry on this topic.

Bottom line: In the mid-1970s, France instituted a planned energy policy to improve energy independence and security by building a fleet of pressurized water reactors (PWRs). Reliance on nuclear generation for electricity supply reached as high as 80 percent. As of 2022, it stands at about 70 percent and is expected to stabilize at about 50 percent by 2035. The current fleet of PWRs and any new third-generation PWRs—EPRs—are envisioned to continue to dominate the French nuclear fleet through most of the 21st century. According to an Orano representative who presented to the committee in September 2021, the short-term objective (by the end of the 2020s) is to demonstrate mixed oxide fuel in the 1,300-MWe PWRs. The midterm (in the 2040s) ambition is deployment of next-generation mixed oxide fuels to support multirecycling PWRs. In the second half of the 21st century, a fast reactor prototype might be started, with the potential for a future power reactor fleet composed almost exclusively of fast reactors using a closed fuel cycle (Gay, 2021). This scenario, if implemented, would result in France being nearly energy independent for its electric power generation, assuming that the renewable sources also help substitute for use of fossil fuels for electricity production.

2.6 INSIGHTS ABOUT MERITS AND VIABILITY OF FUEL CYCLE OPTIONS FOR EXISTING LWR TECHNOLOGIES

The two partial fuel cycles referred to as once-through and monorecycle were not anticipated to be options during the early days of nuclear power development and deployment. Spent LWR fuel was intended to be reprocessed rather than disposed of in a geologic formation, and recovered plutonium was intended to be recycled in fast reactors, rather than in LWRs, given the much better nuclear properties of plutonium in a fast neutron environment (i.e., in fast reactors) compared with those in a thermal neutron environment (i.e., in LWRs). Both once-through and monorecycling systems are the direct result of adjustments required in light of the lack of progress in advancing fast reactors commercially.

The once-through option is likely to remain an “open” option, meaning that the spent fuel will be stored eventually in a geologic repository and is unlikely to enter a more advanced fuel cycle. The monorecycle option, on the other hand, may end up as either an open fuel cycle or become part of an advanced fuel cycle, as illustrated in Figure 2.3. Many countries (e.g., Belgium, Germany, Switzerland) that originally had reprocessing contracts have generally opted for monorecycling rather than storage but have no intent to further reprocess the spent mixed oxide fuel, which is now earmarked for final disposal. In this case, the monorecycle option will end up being an open fuel cycle. For France, the spent mixed oxide fuel may enter an advanced fuel cycle if and when fast reactors become commercially or strategically viable.

The front end of the once-through fuel cycle for thermal LWRs is well proven in dozens of countries, providing low-enriched uranium fuel (<5 percent uranium-235) at relatively low fueling costs. Globally, considering the current demand for nuclear power, uranium supplies are abundant and plentiful enough based on known resources at current uranium prices to ensure supply of these reactors for at least the next century. The series of steps to fuel an LWR the first time are the same whether or not the fuel cycle includes reprocessing. After reprocessing, a dedicated mixed oxide fuel fabrication facility is needed, which—because of the presence of plutonium—must meet safety and security requirements additional to those required for uranium oxide fuel fabrication. If the reprocessed uranium is recycled, the conversion and enrichment facilities will have to be appropriately modified to accommodate the added beta-gamma radiation associated with reprocessed uranium, as well as to add a separate, dedicated centrifuge cascade if this path is chosen to reenrich reprocessed uranium.

In contrast, the back end of the once-through fuel cycle is not complete. As Chapter 5 makes clear, all nuclear fuel cycles, open or closed, will require geologic repositories. The lack of geologic repositories for permanent

disposal of the spent nuclear fuel and high-level waste presently constitutes the biggest challenge for the national programs that have adopted an open fuel cycle approach. As a result, by both default and necessity, the nuclear industry has figured out how to temporarily manage waste safely in interim storage. In particular, utilities have had to increase the capacity of cooling pools for storing spent fuel assemblies at reactor sites and transfer an increasing fraction of their spent fuel inventory into dry (casks) or wet (pools) interim storage systems. In the United States, dry storage has been widely implemented at the reactor sites. While this system is working, it is not considered a permanent solution. Current practices are optimized for at-reactor-site storage, but not necessarily for transportation and disposal (Freeze et al., 2021), and there may be a need to repackage all of U.S. existing spent fuel prior to disposal in a geologic repository. Similarly, storage approaches are being used for the spent mixed oxides, high-level waste, and other products created by monorecycling.

A merit of the once-through cycle is its proliferation and theft resistance. Chapter 6 examines nonproliferation and security risks for advanced reactors and their associated fuel cycles. Low-enriched uranium at the 3–5 percent enrichment level is not weapons-usable material. Moreover, by not reprocessing and thus not separating out plutonium from highly radioactive fission products, the once-through cycle provides further protection against theft or unauthorized access to the plutonium; this is known as the “spent fuel standard” (DOE, 1999).

As was previously discussed, the original motivation for an advanced fuel cycle relying on the recovery of plutonium from the LWR spent fuel and recycling of the plutonium in a fast reactor was to extend fissile resources. As concerns for the adequate supply of natural uranium resources progressively subsided, the interest in such an advanced fuel cycle shifted to its potential merit to waste disposal, and more specifically to separation and transmutation of minor actinides. These topics are reviewed in more detail in Chapter 4.

This page intentionally left blank.