In view of global warming and the world energy crisis, developing a sustainable carbon-free energy mix based on nuclear and renewable energies is mandatory. Improving the use of uranium in nuclear fission reactors involves development of both reactor technology and the accompanying fuel cycle in order to maximize natural fissile resources and reduce the production and impact of waste. This leads to the concepts of plutonium multi-recycling in pressurized water reactors (PWRs) and a closed fuel cycle in fast reactors. Herein, we describe the efforts that are currently underway in France to make these concepts a reality before the end of the century, as well as the associated scientific and technical challenges. We mainly focus on the separation and recycling steps that are at the heart of the Plutonium Futures conference and pay less attention to the front and back end of the fuel cycle.
We begin by detailing the current French fuel cycle, which includes the benefits of plutonium mono-recycling (Part 1). Then we highlight the key points of the French energy policy, translated in the Energy Transition Law for Green Growth and the Multiannual Energy Plan, and its consequences in terms of the nuclear reactor agenda and associated fuel cycles (Part 2). We present both medium- and long-term scenarios to achieve independence from natural uranium and secure long-term sustainability (Part 3). This can be achieved in the long term through plutonium multi-recycling in Generation IV fast neutron reactors (FNRs, Part 5). In the medium term, meanwhile, a step-by-step scenario can be adopted, with benefits gained at each stage, including plutonium multi-recycling in PWRs (Part 4). However, such a trajectory induces new technical issues, related for instance to a continuous increase in the plutonium throughput, thermal power, and radioactivity, and we present some of our promising R&D pathways to overcome these challenges. Finally, we describe the possible management of minor actinides via separation and transmutation for a long-term goal (Part 6).
1. The current French nuclear fuel cycle situation: Back to basics
Fig. 1 shows a simplified description of the current plutonium mono-recycling French fuel cycle, including an optional reuse of uranium from reprocessing, and associated major throughput values. Each year, in order to produce around 1,000 t (metric tons) of uranium dioxide fuels (UOX) for the 56 PWRs in the French fleet, 8,000 t of natural uranium is extracted, purified, and converted into uranium tetraflouride and then uranium hexaflouride. Uranium hexafluoride is enriched in uranium-235 isotope content from 0.71% (natural uranium) to 3–5% by ultra-centrifugation at the Georges Besse II enrichment plant in Pierrelatte. This produces a tail of around 7,000 t of depleted uranium, with a uranium-235 content around 0.3%—a huge, accumulated resource of 320,000 t of depleted uranium is thus available, either for the current MOX (mixed oxide: a mixture of uranium and plutonium dioxide) fuel fabrication, for possible re-enrichment operations, or for future use in FNRs. The UOX fuel is manufactured with 1,000 t of enriched uranium, then irradiated in the French PWRs and finally reprocessed in the UP2-800 and UP3 Orano plants at La Hague after a defined cooling period. This operation allows France to separate around 940 t of reprocessed uranium and 10 t of plutonium annually, while conditioning the remaining 50 t of fission products and minor actinides in 360 t of borosilicate glass containers that are stored at La Hague for transfer to the future deep geological repository, CIGEO, located in a thick Callovo-Oxfordian clay layer at Bure.
Annually, 10 t of plutonium dioxide powders are mixed with depleted uranium dioxide to fabricate MOX fuel pellets by powder metallurgy at the Orano MELOX plant in Marcoule. A total of 120 t of MOX fuel is produced each year and can be used in twenty-four 900 MWe reactors in the French fleet (Fig. 2). After irradiation, the used MOX fuel is stored at the power plants in pools onsite and then at La Hague, waiting for radioactive decay and future reuse. The used MOX fuel contains valuable plutonium, at an average concentration of around 6%, which can be recovered via reprocessing. This has already been demonstrated industrially in France with an operation that used 73 t of MOX at La Hague, which ran from 1992 to 2008. Used MOX fuel is not considered as nuclear waste in France, but rather as a valuable fissile material—it is envisaged to reuse plutonium either by multi-recycling in PWRs or storing for future use in FNRs. Reprocessed uranium can also be re-enriched up to a commercial assay (re-enriched uranium, REU) and reused in the four Cruas 900 MWe PWRs: uranium recycling was performed for 20 years, 1993 to 2013, and could restart soon. In a mono-recycling strategy, used REU dioxide fuels (80 T/y) are stored in pools.
It is insightful to compare the current plutonium mono-recycling strategy with an open cycle (for a given power generation around 350 TWh/y). Annually, the plutonium mono-recycling strategy allows:
- Savings of 1,000 t natural uranium (8,000 versus 9,000 t)
- Savings of waste storage/disposal capacities [(120 t used MOX + optionally 80 t used REU + 360 t high activity glass) versus 1,000 t used UOX]
- Reduction of the net production of plutonium by ~30% (7 t versus 10 t) since a quantity of recycled plutonium is transformed into fission products in the reactors.
This illustrates the benefits of the current French fuel cycle option. We will now describe the benefits that can be gained going beyond mono-recycling, towards a multi-recycling strategy, and finally towards the closed fuel cycle.
2. The French Energy Transition Law for Green Growth and Multiannual Energy Plan
The Energy Transition Law for Green Growth (2015) and Climate Energy Law (2019) aim to achieve carbon neutrality by 2050 and diversify France’s electric mix, with a view to increase energy resilience. The successive Multiannual Energy Plans (MEP 2016–2023 and 2019–2028) set out the government’s strategic priorities and measures to be taken in the next 10 years. The latest Multiannual Energy Plan, published in April 2020, reaffirmed that nuclear is a long-term option but as part of a more balanced electricity mix. As such, nuclear power should decrease to 50% of the electricity generation mix by 2035 (down from 62% in 2022). Additionally, the plan retains the option of building new nuclear reactors and reaffirms the strategy of reprocessing and recycling nuclear fuel until at least the 2040s. Finally, a closed fuel cycle strategy is planned for a more distant period. The total nuclear capacity has been set at a limit of 63.2 GWe, corresponding to the total nuclear power supplied to the grid before the shutdown of the two 900 MWe reactors at Fessenheim in June 2020. Importantly, an increase in renewable energy production was also targeted, with a goal of attaining a 33% share in the electricity mix by 2030, up from 24% in 2022.
More recently, on February 10, 2022, President Macron expressed his wish that, in addition to pursuing the development of renewable energy sources, nuclear energy should be supported and developed to guarantee France's energy independence and achieve carbon neutrality by 2050.
More recently, on February 10, 2022, President Macron expressed his wish that, in addition to pursuing the development of renewable energy sources, nuclear energy should be supported and developed to guarantee France's energy independence and achieve carbon neutrality by 2050. The lifespan of the current reactors should be extended in all cases where possible and, furthermore, a new construction program for six EPR2 (European pressurized reactor) type reactors be undertaken and that studies be started for eight additional reactors.
Despite these important commitments, shutting down the oldest reactors of the fleet should have significant consequences on the current plutonium mono-recycling operation, which is performed in 24 of the oldest 900 MWe reactors of the French fleet (Fig. 2). The MEP strategy states that the treatment-recycling plan is “a major challenge for reducing the volume of radioactive waste.” Thus, in order to maintain the current fuel cycle strategy, the use of MOX fuel in the current 1300 MWe reactors has been scheduled, pending the commissioning of EPR2.
Finally, beyond 2040, strategic decisions within the fuel cycle policy will be taken, based on the R&D conducted in the field of fuel cycle closure, with a two-step approach:
Medium term: plutonium multirecycling in Generation III LWRs using MOX2 fuel assemblies with the aim of stabilizing plutonium and spent fuel inventories. The first objective is to irradiate test fuel assemblies in a PWR reactor, with possible industrial deployment around 2040.
Long term: complete closure of the fuel cycle with plutonium multi-recycling in FNRs in the second half of the 21st century, which will allow France to reach self-sufficiency and independence from natural uranium.
3. The need for a closed fuel cycle in the long term and the benefits of a step-by-step deployment
The International Atomic Energy Agency (IAEA)-Nuclear Energy Agency (NEA) 2020 joint publication “Uranium Resources, Production and Demand” (commonly known as the Red Book) analyzes the balance between natural uranium production and demand up to 2040 (Fig. 3) at a reasonable cost (considered as less than $50/ lb U3O8). By 2040, the projected demand for uranium is covered by the overall potential production, by totaling the existing, planned, and prospective capabilities at an acceptable cost. However, the situation appears to become more challenging in the second half of the 21st century. This is highlighted in the 2018 meta-study by the IPCC (Intergovernmental Panel on Climate Change) “IPCC Special Report on Global Warming of 1.5°C” (Fig. 4).
The IAEA 2020 Red Book states that meeting the highest projection of nuclear electricity demand by 2040 would require the consumption of approximately 28% of the total listed resources in 2019 at a cost of less than $130/kg uranium. We conclude that natural uranium resources at moderate cost will probably be lacking before the end of the 21st century; the spot price of natural uranium should make it largely unaffordable sometime in the second half of the century. This explains the French strategy, moving towards a full closure of the fuel cycle before the end of the 21st century. We now explain a possible way to reach this goal.
A closed fuel cycle requires the full deployment of a fleet of fast neutron reactors. Indeed, the fission/absorption ratio of odd and even isotopes of uranium, plutonium, americium, and curium strongly depends on the neutron energy spectrum (Fig. 5). While fissile isotopes of uranium and plutonium are likely to fission in both thermal or fast neutron spectra, the fission/absorption efficiency is strongly reinforced in a fast spectrum for the other isotopes: uranium-238, neptunium-237, plutonium-238, plutonium-240, plutonium-242, americium-241, americium-243, and curium-244.
Therefore, the multi-recycling of plutonium in thermal spectrum reactors induces a higher consumption of neutrons and an associated increased production of minor actinides. In contrast, the full deployment of fast neutron reactors (Fig. 6) makes it possible to use every actinide isotope and feed the FNR fleet with only recycled plutonium and depleted fertile uranium (without consuming natural uranium and also avoiding the need to maintain a front-end cycle industry).
There are numerous benefits that could be gained deploying a fast nuclear reactor fleet with a closed fuel cycle strategy: no need for natural uranium supply and front-end operations, zero net production of plutonium, no need for large used fuel storage, and decreased production of minor actinides. Nevertheless, there are some important technological issues to overcome to reach this goal. Indeed, assuming a constant electricity output, plutonium throughput would be expected to increase by a factor of eight to ten compared to the mono-recycling currently used in France. In addition, a complete fleet of fast reactors must be deployed, which entails overcoming technical and financial issues and developing an industrial supply chain for reactor components and fuel (fabrication and processing plants adapted to SFR MOX fuel). Therefore, considering that the risk of a natural uranium shortage should not occur for several decades, France plans to adopt a stepwise approach towards a complete closed fuel cycle, with accumulating benefits at each new stage.
After mono-recycling, a possible next step would be to begin plutonium multi-recycling in PWRs by the 2040s by adapting the MOX fuel assemblies and the fuel cycle processes to progressively larger quantities of plutonium, of reduced fissile quality and more highly irradiating. This should not only allow used fuel and plutonium inventories to be stabilized, but also prepare the reprocessing and recycling industry for a future closed fuel cycle.
Before achieving full closure of the fuel cycle, the final intermediate step will be the progressive replacement of the EPRs in the fleet with FNRs. This will gradually decrease the production of minor actinides, and therefore also the volume of glass waste canisters and associated disposal footprint.
4. Multi-recycling of plutonium in PWRs (EPRs)
Here we examine in more detail the origin of the plutonium amounts used in the PWR multi-recycling stage. Fission quality (FQ) of a plutonium batch is defined as its proportion (percent contained) of fissile isotopes of plutonium (plutonium-239, plutonium-241). Plutonium derived from UOX fuel typically has an FQ of 63%, whereas that from used MOX reprocessed fuel is lower, 53% or less. Each cycle in a multi-recycling scheme in a PWR will also decrease the FQ. Therefore, in order to maintain a given burn-up reactivity, it is mandatory to compensate for the decreasing FQ. This can be achieved by using one of two MOX2 fuels concepts (Fig. 7).
The MOX-MR scheme mixes plutonium batches coming from reprocessed used UOX, current MOX used fuels, and a minor amount of multi-recycled plutonium, leading to an FQ ranging from 52.5 to 55%. The plutonium is then blended with depleted uranium dioxide. The lack of reactivity in these assemblies is accounted for by adjusting the reload size: MOX-MR assemblies are therefore very similar to current MOX assemblies.
The MIX scheme uses a similar source of plutonium batches but mixed to give a slightly lower FQ, ranging from 53 to less than 50%. The lower neutron reactivity is compensated for by using enriched uranium dioxide instead of depleted. Because both schemes use additional supplies of uranium, the system is not free from front-end operations at this stage (namely, natural uranium mining, purification, and enrichment).
It is possible to reach a steady state with zero net production of plutonium in the PWR fleet by operating an appropriate mix of UOX (naturally enriched uranium oxide, NEU) and MIX amounts (Fig. 8). Let y represent the share of the MIX fuel assembly throughput in PWR cores (y = MIX/(MIX+UOX)) and x and x’ the average plutonium contents in the fresh MIX fuel (x) and in the used MIX fuel (x’) (averaged values of the three different fuel zones in the fuel assembly). To comply with the safety constraints in the core, the local plutonium content in the fresh fuel must not exceed 12%. Incorporating a wide safety margin from this limit, a value of 8% has been chosen. The plutonium concentration in used UOX fuel is around 1% and x’ is roughly 6%.
Basic calculations lead to the share of the MIX fuel y ≈ 1/3, which means that, in order to reach a zero net production of plutonium, one third of the fuel used in the PWR fleet should be MIX fuel and two thirds should be UOX fuel. It is also easy to calculate the mass throughput of plutonium (mPuMIX) and compare it with the current plutonium mono-recycling value (mPumono ≈ 10 T/y). Assuming a constant value of electricity production, it follows that the ratio mPuMIX/mPumono ≈ 2.7 or mPuMIX ≈ 27 T/y. It is worth noting that attaining a steady state, which means zero increase of the used fuel and plutonium inventories, leads to a large increase in the quantity of circulating plutonium.
Furthermore, the average plutonium FQ slightly decreases with each cycle (see “Pu multi-recycling scenarios towards a PWR fleet for a stabilization of spent fuel inventories in France” by Courtin et al., published in the Journal of Nuclear Science and Technology in 2021) and consequently there is an increase in isotopes with even-numbered mass numbers (plutonium-238, plutonium-240, and plutonium-242). This has several consequences: increasing the thermal power and radioactivity, as well as increasing the production of minor actinides and the accompanying production of helium. This has been taken into consideration in plans to adapt the fuel cycle process to a multi-recycling scheme, with specific R&D studies already performed, as we show below through two examples.
Fabrication: Cryo-milling process development
To limit the release of helium and other gaseous fission products (which leads to a pressure increase in the fuel rod), an ongoing goal of CEA (French Alternative Energies and Atomic Energy Commission) and Orano R&D is to optimize the grain size and increase (U,Pu)O2 homogeneity through advanced fabrication processes. There are three major steps in the MOX fabrication process: powder preparation (with a feed of plutonium dioxide and uranium dioxide powders), pellet shaping, and sintering. Grain size optimization and oxide homogeneity may be achieved by adapting either the first step, powder preparation, or the last, sintering. The first pathway we have studied is changing the current roller-ball mill process into a new cryo-milling approach (Fig. 9).
Experiments show that total cryo-milling may reduce the time required for powder preparation, increase homogeneity, and lead to several other operational benefits such as reduced waste production.
Reprocessing / dissolution: Voloxidation step
To validate these initial results, complementary experiments were performed. Using larger grain sizes of fuel, with potentially greater amounts of plutonium, slows down the dissolution of used fuel pellets in nitric acid at the front-end reprocessing step. Slower dissolution is also observed with MOX fast-neutron fuel pellets, owing to the significantly higher plutonium content required for achieving a fast-neutron spectrum. To facilitate the dissolution of these pellets, CEA and ORANO are currently studying the voloxidation pathway. Voloxidation induces an over-oxidation of UO2 into U3O8, leading to swelling of the uranium lattice and a fragmentation of the oxide. The larger surface area of the oxide significantly increases the rate of dissolution. The impact of this oxidizing treatment and the behavior of plutonium at the dissolution step is currently under study and could also be beneficial for SFR fuels. These studies aim to increase the TRL (technology readiness level) of the nuclearized voloxidation process.
5. Long-term closed fuel cycle in fast neutron reactors
A simplified view of the closed fuel cycle scenario is shown in Fig. 6. Fast neutron reactors allow higher burn-up values, roughly 150 GWd/T (versus 50 GWd/T for the current PWR fleet) but they require higher plutonium contents, typically 22 to 28%. Assuming that electricity production is fixed at the current value (around 380 TWh/y) and considering the above increased burn-up value, this gives an annual MOX fuel throughput value for FNRs of around 400 T/y (versus 1200 T/y for the current fleet), which also roughly yields the amount of depleted uranium consumed each year. Considering the above range of plutonium content, we obtain the value of annual plutonium throughput at roughly 90 to 110 T/y plutonium in the FNR fleet and fuel cycle plants. We conclude that, compared to the current plutonium mono-recycling value (around 10 T/y), the inventory of plutonium in the closed fuel cycle is multiplied by a factor of roughly eight to ten and by a factor of three compared to the multi-recycling of MIX fuel in PWRs, which is a considerable challenge. More precisely, the challenges and R&D goals associated with a closed fuel cycle are the following:
- Larger quantities of MOX fuels to be treated. Therefore, it is mandatory to develop simplified extraction processes (see below).
- Increased plutonium content in the MOX fuels. This makes dissolution slower and less efficient (see the voloxidation pathway above).
- Strongly increased burn-up values. This tests the integrity of the materials used in the reactor. Oxide dispersion strengthened (ODS) ferritic steel is being developed as an advanced cladding tube because of its excellent swelling resistance and high temperature strength. This work is important in the framework of the collaborative European MATISSE project, which aims to develop materials for safe and sustainable nuclear power.
- Increased content of fission products and platinum-group metals in the used fuel (associated with high burn-up values). This would likely require some improvements in high-level waste conditioning (N.B., in this definition, the radioactivity of high-level waste is of the order of several GBq/g).
CEA, with the support of partners Orano and EDF, are conducting several R&D projects to develop future closed fuel cycle processes. As an example of one of these projects, we describe the principle of a simplified separation for fast MOX fuel reprocessing (see Sorel et al., Proceedings of the GLOBAL 2017 conference, Paper A221.). Due to the limited selectivity and properties of the TBP extractant molecules used in the PUREX process, the current plutonium/uranium separation combines two steps: an initial extraction cycle (with a specific redox operation to re-extract plutonium from the organic phase), followed by two cycles for uranium and plutonium purification. The use of new, highly selective mono-amide type solvents (DEHiBA, DEHBA, etc.) simplifies and reduces the process to a single extraction cycle, with no redox operation as plutonium and uranium are simply re-extracted by decreasing the acidity of the aqueous phase (Fig. 10). This simplified process has been successfully tested in shielded cells in the Atalante hot facility in Marcoule, France, first with surrogate solutions and then on real PWR MOX fuel dissolution solutions.
CEA, supported by Orano and EDF, continues to study new molecules to improve extraction schemes and separation coefficients for different elements. These studies will also address the separation process as a whole, including aspects such as the implementation of industrial equipment and the management of secondary flows in scrubbing operations that remove degradation products, in order to improve the TRL of this process.
6. A path towards transmutation of minor actinides
Beyond the multi-recycling of plutonium, which is clearly a high priority, the separation and transmutation of minor actinides is a more distant future goal. This aims to decrease the short- and medium-term thermal power of high-level waste, to reduce the volume of waste delivered to a disposal facility, and reduce its long-term radiotoxicity. The French reference path involves using sodium-cooled fast reactors (SFR) to transmute americium and even curium. Molten salt reactors are a new option that we also currently study, with some potential benefits (e.g., no handling of solid fuels and targets) but also with a low TRL at present.
A great deal of experience has been gained following both the first Waste Management Act (1991–2006) and the second Waste and Radioactive Products Management Act (2006–present). CEA has used two major R&D facilities for this separations research: the PHENIX SFR (closed in 2009) and the Atalante hot facility, both in Marcoule, France. The first step is to separate americium and curium from fission products and other actinides uranium, neptunium, and plutonium, using a chained PUREX/DIAMEX/SANEX process. Following this, doped targets or fuels are fabricated using the minor-actinides and then they are finally irradiated in a fast neutron reactor (Fig. 5). The feasibility of the transmutation process has been proven through several experiments, in particular for americium-241, and also to a lesser extent americium-243 (see the ECRIX results in Fig. 11). These results and more information can be found in the second volume of CEA experimental work published in 2012 (“Tome 2: La séparation et la transmutation des éléments radioactifs à vie longue”). Despite their proven feasibility, these processes would need some improvements before moving towards a potential industrial use.
7. Summary
The use of nuclear energy will increase and diversify in the coming decades: it is therefore necessary to preserve natural uranium resources in a competitive world. France has chosen to recycle reusable fissile materials by mono-recycling of plutonium and by commissioning feasibility studies on multi-recycling and fuel cycle closure, which are underway. The multi-recycling of plutonium in PWRs would be a first step before the complete closure of the cycle using FNRs, thus allowing France’s independence from natural uranium resources.
This path towards multi-recycling and closing the fuel cycle before the end of the 21st century requires developing new processes; in particular, accounting for the increasing throughput of plutonium (by a factor of three to ten), but also its changes in isotopic composition, thermal power, and radioactivity, and the future diversity of fuels.
To make these developments possible, CEA with its industrial partners Orano, EDF, and Framatome, are conducting several important R&D programs on innovative reprocessing and fuel fabrication processes, in order to prepare the French nuclear industry for future fuel cycle goals.