|
|
| Fig. 1: Relative composition of a 47.5 GWd/t UOx spent nuclear fuel after 4 years irradiation in PWRs, divided into uranium, plutonium, and fission products plus minor actinides. [9] This shows that conventional reprocessing mainly recovers the large uranium/plutonium streams. The "fission products + minor actinides" slice most relevant to advanced waste minimization represents a small fraction of the overall mass. (Image source: L. Garcia) |
One of the largest obstacles to the deployment of future nuclear reactor fleets is how to dispose of the spent fuel waste they would create. Claims that nuclear fuel can be "recycled" are correct but incomplete. Existing commercial reprocessing can successfully recover uranium and plutonium from spent fuel assemblies and send a portion of that material back into the fuel cycle as mixed oxide (MOX) fuel. The Orano-operated La Hague plant in Northern France is the most prominent example of how to reprocess spent fuel, recycle part of it into new fuel, and turns the fraction that cannot be reused into glass (via vitrification). [1] There are ample reports already that tell French MOX story quite well. [2,3] The harder question is: does this kind of recycling actually solve the nuclear waste problem?
No, it does not. Existing conventional reprocessing should be understood mainly as a fuel-cycle strategy, not a way to eliminate high-level waste. Reprocessing allows for a larger portion of uranium fuel to be fully utilized, chemically changes the form of the waste, and can reduce some burdens on waste repositories, but the economics are still sharply contested. [4-6] However the waste problem is largely caused by the residual fission products and the minor actinides that remain once the relatively straightforward chemistry of uranium and plutonium recovery has been done. [4,5] So there is a clear difference between reprocessing and actually solving nuclear waste.
The current industry standard for reprocessing is PUREX, shorthand for the plutonium-uranium extraction process. In PUREX, spent oxide fuel is chopped up to increase surface area, then dissolved in nitric acid, and contacted with an organic solvent containing tributyl phosphate (TBP). In this mixture the uranium(VI) forms a neutral nitrate molecule, and this process can generally be expressed as:
Plutonium can also be carried into the organic phase in the same way and later on be separated from uranium by redox control, since Pu(III) is much harder to extract than than Pu(IV). [7] This process can quite effectively extract bulk uranium and plutonium streams from spent fuel. This is why reprocessing can support MOX fabrication at scale and why France has stayed more committed to it than most other nuclear countries. [1,4]
However if one says that this is "95% recyclable" that statement can be incredibly misleading if it is heard as "95% of the waste problem is solved." Because it isn't. If I take the IAEA reference composition for a standard spent 900 MWe PWR fuel assembly with 500 kg initial uranium: the waste rules out to about 470 kg uranium, 5 kg plutonium, ~0.5 kg minor actinides, and 25 kg fission products, where the recyclable share is about 95%. [4] By mass, that is correct, but in terms of waste significance this is not true.
An examination of the constituents in the waste further underscores this point with the minor-actinide to uranium mass ratio being
while the plutonium to uranium ratio remains
As shown in Fig. 1, the spent-fuel mass is overwhelmingly uranium, with plutonium, minor actinides, and fission products constituting a small fraction of the total mass. Commercial reprocessing is designed for the large, easily recoverable U/Pu streams, however not for perfectly isolating the tiny chemically difficult fraction that dominates the deeper waste argument. [4,5]
 |
| Fig. 2: Evolution of ultimate waste long-term toxicity vs. time for once-through, twice-through, and fully closed fuel cycles. The conventional recycling reduces long-term toxicity, but much larger reductions require a more fully closed cycle with actinide management beyond ordinary MOX recycle. (Image source: L. Garcia, after Taylor. [5,9]) |
The "reprocessing solves waste" argument completely falls apart on physical grounds. The waste burden changes with time, and in the shorter term, heat output is driven mainly by fission products such as 137Cs and 90Sr. [4] On longer timescales, the transuranic inventory becomes increasingly important, and after uranium and plutonium are both separated and recycled, but the remaining burden does not vanish, and minor actinides become central, and long-lived fission products like 129I and 99Tc still matter for long-term dose and repository performance. [4,5]
Taken together, Figs. 1 and 2 show the nuclides that dominate the most difficult long-term waste burden represent only a small fraction of the total spent-fuel mass, which is the central paradox of spent-fuel management. In Fig. 2, the once-through case remains above the uranium-ore reference line for an incredibly long time, the twice-through case slightly improves on this, and only the fully closed case produces a much more dramatic reduction in long-term toxicity. In this case, the "twice-through" scenario corresponds to a limited recycle strategy, while the "fully closed" case represents a much more ambitious cycle with much deeper actinide management. [5]
So MOX recycle is not the same thing as waste elimination. MOX consumes plutonium which essentially extracts additional energy value from the fuel cycle, and this process is useful. However even the most advanced closed cycles still require a deep geological repository, which exposes these cycles to the same political issues that nuclear waste disposal faces to begin with. Advanced fuel cycles can reduce the waste burden on a geological repository through lower thermal loading volume and lower long-term radiotoxicity, but every single credible fuel cycle concept still requires a repository for some waste forms. [5]
The economics also reinforce the point that this is not a simple industrial triumph. Even at La Hague, the world's best-known commercial reprocessing site, Reuters reported in 2015 that foreign customers had dwindled and EDF was pressing for lower prices. [8] Some opponents of reprocessing decry large energy inputs to reprocess the waste. Although it's true that reprocessing is an energy and infrastructure-intensive industrial process, existing analyses indicate it is not the largest contributor to the overall environmental footprint of the nuclear fuel cycle. For example, in the French twice-through cycle, reprocessing contributes only a modest share of total fuel-cycle GHG emissions compared to the mining, reactors, and disposal. [9] Reprocessing has persisted in France because it is deeply embedded in the national strategy for energy and fuel recycling. But it has not made the waste problem trivial.
In order to attack the most difficult waste problem it becomes necessary to go beyond U/Pu recycle and look at the longest-lived waste, and then it becomes a different chemistry problem. The main step ceases to be bulk extraction of Uranium and Plutonium in this case, and becomes the selective separation of minor Actinides (especially americium and curium) from chemically similar lanthanides and from the rest of the high-level waste. In this particular case partitioning and transmutation becomes much more than just reprocessing. [5,10]
Separating these species becomes quite challenging for actinides. Am/Ln separation is incredibly intriguing because it could reduce the long-term radiotoxicity of nuclear waste, but it is extremely difficult because Am(III) and the lanthanides have very similar ionic radii and the same +3 oxidation state. [10] A potential avenue is to oxidize americium to even higher oxidation states that lanthanides do not easily oxidize to, but even that is hard. Li et al. finds the redox potentials of the high-valent Am/Am(III) couples to be roughly 1.68 V for Am(VI)/Am(III), 1.73 V for Am(V)/Am(III), and 2.62 V for Am(IV)/Am(III) versus SCE in 1 M HClO4. [10] These advanced separation methods are also expected to increase fuel-cycle costs. [6]
There are clear engineering problems that need to be solved before this becomes commercially viable. Advanced aqueous flowsheets that have been proposed such as DIAMEX/SANEX can (in principle) push actinide recoveries toward 99.9%. But separating americium and curium cleanly from the bulk lanthanides is a much harder task, and it is a serious issue since lanthanides interfere with subsequent transmutation performance. [4,10]
The efficiency required here is astounding. If I start from a reference value of roughly 0.5 kg of minor actinides in a spent PWR assembly, 99% recovery would still leave 5 g behind, whereas 99.9% recovery leaves 0.5 g. This is a tenfold difference that remains important since the whole point of partitioning is to keep small actinide masses from dominating the long-term waste inventory. Commercial reprocessing is good at the easy part of isolating bulk U/Pu. Advanced partitioning is difficult because of the relative scales on which it operates. [4,5]
Transmutation can only become possible once the separation problem has been solved. In order to achieve this, the separated actinides must be successfully transformed into suitable targets or fuels. They can then be irradiated in a fast reactor or an accelerator. [5] At present this is far beyond the realm of conventional reprocessing. This is an advanced, multi-step waste-minimization program in which chemistry is the gating step.
Reprocessing is a really useful technology that extracts the bulk of the mass from nuclear waste that can be reused in future fuel assemblies. Commercial plants such as La Hague show that spent fuel can be chemically treated so uranium and plutonium can be recovered so that part of that inventory is returned to reactors as MOX. [1,4] However this does not solve nuclear waste. Once the bulk U/Pu layer has been peeled back, the chemically stubborn inventory that remains is more difficult to address.
Reprocessing is a solid first step towards isolating high level waste, but partitioning and transmutation are necessary to achieve a final solution to nuclear waste. Reprocessing aims to achieve complete uranium and plutonium recovery whereas partitioning and transmutation is a much more intense attempt to reduce repository strain by isolating and destroying minor actinides and selected long-lived fission products. As Fig. 2 makes abundantly clear, ordinary recycling helps reduce the mass, but much larger reductions in long-term toxicity can happen only in more advanced closed-cycle methods. A geological repository also still remains necessary. [5] Thus, reprocessing is not the same thing as solving nuclear waste.
© León Garcia. The author warrants that the work is the author's own and that Stanford University provided no input other than typesetting and referencing guidelines. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] B. Mallet, "Focus: France Seeks Strategy as Nuclear Waste Site Risks Saturation Point," Reuters, 3 Feb 23.
[2] K. Friedman, "NuClear or Unclear: French Nuclear Recycling and Its Future," Physics 241, Stanford University, Winter 2024.
[3] C. Cranmer, "Analysis of Commercial Reprocessing Technologies," Physics 241, Stanford University, Winter 2024.
[4] "Spent Fuel Reprocessing Options," International Atomic Energy Agency, IAEA-TECDOC-1587, August 2008, Annex II-3.
[5] R. Taylor et al., "A Review of Environmental and Economic Implications of Closing the Nuclear Fuel Cycle - Part One: Wastes and Environmental Impacts," Energies 15, 1433 (2022).
[6] R. Taylor, W. Bodel and G. Butler, "A Review of Environmental and Economic Implications of Closing the Nuclear Fuel Cycle--Part Two: Economic Impacts," Energies 15, 2472 (2022).
[7] E. R. Irish and W. H. Reas, "The Purex Process - A Solvent Extraction Reprocessing Method For Irradiated Uranium," General Electric, Hanford Atomic Products Division, HW-49483A, April 1957.
[8] E. Jarry, "Crisis for Areva's La Hague Plant as Clients Shun Nuclear," Reuters, 5 May 15.
[9] C. Poinssot, B. Boullis, and S. Bourg, "Role of Recycling in Advanced Nuclear Fuel Cycles," Reprocessing and Recycling of Spent Nuclear Fuel, ed. by R. Taylor (Woodhead Publishing, 2015), p. 27.
[10] B. Li et al., "Exploiting the Coordination Chemistry of High-Valent Americium For Actinide/Lanthanide Separations," Commun. Chem. 8, 260 (2025).