HORNIG's Real Test Is Not the Fuel Concept — It Is Whether a US Transuranic Supply Chain Exists

ARPA-E's High-performance Optimized Recycled Nuclear Isotopes for Gen IV reactors (HORNIG) programme has selected eight projects to share a funding ceiling of up to $50 million, aimed at turning transuranic elements in used fuel into qualified fuel for advanced reactors. The selections were announced on 29 October; the wider press cycle, including the Deputy Secretary of Energy's appearance at Texas A&M, ran into early November.

Two points about the money deserve setting straight before anything else. First, the ceiling is not the spend. The eight awards sum to about $44.9 million — roughly 90% of the programme ceiling — leaving some $5 million unallocated or held in reserve. Second, the awards are small by capital-project standards but unusually large by ARPA-E standards, where individual awards have historically been capped near $10 million. That sizing tells you what this programme is: not a reactor programme, and not a demonstration plant, but a bet on the fuel front end.

The composition of the portfolio

Figures from ARPA-E project descriptions as reported by the American Nuclear Society; note Sandia's own release rounds its award to $3.38 million and World Nuclear News rounds several figures to one decimal place.

The distribution is worth reading closely. Three awards — nearly $21 million, or almost half the money — go to molten salt and liquid fuel concepts (INL's chloride salt, Texas A&M's thorium-bearing salt) or to a metallic fast-reactor fuel (Oklo). Two go to coated-particle and ceramic routes (Standard Nuclear's TRISO, OSU's nitride). One goes to a light-water reactor fuel assembly concept. Two relatively small awards fund the unglamorous work — feedstock logistics, cost modelling, manufacturing scale-up constraints — that determines whether any of the others matter commercially.

That last pair is not a rounding error. Exodys ($1.3 million) is tasked with a common framework for evaluating TRU feedstocks, fuel-cycle costs, transport and deployment requirements across all HORNIG concepts. Sandia's TRU-PATH ($3.4 million) covers safety, safeguards, technoeconomics and production constraints at commercial scale. Together they are the programme's admission that the fuel chemistry is the easier half of the problem.

What is actually hard here

Reprocessing-derived fuel is not new physics. Mixed oxide (MOX) fuel containing plutonium has been fabricated and irradiated at industrial scale in France, and the feasibility of recycling transuranics into new fuel forms has been shown at laboratory scale for years. What has not been shown is a US supply chain that can deliver separated transuranics in the quantities, forms and purity a fuel fabricator needs — with the materials accountancy that safeguards require.

That is where HORNIG's predecessor matters. CURIE (Converting UNF Radioisotopes Into Energy), launched in 2022, funds separations, accountancy, monitoring and facility-design work to recover actinides from used fuel. HORNIG assumes CURIE's output. According to ANS's reporting, only one of the eight HORNIG selectees — INL — also advanced a project through CURIE. In other words, the fuel developers being funded to consume a feedstock are, with one exception, not the teams being funded to produce it. The two programmes are running on parallel tracks rather than a linked one.

This is the central risk in the portfolio, and it is a scheduling risk rather than a scientific one. A TRISO fabricator can develop a kernel and coating process; a nitride developer can demonstrate a low-temperature synthesis route; a metallic fuel team can produce qualification-relevant data. None of that closes the loop without kilogram-scale quantities of separated americium, curium and plutonium in a specified form, from a facility that does not yet exist at commercial scale in the United States.

The seven-year qualification target

ARPA-E's stated programme metric — qualification and regulatory acceptance within seven years, against a current baseline ANS reports as more than a decade — is the most aggressive commitment in the package. It is achievable in principle only if testing, modelling and manufacturing development run concurrently rather than sequentially, which is precisely what the programme asks applicants to do.

The comparison with historical practice is instructive. Qualifying a new LWR fuel design conventionally takes a decade or more because irradiation campaigns are slow and the regulator wants statistically robust data across burnup, temperature and transient conditions. Compressing that to under seven years means fewer, better-instrumented irradiation experiments; heavier reliance on modelling and simulation to substitute for test matrix coverage; and early, informal engagement with the NRC on the acceptance criteria themselves. None of that is impossible, but it shifts the burden onto data quality and regulatory pre-alignment in a way that a research grant cannot guarantee. HORNIG funds the work; it does not fund the NRC's review capacity or its willingness to accept a novel evidence base.

The Westinghouse award illustrates a related subtlety. PowerBlock is described as additively manufactured metallic fuel for light-water assemblies targeting 24-month cycles above 4,200 MWt. Westinghouse is the largest incumbent fuel supplier — roughly 40% of the global fleet's fuel, per ANS — and it currently supplies no metallic fuel. So a programme framed around transuranics and Gen IV reactors is also funding the incumbent's effort to manufacture a metal fuel for the existing LWR fleet by laser powder bed fusion. Whether PowerBlock ultimately carries TRU or conventional uranium, the manufacturing capability being built is dual-use. That is probably the point.

Commercial read-across

The programme's most consequential near-term effect may be on the HALEU debate. Standard Nuclear's project is explicitly positioned as an alternative to high-assay low-enriched uranium: an MCO TRISO using recycled transuranics and uranium, diversifying fissile feedstock while drawing down commercial spent fuel inventories. The company — which acquired Ultra Safe Nuclear's assets after its bankruptcy and has TRISO plants under construction in Idaho and Tennessee — is arguing that recycled material is a second supply route for the same customers currently queuing for HALEU allocation.

That argument is compelling in principle and unproven in practice. Recycled TRU feedstock carries handling, transport and safeguards costs that HALEU does not, and the fuel must still be qualified. Exodys's cost framework is the mechanism by which the programme will find out whether the economics actually clear. Watch that small award more closely than the headline ones: if the levelised cost of fuel from recycled transuranics cannot approach the programme's target, the diversification case weakens considerably, whatever the physics delivers.

For Oklo, FORGE extends an existing arc. Aurora is a liquid-metal-cooled, metal-fuelled fast reactor; FORGE develops U-TRU-Zr from electrorefiner product, and Oklo separately holds ARPA-E funding for electrorefining commercialisation. The company did not, as far as can be established, issue its own announcement of the HORNIG selection. The technical logic is nonetheless coherent: a fast reactor is the natural burner for transuranics, and vertical integration into recycling would give Oklo a fuel story its competitors cannot easily replicate.

What to watch

The measure of HORNIG in five years will not be how many fuel forms reached irradiation. It will be whether a separated-TRU feedstock of specified quality became available on schedule, whether the NRC accepted a compressed qualification evidence base, and whether Exodys's cost work showed recycled fuel competing with enriched uranium rather than merely with HALEU scarcity. If the feedstock side slips — and the CURIE-to-HORNIG discontinuity suggests it might — the eight projects will produce excellent metallurgy and a stranded commercialisation pathway.

A programme that funds the engine but only partly funds the pipeline is a deliberate risk allocation, not an oversight. It is also the reason to read the cross-cutting awards as the leading indicators rather than the supporting cast.