US Nuclear Fuel Cycle

• The USA is highly experienced in the nuclear fuel cycle, having operated the full range of fuel cycle facilities either commercially or for research.
• Since a high in 1980, domestic uranium mining has withered and now supplies only around 5% of the country’s needs, though that figure has begun to increase.
• Between 1977 and 2005, government policy did not allow reprocessing of used fuel for commercial reactors.
• The federal government is supporting a major expansion of nuclear power and uranium enrichment with bipartisan support.

Introduction

The initial build-out of infrastructure for uranium mining, enrichment and conversion in the 1950s was driven by military needs. Soon, those drivers declined and were overshadowed by the demands of a large and growing commercial nuclear power sector.

Domestic uranium mining peaked in 1980 before declining as reliable international markets met demand for uranium more cheaply. In 2024 the USA had 2.5% of global uranium resources and provided 0.4% of global production.

In the 2000s, legacy gaseous diffusion facilities for uranium enrichment were made uneconomic by energy efficient centrifuges in operation overseas. The need for domestic enrichment was mitigated by a large supply of low-enriched uranium (LEU) sourced from an intergovernmental agreement with Russia to downblend highly enriched uranium from its weapons stocks. From 1994 to 2013 this highly successful 'Megatons to Megawatts' programme provided enough LEU to fuel half of the US power reactor fleet. Even after that ended, Russian enricher Tenex continued significant supply of enrichment services to the USA.

The USA has therefore long been reliant on imported nuclear fuel in the front end of the fuel cycle (mining, conversion and enrichment), despite its extensive knowledge and resources. This potential security of supply issue was brought into sharp focus in 2022-4 with pieces of legislation that prohibited imports of Russian uranium (with some exceptions). As well as fuel for current reactors this also took away the only commercial source of the high-assay low-enriched uranium (HALEU) that advanced reactor designers were beginning to demand.

Current bipartisan policy efforts focus on establishing US self-sufficiency by urgently incentivizing production of LEU for the current fleet needs and preparing a US supply chain of HALEU.

In the back end of the fuel cycle (used fuel storage, reprocessing, recycling, decommissioning and waste disposal) US success has been patchy. Despite successful facilities in the scientific and military sectors, shifting government policies led to the failure of reprocessing, recycling and used fuel disposal for the commercial fleet. Used reactor fuel is being safely managed mainly at reactor sites by the commercial industry in lieu of the federal disposal facility mandated by law.

The US strategy for decommissioning power reactors has been a notable success. Sufficient funds were built up over time for retired power plants to be fully dismantled and decommissioned entirely by the commercial industry as a highly profitable exercise.

Security of supply

Lawmakers in the USA have voiced concern that the country relies heavily on imported uranium and fuel cycle services. As of 2026, US nuclear power plants consume about 19,000 tonnes of uranium per year, while US uranium mines produce about 260 t.

In July 2018 the US Department of Commerce launched a ‘Section 232’ investigation into whether the USA’s reliance on imported uranium posed a threat to national security. Such processes result in a petition to the President who could then choose to address the matter using restrictions or tariffs. In July 2019 the petition was rejected by President Donald Trump, who instead created the Nuclear Fuel Working Group to report on ways to “reinvigorate the entire nuclear fuel supply chain.”

In September 2022 as part of the $47 billion supplemental funding request in the short-term spending bill of President Joe Biden’s administration, $1.5 billion was requested “for the acquisition and distribution of low-enriched uranium (LEU) and high-assay LEU (HALEU)… to reduce the reliance of the United States and friendly foreign countries on nuclear fuels from the Russian Federation and other insecure sources of LEU and HALEU.”

At the end of September 2022, in an amendment of the National Defense Authorization Act, a bipartisan group of US Senators introduced a ban on Russian nuclear fuel imports. The amendment would see the prioritization of domestic production of LEU and the acceleration of efforts to establish a domestic HALEU capability.

In October 2023, the Nuclear Fuel Security Act was approved by the House of Representatives’ Energy and Commerce Subcommittee on Energy, Climate, and Grid Security. The Act aims to establish and expand critical US nuclear fuel programmes to boost domestic uranium mining, production, enrichment, and conversion capacity.

In May 2024, President Biden signed the Prohibiting Russian Uranium Imports Act. The legislation banned the import of Russian-produced unirradiated LEU as well as unirradiated LEU that has been swapped for banned uranium.

The bill came into effect in August 2024 and will last until the end of 2040. Waivers may be granted to allow the import of limited amounts of LEU until 1 January 2028 if one of the following criteria is met:

• The import must be necessary to maintain the viability of a US nuclear energy company that is critical to the US nuclear energy fuel supply chain.
• The import must be intended to support an existing arrangement to provide fuel for a nuclear power plant in another country and thus minimize the likelihood of that country seeking a non-US fuel supplier.

The US Department of Energy (DOE) has committed to acquiring LEU generated from new sources of US domestic enrichment capacity. These can include new enrichment facilities or projects that expand the capacity of existing enrichment facilities. The DOE intends to sell the LEU to utilities operating US reactors to support clean energy generation and sever reliance on Russian imports.

In July 2024 Centrus received a waiver to allow it to import LEU from Russia for delivery to US customers in 2024 and 2025.

At the end of 2025, uranium was added to the US Geological Society’s list of critical minerals.

Details of specific financial awards and actions taken to bolster security of supply in various areas of the fuel cycle are described below in the sections on HALEU demonstration and Commercial HALEU production, below.

Uranium resources and mining

The USA ranks 14th in the world for known uranium resources in the category up to $130/kgU ($50/lb U₃O₈), with 67,800 tU (reasonably assured resources, 2023), about 1% of world total.

Some 40% of resources are in New Mexico and amenable to in-situ leaching (ISL, also referred to as in situ recovery, ISR). Resources on the western Colorado Plateau require conventional mining and milling, as does high-grade breccia pipe mineralization in northwest Arizona. Smaller resources are in Utah, Nebraska and Texas. Production potential is about 50% underground mining, 26% open pit and 24% ISL.

In the 1950s, the USA had a great deal of uranium mining, promoted by federal subsidies. Peak production since 1970 was 16,800 tU in 1980, when there were over 250 mines in operation. This abruptly dropped to 50 in 1984 when 5700 tU was produced, and then there was a steady decline to 2003, by which time there were only two small operations producing a total of under 1000 tU/yr, or about 5% of the uranium consumed by US nuclear plants. Most US production has been from New Mexico and Wyoming.

Exploration expenditure and interest in re-opening old mines has tracked the uranium price through lows in the 1990s, highs around 2007-2008 and lows again through the 2010s.

In mid-2009, the Nuclear Regulatory Commission (NRC) issued a generic environmental impact statement (EIS) on ISL mining in western USA. This streamlined the requirement for a supplementary EIS for each new mine.

Since 2022, exploration activity has been on the increase, and decisions have been made to restart many mines with some reaching production in 2024.

Details of US uranium mining are in US Nuclear Fuel Cycle Appendix 1: US Uranium Mining and Exploration.

See also, Paducah Laser Enrichment Facility (PLEF), below, which would re-enrich depleted uranium hexafluoride tails to create a secondary supply of natural-grade uranium.

US uranium production, tonnes

Conversion

Metropolis Works in southern Illinois is the only conversion plant in the USA. It was built in the late 1950s to fulfil a government contract before being mothballed in 1964 and then restarted as a commercial facility in 1967.

The plant is owned and operated by Solstice Advanced Materials (spun off from Honeywell in 2025) and its output is marketed by ConverDyn, a partnership between Honeywell and General Atomics.

Metropolis’s original nameplate capacity was up to 15,000 tU/yr, but this was reduced to 7000 tU/yr in 2017 in light of poor market conditions. Later in 2017, the plant was idled. In March 2020, the NRC renewed Metropolis’ operating licence for an additional 40 years to March 2060.

In December 2022 ConverDyn received a $14 million award for uranium conversion services from the DOE under its $75 million programme to create a domestic uranium reserve recommended by President Trump’s Nuclear Fuel Working Group. Metropolis was restarted after being upgraded in July 2023 with a licence to produce up to 15,000 tU per year, but producing at 50-70% capacity. Solstice said in February 2026 it was working to ‘de-bottleneck’ the plant and increase output. ConverDyn said in April 2026 it was considering building an entirely new conversion plant at the Metropolis site.

In September 2025, Uranium Energy Corp launched a subsidiary to develop a new US conversion facility to be co-located with a new refining plant near its uranium mining operations in Wyoming. By March 2026 it had secured state regulatory approval and had submitted a Letter of Intention to the NRC. The plant could have a capacity of 10,000 tU per year, according to statements.

Enrichment

Three uranium enrichment plants – at Oak Ridge, Tennessee; Paducah, Kentucky; and Portsmouth, Ohio – were built by the federal government in the 1940s and 1950s for defence purposes and operated for about 25 years. From 1969 to 1992, the US Department of Energy (DOE) and its predecessor agencies sold some of the plants’ enrichment services commercially under full-cost recovery contracts that required utility customers to pay for the plants’ future decommissioning. These fees were collected once, and then further fees were levied from 1992 for decommissioning.

The National Nuclear Security Administration (NNSA) requires a steady supply of HALEU at 19.75% for research reactors as well as defence purposes, including tritium production and naval reactors. NNSA can only use uranium for national security that has been enriched by US-origin technology and this has positioned it as the custodian of the only such technology – the American Centrifuge AC100 centrifuges. These were developed at Oak Ridge National Laboratory in the 1980s before a failed attempt at commercialization by USEC (now Centrus) in the 2010s, after which ownership reverted to the DOE. Now, one cascade of AC100M centrifuges is operated by Centrus in Piketon as a demonstration of commercial HALEU production under DOE funding.

The US power reactor fleet requires about 13 million SWU per year in enrichment.

The USA has two operating enrichment plants: Urenco USA’s facility at Eunice, New Mexico has capacity of 4.8 million SWU per year and is being expanded; and Centrus’ American Centrifuge plant at Piketon, Ohio, although this produces HALEU not yet used by commercial reactors.

Enrichment capacity in operation

Enrichment capacity planned and proposed

LEU

Urenco USA

Urenco USA, the US subsidiary of Urenco (jointly owned by Germany, the Netherlands and the UK), has a centrifuge enrichment plant at Eunice, New Mexico. It uses 6th generation Urenco technology and was planned by the Louisiana Energy Services (LES) partnership – comprising Urenco, Exelon, Duke Power, Entergy, and Westinghouse. Construction of the $1.5 billion plant was licensed by the Nuclear Regulatory Commission (NRC) in mid-2006 when as agreed the three utilities then passed their share to Urenco.

In June 2006 the NRC issued a construction and operation licence Urenco, permitting enrichment up to 5% U-235. Commercial operation began in mid-2010, with phase 1 capacity of 1.6 million SWU/yr (21 cascades of TC-12 centrifuges) being reached in 2012. Phase 2 of 2.1 million SWU/yr was inaugurated in August 2012 with the commissioning of the first new type of centrifuge, and completion to 3.7 million SWU then licensed capacity was reached in April 2014. Phase 3 began operation in 2015, taking capacity to 4.8 million SWU/yr but with the second part of phase 3 deferred until market conditions improve.

A new TC-21 centrifuge design with 60% greater capacity is being used in phases 2&3, and the plant is licensed to enrich to 5.5% U-235.. HALEU of 18-19% U-235 – for new designs of advanced reactors – would require a further cascade.

In November 2012 it was reported that Urenco had approached Korea Electric Power Corporation (Kepco) to invest in phases 2&3 of the plant, which would have had strategic significance for Korea in relation to its nuclear agreement with the USA. Kepco buys 2.9 million SWU/yr of enrichment services from several international sources. This did not proceed.

The plant provides one-third of US enrichment demand from 63 cascades in operation and is a major step forward in underwriting new US nuclear generating capacity and in ensuring security of fuel supply, with flexibility of operation enabling more energy input to produce more fuel from the same natural uranium feed if required.

In May 2014 Urenco applied to the NRC for a licence amendment to allow it to use high-assay tails (c. 0.4% U-235) as feed in part of the new plant expansion applied for in 2012, and also to increase capacity to 10 million SWU/yr. This was approved in March 2015. The expansion would involve three phases, with three new separation building modules.

In February 2019 Urenco said it was planning production of HALEU. It was considering construction of a dedicated HALEU plant for this, particularly to supply a new generation of advanced reactors. In May 2020 the NRC approved a licence amendment for the plant to enrich to 5.5%.

In May 2025 Urenco USA started up one new gas centrifuge cascade as of its Eunice. This was followed in September 2025 by the startup of a second new cascade. This project is progressively adding 700,000 SWU/yr capacity to the plant, which was at 4.3 M SWU/yr previously.In 2026 executives said a further expansion of 2.1 million SWU/yr would take place between 2032 and 2036.

In October 2025 Urenco received authorization from the NRC to produce uranium enriched up to 10% (LEU+), with initial production planned for later that year and first deliveries to a fuel fabricator planned for 2026.

Project Ike

Orano’s US subsidiary Orano Federal Services is planning to construct and operate a large enrichment plant to be named Project Ike in a reference to President Dwight ‘Ike’ Eisenhower and his Atoms For Peace speech.

Project Ike will be sited in Roane County, Tennessee, on a largely unused area of land near Oak Ridge. It is planned as a 70,000 m² facility similar to Orano’s George Besse II plant at Tricastin in France, with capacity dependent on customer demand.

In January 2025 Orano notified the NRC that it intended to apply for licences to construct and operate Project Ike in the first quarter of 2026.

In January 2026 the DOE awarded Orano a $900 million contract to support its plans for Project Ike.

Orano’s environmental impact assessment submitted to the NRC in February 2026 said Project Ike would have a maximum capacity of 7.5 million SWU/yr (the same as Georges Besse II). The licence application was accepted by the NRC for an accelerated 12-month review in May 2026.

Orano was proposing construction of the $5 billion project to begin in September 2027 with operation beginning in January 2032. The plant would be gradually ramped up to full production in the period to 2039. Orano intends to supply LEU and LEU+ at up to 10% enrichment.

Paducah Laser Enrichment Facility (PLEF)

In November 2016 Global Laser Enrichment (GLE) reached an agreement with the DOE for it to sell about 300,000 tonnes of high-assay UF₆ tails (above 0.34% U-235) to be re-enriched to natural uranium level (0.7% U-235) at the proposed Paducah Laser Enrichment Plant (PLEF). In June 2020 the 2016 GLE agreement with the DOE was extended.

With a capacity of up to 6 million SWU/yr, PLEF would produce about 100,000 tonnes of natural-grade uranium over 30 years or more. The DOE would dispose of the reduced-assay balance. Under this plan, PLEF would become a new source of secondary supply of uranium. In the long term it would compete with new mines, rather than other enrichment plants.

In 2024 GLE bought land at Paducah adjacent to the cylinder yard where UF₆ tails are stored.

GLE is a project company developing laser enrichment technology from Australia. It operates a test loop at Wilmington, North Carolina and has announced plans for commercial plants at both Wilmington and Paducah, Kentucky.

In 2006 Silex Systems of Australia and GE Energy (now GE Vernova) received US government approval to develop and potentially commercialize the third-generation SILEX (Separation of Isotopes by Laser Excitation) uranium enrichment process using laser technology in the USA. It provided for GE to construct in the USA an engineering-scale test loop, then a pilot plant or lead cascade, and eventually expand to a full commercial plant. GE was funding the development and making a series of licensing payments to Silex Systems.

The SILEX technology was rebadged as Global Laser Enrichment (GLE) which became the name of the project company. In mid-2008 Cameco bought into the GLE project, paying $124 million for a 24% share, alongside GE (51%) and Hitachi (25%).

In February 2019 Silex Systems and Cameco bought the 76% of GLE held by GE and Hitachi for $20 million on a deferred payment basis, leaving 49% with Cameco and 51% with Silex Systems. Cameco has an option to purchase an additional 26% of GLE. The agreement called for Silex and Cameco to pay $300,000 per month to complete construction of the prototype enrichment facility, known as the Wilmington Test Loop, which had been partially built by GE. Final US government approvals were granted in January 2021. Silex said the restructuring would "provide an ideal path to market for GLE and for the continued commercialisation of the SILEX technology in the US.”

In October 2025 GLE announced the completion of demonstration work at the Wilmington Test Loop. The company said attention would turn to deployment at Paducah.

LIST Island

In January 2026 LIS Technologies (LIST) announced plans to build a laser enrichment facility at Oak Ridge, Tennessee, based on its Condensation Repression Isotope Selective Laser Activation (CRISLA) technology. The EU-3 Facility would be constructed adjacent to the site of the former K-25 gaseous diffusion plant on a 206 acre (83 ha) parcel of land known as Duct Island. LIST bought the land for $8 million and has announced a name change to ‘LIST Island’.

LIST plans to begin construction in 2026, with first production following in 2030. LIST said the plant would represent an investment of $1.38 billion. The company did not reveal proposed capacity.

In December 2025 the state of Tennessee awarded LIST a radioactive material licence for its Test Demonstration Facility, also at Oak Ridge.

USEC, Centrus and SWU supply from Russia

From 1994 until 2013 USEC was the agent for supply to the USA of downblended Russian uranium from weapons stockpiles (see Megatons to Megawatts section of US Nuclear Power Policy paper and Military Warheads as a Source of Fuel). This arrangement met about half of US demand and represented about 5.5 million SWU/yr.

In March 2011 USEC signed a further contract with Russian enricher Tenex for supply of low-enriched uranium from 2013 to 2022, ramping up to about half of earlier levels, with an option to match those levels. The 2011 contract covered the supply of 21 million SWU, worth $2.8 billion. The supplies came from mined uranium enriched in Russia, rather than recycled weapons. The arrangement saw USEC purchase the SWU as LEU and deliver a corresponding amount of natural uranium to Tenex for the uranium component of the LEU, so like its predecessor, the contract yields no net uranium for the USA.

This renewed supply of Russian enrichment to the USA was intended to bridge the gap until USEC could open its own enrichment plant, the American Centrifuge. However, this project ultimately failed, resulting in USEC’s bankruptcy and rebranding as Centrus Energy in 2014. (See USEC/Centrus American Centrifuge plant, in the Appendix)

In December 2015 the 2011 agreement was modified to reflect lower enrichment demand and extended to 2026 for delivery of the balance of the contracted SWU (17 million SWU over 2016 to 2022).

In July 2024 Centrus received a waiver under the Prohibiting Russian Uranium Imports Act to allow it to continue importing LEU from Russia for delivery to US customers in 2024 and 2025.

See also Cancelled enrichment facilities in the Appendix.

HALEU demonstration

Several initiatives have been taken by the DOE to establish a domestic capability and market for high-assay low-enriched uranium (HALEU). This is uranium enriched to between 5% and 19.75% U-235 compared to the 5% typical of LEU and natural level of around 0.7%. Several advanced reactors being developed in the USA and other countries would require HALEU, but licensing conditions demand dedicated enrichment cascades, storage and transport containers. This would require significant investment in new capacity several years before firm orders for the fuel could be placed, while the lack of a HALEU supply chain would itself hinder those orders.

This impasse was foreseen in 2016 and suggestions were made by the Nuclear Infrastructure Council to create a strategic reserve of HALEU to avoid US technology developers having to rely on imported HALEU, Russia’s Tenex being the only commercial supplier of HALEU at the time. Suggestions were made to make use of recycled naval fuel and to limit the downblending of unwanted weapons stocks to HALEU levels rather than LEU levels. (See Military surplus and other government stocks, in the Appendix.)

In June 2019 the DOE awarded a $115 million contract to American Centrifuge Operating, a subsidiary of Centrus Energy, to demonstrate the production of HALEU. The project involved construction of a cascade of 16 of the very large AC-100M centrifuges in Piketon, Ohio to produce some 19.75% enriched uranium by June 2022. It should also demonstrate the capability to produce HALEU with existing US-origin enrichment technology, providing the DOE with a "small quantity" of it for use in research and development "and other programmatic missions".

In June 2020 the NRC accepted an application from Centrus to enrich at up to 20% at Piketon and issued the licence around one year later. In November 2022, the DOE awarded Centrus $150 million of cost-shared funding to finish the cascade, complete final regulatory steps, begin operating the cascade, and produce up to 20 kg of HALEU, which Centrus delivered in November 2023.

It was then expected that production would continue in 2024 at a rate of 900 kg of HALEU per year. However, in February 2024 it was announced that delays in the supply of 5B cylinders from the DOE would prevent the delivery of this amount of material in 2024.

In June 2025 the DOE extended its HALEU production contract with Centrus for one year, until 30 June 2026. Later that month, Centrus announced it had met its production target of 900 kg of HALEU for the second phase of its contract with the DOE and would begin production under the third phase.

In December 2025 the DOE announced $11 million in awards to five US companies to develop and license new or modified transport packages for HALEU. The companies selected under Topic Area 1 for new package designs are NAC International, Westinghouse Electric Co., Container Technologies Industries, Centrus’ American Centrifuge Operating, and Paragon D&E. Under Topic Area 2 NAC International was selected to modify an existing package design for NRC certification.

In parallel, the DOE has made small quantities of HALEU available to selected US reactor developers through other channels such as downblending used fuel from research reactors and deconversion.

In 2025 the DOE estimated domestic HALEU demand could reach 50 tonnes per year by 2035.

Commercial HALEU production

Piketon, Ohio

In September 2025 Centrus announced plans for a multi-billion-dollar expansion of its American Centrifuge enrichment plant in Piketon, Ohio, contingent on federal funding, which could create 300 new operations jobs to boost production of both LEU and HALEU.

In January 2026 the DOE awarded Centrus a $900 million task order to produce HALEU at its American Centrifuge facility. Later that month Centrus announced it would invest $560 million over the next few years to expand its centrifuge factory at Oak Ridge. Centrus was also supported in this by Tennessee’s Nuclear Energy Fund. Centrus is targeting an initial HALEU capacity of 12 tonnes per year and first production from new cascades in 2029.

General Matter

General Matter has a lease agreement with the DOE to redevelop 100 acres (approximately 400,000 m²) of land at the Paducah, Kentucky, site which also affords it access to 7600 cylinders of UF₆ tails suitable for re-enrichment.

In January 2026 the DOE awarded General Matter’s a ten-year ‘Indefinite Delivery/Indefinite Quantity’ (IDIQ) contract to build and operate an enrichment facility for HALEU “for US commercial and government fuel needs” based on a series of milestones. General Matter has not revealed details of its plans or technology, saying only that it is “novel, scalable and cost-competitive”.

Defence enrichment needs

The National Nuclear Security Administration (NNSA) needs a reliable supply of unobligated LEU at 19.75% enrichment that can be used in research reactors as well as for defence purposes, notably tritium production and HEU for naval reactors. This must be sourced from a US-developed technology.

There is no such source at present, but in August 2015 the NNSA reported that it had enough unobligated LEU to get to 2038. By then it needed a plant of about 400,000 SWU/yr and favoured reviving the American Centrifuge technology in some form to provide this. Hence it is providing $32 million per year R&D at Oak Ridge to this end, with Centrus as contractor.

In January 2026 BWXT opened the Centrifuge Manufacturing Development Facility at Oak Ridge to support a future enrichment facility for NNSA needs.

Decommissioning of enrichment plants

The three original diffusion uranium enrichment plants – at Oak Ridge, Tennessee; Paducah, Kentucky; and Portsmouth, Ohio – were built by the federal government in the 1940s and 1950s for defence purposes and operated for 42, 60 and 47 years respectively.

From 1969 to 1992, the DOE and its predecessor agencies sold some of the plants’ enrichment services commercially under contracts that required utility customers to pay for future decommissioning and decontamination. Fees were collected for this but not set aside, and in 1992 the Energy Policy Act assessed utilities an annual fee of $150 million for 15 years, pursuant to which some $2.6 billion was paid. The administration’s 2014 budget proposed levying further tax of $2.4 billion over 10 years on nuclear utility customers, which would have been the third time a fee for the same programme was levied. In July 2013 the National Association of Regulatory Utility Commissioners contested this.

Decommissioning of the Oak Ridge diffusion plant was completed in October 2020, and the site was transformed into an industrial park.

Deconversion

Deconversion of the depleted uranium hexafluoride (UF₆) that remains as a byproduct after enrichment has not so far been undertaken on a large scale in the USA. For legal reasons depleted uranium is sometimes considered as a waste in the USA.

UDS plants at Portsmouth and Paducah

Uranium Disposition Services (UDS) was a joint venture of Areva’s fuel division (now Orano), EnergySolutions, and Burns & Roe. It was awarded a $558 million contract by the DOE in 2002 to design and build deconversion plants at Portsmouth, Ohio and Paducah, Kentucky. The contract ran to August 2010. The plants use a process developed by Areva (now Orano) which it employed at Richland, Washington and Lingen, Germany. The 13,500 t/yr Portsmouth plant was completed in mid-2010 and the 18,000 t/yr Paducah one in 2011. Operational testing and reviews were undertaken on both by EnergySolutions as managing partner. At Portsmouth about 250,000 t depleted UF6 is stored, with another 75,000 t shipped from Oak Ridge, and at Paducah some 440,000 t depleted UF6 is stored. Hydrofluoric acid was a commercial by-product.

In December 2010 Babcock & Wilcox Conversion Services won a five-year $428 million contract from the DOE for uranium deconversion operations at both Portsmouth and Paducah, using these UDS plants. Deconversion proceeded at about 20,000 tU/yr to 2016.

In September 2016 Mid-America Conversion Services, a joint venture of Atkins, Fluor and Westinghouse was awarded a $318 million, five-year contract to take over operation of the UDS plants from 2017. Products were depleted uranium oxide for reuse or disposal and hydrofluoric acid for industrial use.

Hobbs

In December 2009, International Isotopes (INIS) subsidiary Fluorine Products applied for a licence to build and operate a 6500 t/yr deconversion plant and fluorine extraction facility 35 km west of Hobbs, New Mexico, which is 70 km by road from the Urenco USA enrichment plant at Eunice.

In April 2010, INIS signed a five-year agreement to provide toll deconversion services for depleted uranium tails from Urenco from 2014. In October 2012 the NRC issued a licence to build and operate the $125 million plant, but in August 2013 the project was put on hold.

INIS had hoped to start production in 2014, subject to raising $75 million capital. Initial capacity would have been for about 300 type 48Y cylinders, containing 3600 tonnes of UF6 per year, increasing to about 575 cylinders (approximately 7000 t) in 2016. However in mid-2014 the company said that the project remained on hold until "additional uranium enrichment capacity comes online that can provide us with additional opportunities to contract for deconversion service of that material." Some 1300 to 2300 tonnes of anhydrous hydrofluoric acid with 450 tonnes of fluoride gas would be produced per year for sale by INIS, and the depleted uranium belonging to the enrichment companies would be stored as more stable U3O8. In-mid 2010, the first part of a $65 million loan guarantee application by INIS was approved by the US Department of Energy (DOE), and the company has been invited to submit the second part of its application².

Richland

Framatome operates a small deconversion plant in association with its fuel fabrication plant in Richland, Washington state.

Piketon

In March 2026, Centrus and Oklo announced they were discussing a joint venture to set up a deconversion facility alongside Centrus’ enrichment plant at Piketon which would meet the regulatory requirements for deconverting HALEU UF₆.

Uranium fuel fabrication

The USA has three main fuel fabrication facilities to convert enriched uranium oxide into solid pellets for fuel rods.

Westinghouse has a 1600 t/yr fuel fabrication plant for PWR and BWR fuel at Columbia, South Carolina.

Global Nuclear Fuel-Americas, owned by GE Vernova, Hitachi and Toshiba, has a 1000 t/yr plant for BWR fuel at Wilmington, North Carolina. It is licensed to produce fuels at up to 8% enrichment.

Framatome has a 1200 t/yr plant for PWR and BWR fuel at Richland, Washington. In 2009 this was the first to receive a 40-year licence extension from the NRC.

In September 2024 Framatome submitted a licence amendment for the Richland facility to begin producing TRISO fuel. A year later Framatome announced a partnership towards this goal with Standard Nuclear. In January 2026 Standard Nuclear announced investment of $140 million which would support construction of TRISO fuel fabrication infrastructure. Standard Nuclear said it would achieve production rates of 2 t per year “across multiple, strategically located sites” by mid-2026.

Decommissioning reactors

Several US nuclear power plant sites have been completely decommissioned and released by the NRC for unrestricted use.

The NRC requires nuclear licensees to make arrangements in advance to fund the decommissioning of their reactors. Usually, a trust fund is set up outside the licensee’s control to grow from direct deposits and by accumulating interest over time, the licence period of a US reactor being 40 years with 20-year extensions very common. Other possible options acceptable to the NRC can be insurance policies or guarantees from a parent company.

After a nuclear power plant is shut down and its fuel is removed from the reactor core, NRC licensees have two options to complete its decommissioning.

The major phase of decommissioning is known as ‘Decon’, which sees the operator promptly dismantle the plant components and equipment that contain the vast majority of radioactivity. With as much as 99% of the radioactivity hazard removed, the remaining work can take place more easily. Some operators postpone this for a period of ‘Safstor’ (safe storage), which places the plant in a state of protected storage for a number of years. During this time, radioactivity levels reduce while at the same time the decommissioning fund has extra time to grow.

Decon takes 5-10 years and a plant can be in Safstor for no longer than 50 years. The NRC mandates that the entire process must be completed within 60 years. Owners of single-unit plants generally opt for Decon, while Safstor is preferred for units at plants where other reactors are still in operation. Many operators combine the two approaches.

The NRC regularly reports on the adequacy of decommissioning funds set aside during operation into a trust fund or similar. In 2023 it confirmed that commercial reactor operators had sufficient funds for decommissioning their plants. The total amount accumulated by all licensees was $67 billion at the end of 2022.

Under US law, the DOE has been responsible for the management and disposal of highly radioactive used nuclear fuel since 1998 (see Used fuel management and disposal, below). However, the DOE has not yet prepared a disposal route. In the meantime, some decommissioned reactor sites still host small facilities that hold their used nuclear fuel in dry casks.

See the Nuclear Energy Institute’s Decommissioning Status of Shutdown US Nuclear Plants.

Private industry decommissioning

Funding provisions for decommissioning have been so successful that from around 2010 a new business model emerged in which a decommissioning specialist would buy a shutdown reactor for a nominal sum and transfer its licences and decommissioning fund. Its intention is to perform the decommissioning so efficiently that significant funds are left over.

NorthStar Group bought Vermont Yankee from Entergy in January 2019 and commenced decommissioning with support from Orano USA (reactor vessel segmentation and used fuel management support), Waste Control Specialists (waste management, packaging, transport and disposal) and Burns & McDonnell (engineering and regulatory support).

Holtec International bought Oyster Creek from Exelon in July 2019, Pilgrim from Entergy in August 2019, Indian Point from Entergy in May 2021 and Palisades from Entergy in May 2022. However, after beginning to decommission Palisades, Holtec has since begun working towards restarting it. See Nuclear reactor restarts in the information page on US Nuclear Power.

Used fuel management and disposal

Introduction

Four landmark decisions have created a disorganized landscape for storage of used nuclear fuel in the USA.

1977: A policy decision by President Jimmy Carter to “defer indefinitely the reprocessing of spent commercial nuclear fuel” based on global non-proliferation concerns. This solidified statements by President Gerald Ford the previous year that non-proliferation should take precedent over national and economic interests and argued for a three-year hold on commissioning the Barnwell reprocessing plant. Although no federal policy was ever codified, the combined effect of these statements led to the closure of facilities already built and under construction for reprocessing. It also resulted in all used fuel in the USA being considered as waste by federal agencies, regardless of the wishes of its private owners and its potential as a resource to manufacture further fuel.

1982: The Nuclear Waste Policy Act established federal responsibility for all civil used fuel, including a timetable and procedures for the building of two repositories. It obliged the federal government, through the DOE, to begin removing the used fuel from nuclear facilities by 1998 for disposal in a federal facility. This was to be paid for by a Nuclear Waste Fund built up from a small fee on utilities per kilowatt hour they generated.

1987: The Nuclear Waste Policy Act was amended in 1987 to designate Yucca Mountain in Nevada as the sole initial repository for 70,000 tonnes of high-level waste. This met strong and persistent opposition from Nevadan representatives.

2009: President Barack Obama’s energy secretary cut funding to the Yucca Mountain project and his appointed chairman of the NRC directed staff to stop working on its licence application (see Yucca Mountain and federal responsibility from 1987, below). This left the federal government with full responsibility for used fuel, but no route for management or disposal.

After the demise of the Yucca Mountain project, in 2011 a blue ribbon commission made a series of recommendations on consent-based siting of interim storage facilities and geologic disposal, but these have not been acted on.

By 2014 a Government Accountability Office report said that provisions in the Nuclear Waste Policy Act authorizing the DOE to establish consolidated storage facilities have either expired or are unusable because they are tied to progress with the aborted Yucca Mountain repository project. In the early 2020s court decisions also prevented the NRC from being able to license storage sites that are not at existing power plants or under DOE programmes.

Current status

As of 2026, there is a total of approximately 95,000 tonnes of used fuel from power generation in the USA. According to the IAEA, 43% of this is in pool storage and 56% is in dry cask storage. All of this is at current or former power plant sites, apart from a small amount at the never-used Barnwell reprocessing plant, which was retained as a pool storage facility.

Federal responsibility to pay for used fuel storage and eventual disposal remains.

Four executive orders were issued by the Trump administration in May 2025 intended to reconsider the country’s approach to the back end of the nuclear fuel cycle. Among other things they requested from administration departments:

• A recommended policy to support the management of used nuclear fuel and high-level radioactive waste and the development of an advanced fuel cycle.
• A review of statutory authorities and legislative changes necessary to achieve the above.
• An evaluation of reprocessing and recycling of used nuclear fuel, including that from power reactors, naval reactors and other facilities under the DOE.
• An analysis of legal, budgetary and policy considerations relevant to the idea of transferring used nuclear fuel from reactor sites to a government-owned by privately operated reprocessing facility.
• Recommendations for the disposal of wastes generated by reprocessing and recycling.
• To identify all useful uranium and plutonium material within DOE inventories that may be recycled or reprocessed for advanced reactors and to release at least 20 t of HALEU for qualifying private sector projects.

In response to these and the outstanding question of disposal of used fuel a number of reprocessing and recycling initiatives have emerged.

Argonne National Laboratory is the creator and custodian of a reprocessing (electrorefining) process that was used in the recycling and refabrication of metal fuels used in US fast reactors including EBR-II. In 2022 Argonne began a partnership with Oklo to commercialise the technology. As of 2026 Argonne had a cooperative research and development agreement (CRADA) to the same ends with BLSK Energy.

In September 2025 Oklo announced plans for a reprocessing (electrorefining) plant at Oak Ridge as well as a co-located fuel manufacturing plant which would make the metal HALEU fuel it requires for its sodium-cooled Aurora reactors. Oklo is in pre-application talks with NRC on these, as well as Aurora.

In October 2025 the DOE announced that it would supply plutonium to selected companies wishing to demonstrate reprocessing and recycling technologies and create fuel for advanced reactors. DOE said it had a stock of 20 tons of plutonium for this which it had identified as surplus. In May 2026 the DOE announced it would enter advanced negotiations with a shortlist of companies (Exodys Energy, SHINE Technologies, Standard Nuclear, Flibe Energy, and a partnership of Oklo and France-based Nucleo).

In June 2026 Curio sent a letter of intent to the NRC declaring its goal of building a reprocessing (pyro-oxidation) plant and requesting to begin pre-licensing discussions. Curio said it had a proprietary process that would reprocess a variety of used fuels and output uranium hexafluoride, transuranics, and other commercially valuable isotopes. The company is working to scale up the process after successful laboratory scale demonstration in 2025.

Recovering costs

US utilities have successfully sued the federal government for breach of contract to recover the costs they incurred after the government failed to take responsibility for the used fuel in 1998 as mandated by the Nuclear Waste Policy Act.

Settlements for this came from the taxpayer-funded Judgement Fund, and not the Nuclear Waste Fund, which remains ringfenced for final disposal only. By 2020 these settlements reached almost $9 billion, according to the Government Accountability Office, which estimated that liabilities of a further $30.8 billion remained.

Private companies have proposed interim storage areas for dry casks but none of these have gone ahead (see section on Commercial proposals for interim used fuel storage, in the Appendix).

Waste Confidence Rule

The DOE’s responsibility for used fuel disposal under the Nuclear Waste Policy Act of 1982 was formalized with each utility by standard contracts. These included the 0.1¢/kWh fee to the Nuclear Waste Fund and the date of 1998 by which DOE should remove used fuel from reactor sites, transport it to a repository and permanently dispose of it. These contracts dovetailed with NRC findings that (i) interim storage of used fuel is safe, (ii) a geologic repository is technically feasible, and (iii) a repository would be available when needed. Together this framework was known as the Waste Confidence Rule. It supported NRC to license new reactors without needing specific information on how used fuel would be stored and disposed of decades into the future.

Having missed its 1998 deadline, the DOE changed its standard contract for new reactors from 2000. The new text deferred the start of DOE’s responsibilities to ten years after the end of the reactor’s operating licence. In practice this would be 50 or 70 years into the future, obligating the utility to make full arrangements for interim storage and weakening its legal claims for compensation.

NRC reviewed and revised its findings in 1990 and 1999, and made a major update in 2010 which acknowledged that a repository could be significantly delayed. The revision affirmed that used fuel could be stored safely at reactor sites for at least 60 years after the end of operation, which enabled nuclear reactors to have their operating licences extended. A legal challenge to NRC’s assessment of the safety of interim storage over these time periods prompted NRC to undergird its findings with a generic environmental assessment which it incorporated into the Continued Storage Rule in 2014.

Post-Yucca Mountain policy vacuum

Following the Obama administration’s effective cancellation of the Yucca Mountain project from 2009 (see Yucca Mountain and federal responsibility from 1987 in the Appendix), the Blue Ribbon Commission on America's Nuclear Future was set up to recommend a new strategy. It was intended to "conduct a comprehensive review of policies for managing the back end of the nuclear fuel cycle, including all alternatives for the storage, processing, and disposal of civilian and defence used nuclear fuel and nuclear waste."

The blue ribbon commission’s report to Congress in January 2012 by recommended the development of centralized interim storage, establishing a new organization outside the DOE to manage the US used fuel programme, assured access to the Nuclear Waste Fund in a trust rather than via the Treasury, continued pursuit of a geological repository, and continued debate on reprocessing used fuel for recycle (strongly supported by the industry). More generally the report said: "This nation's failure to come to grips with the nuclear waste issue has already proved damaging and costly. It will be even more damaging and more costly the longer it continues." Congress expected the DOE to develop a new strategy for managing used nuclear fuel and other nuclear waste within six months in response to the report.

In 2012 Nye County, Nevada wrote to the secretary of energy agreeing to host the repository in line with the blue ribbon commission’s recommendations for consent-based siting of such facilities.

In January 2013 the DOE announced a new approach based on the report, including setting up a new organization to manage the siting, development and operation of the future waste stores. It envisaged a consent-based approach for setting up a consolidated interim storage facility (CISF), with a 'pilot interim store' being in operation in 2021, with a priority on taking used nuclear fuel from current shutdown power plant sites. By 2025 a larger 'full-scale interim store' was supposed to open, and by 2048 an underground disposal facility was hoped to be in place to permanently store and dispose of the material. The mandate for the new organization would exclude reprocessing of used fuel.

However, there has been no progress regarding this organization. Instead, the Trump administration in 2017 abandoned the draft consent-based approach for interim storage in favour of reviving the Yucca Mountain project, though nothing has happened to effect this either.

Nuclear Waste Fund

The Nuclear Waste Fund is an accounting mechanism in the federal budget that records the cash flows associated with the civilian nuclear waste programme. The fund primarily grew by collection of fees from utilities (0.1 ¢/kWh), with the balance then available for the DOE to appropriate for spending on civil used fuel disposal. Unspent balances have been available for the Treasury to invest in nonmarketable Treasury securities with the proceeds being added to the fund.

Following the cancellation of the Yucca Mountain project, utility contributions to the Nuclear Waste Fund were set to zero in 2014 after a Court of Appeals decision. "Because the Secretary [of Energy] is apparently unable to conduct a legally adequate fee assessment,” the court ruled, “the Secretary is ordered to submit to Congress a proposal to change the fee to zero until such a time as either the Secretary chooses to comply with the [Nuclear Waste Policy] Act as it is currently written, or until Congress enacts an alternative waste management plan."

US nuclear power utilities paid in a total of $24.6 billion (non-adjusted dollars) between 1983 and 2025. At the end of FY2025 the Nuclear Waste Fund stood at $51.4 billion, having grown by $1.9 billion that year from investment.

Military Radioactive Waste

In 1985 a decision was made to dispose of military high-level waste alongside used fuel from power plants and in 1987 the amendment to the Nuclear Waste Act specified that this would be at Yucca Mountain. Following his cancellation of this project, the President Barack Obama in 2015 authorized the DOE to plan a separate repository for military high-level radioactive waste.

The vast majority of used fuel from US nuclear power plants is fairly standardized in form – long fuel assemblies of zirconium tubing filled with ceramic fuel pellets that contain uranium oxide and the radioactive fission products which are the actual waste. By contrast, waste from military activities and research comes from a range of activities and processes and is therefore much less standardized in radioactive signature, heat production, packaging and options for disposal.

The DOE said in 2015 that some defence waste had decayed to the point that a repository could have a simpler design and present lesser challenges in transport and licensing. Much of DOE’s high-level waste has been or will be vitrified, which also allows for a simpler repository design.

The USA's only high-level radioactive waste vitrification plant has been operating at the Department of Energy's Savannah River Site since 1996. The Defense Waste Processing Facility converts waste-containing sludges into borosilicate glass, which is currently stored onsite in stainless steel canisters. By 2016 the facility had processed over 4000 canisters of vitrified waste since operations began, and it is expected to operate for a further 20 years.

The Waste Isolation Pilot Plant (WIPP) is the DOE’s deep geologic radioactive waste disposal facility for legacy transuranic waste from the US military programme. It is a series of tunnels some 660 metres underground in the Salado salt formation near Carlsbad, New Mexico. It has operated since 1999.

By 2013 WIPP held about 85,000 m3 of TRU waste. Its operation was interrupted in February 2014 by a leak of radioactivity from one of the waste packages after an exothermic chemical reaction in organic absorbent material. In response the DOE made improvements to infrastructure and ventilation to bring WIPP back into operation in January 2017.

WIPP was originally scheduled for permanent closure in 2030, but as of 2026 there is no published end date for its operation.

Low-level waste

The NRC classifies low-level radioactive waste (LLW) into three categories, Class A, B and C, based on increasing concentrations of long-lived radionuclides. Around 95% of all low-level waste is Class A. More stable packaging and deeper disposal facilities may be required for Class B and Class C waste.

There are a number of specialist US facilities which are licensed and regulated by the host states under terms of agreements with the NRC. Otherwise, LLW storage is at reactor sites.

Waste Control Specialists operates the Texas Compact Facility, which has taken some LLW since 1998 and was licensed in 2009 for Class A, B & C LLW, and also to take Vermont's and the DOE’s LLW. The company was licensed in 2012 to take LLW from other states, and in October 2011 it entered into an agreement with Utilities Service Alliance, representing 15 utilities including 17 nuclear power plants with 27 reactors. A separately licensed site in the complex handles radioactive waste from federal sites around the country, and in 2013 started receiving shipments from Los Alamos Laboratory in New Mexico. The two facilities in Andrews County are 580 km west of Dallas, near the New Mexico border, and owned by the state.

The Texas Compact Facility is the only one in the USA to be licensed within the last 30 years that is authorized to dispose of all classes of low-level radioactive waste materials considered suitable for near-surface burial. This includes depleted uranium, classified as Class A LLW. Since the radioactivity of this will increase over time, it will be encased in concrete and buried 30m deep. It accepts waste from the nuclear medicine, research, and power sectors. Some $500 million was invested in setting up the facility. In 2014 it gained approval to expand capacity from 65,000 m³ to 255,000 m³.

EnergySolutions has a facility at Clive, Utah, which accepts Class A LLW (about 90% of all LLW) from all over the USA (a proposal to take low-level decommissioning waste from Italy was abandoned in mid-2010). It has applied for a licence to accept depleted uranium, which is considered Class A LLW by the NRC, and envisages burying up to 700,000 tonnes of it about 3 metres deep. Depleted uranium would be disposed of with other LLW so that its natural increase in radioactivity over thousands of years is matched by the decrease in that of the other LLW, so that the disposal cells remain Class A forever.

EnergySolutions at Barnwell, South Carolina, for Class A, B & C LLW from that state, New Jersey and Connecticut, has operated since 1971, and also processes high activity filters from the nuclear industry.

EnergySolutions at Bear Creek, near Oak Ridge, Tennessee for processing and packaging of radioactive material for permanent disposal claims to be the largest licensed commercial US LLW facility with innovative technologies for radioactive material volume reduction (compaction, melting, incineration). There is nearby capacity for recycling depleted uranium.

American Ecology Corp at Richland, Washington accepts Class A, B & C waste from the Northwest and Rocky Mountain compacts.

Appendix

Cancelled enrichment facilities

Eagle Rock Enrichment Facility

In mid-2007 Areva’s fuel division (now Orano) proposed to build a 3.3 million SWU/yr $2 billion centrifuge plant in the USA to supply domestic enrichment services. It submitted a licence application to the NRC for the Eagle Rock Enrichment Facility in December 2008 and the NRC issued the licence in October 2011. Construction was to commence in 2012 at Idaho Falls, near the Department of Energy's Idaho National Laboratory, and operation was envisaged early in 2014, ramping up to full capacity in 2018.

Areva Enrichment Services, the owner and operator, signed a procurement & construction contract with URS (now AECOM) in February 2011. It would be similar to the Georges Besse II plant in France). In 2009, Areva notified a planned doubling in capacity to 6.6 million SWU/yr, with the first stage being 3.3 million SWU/yr. In May 2010, the DOE granted it a $2 billion loan guarantee. However, in December 2011 Areva announced that it was putting the project on hold as it sought an additional investor, and in May 2013 the projected timeline became indefinite. In March 2017 Areva Nuclear Materials asked the NRC to terminate the licence "since the facility will not be constructed."

USEC/Centrus American Centrifuge Plant

With the supply of LEU from downblended weapons, for which it had been the US agent, coming to an end, the United States Enrichment Corporation (USEC, now Centrus) sought to build an enrichment plant. USEC proposed to build The American Centrifuge at Piketon, Ohio using centrifuge technology it had developed from work carried out by the DOE at Oak Ridge through the 1970s and 1980s. A licence for construction and operation was granted by the NRC in April 2007.

The prototype lead cascade started operation in September 2007 in a programme to refine the design of the very large AC100 centrifuges, which were each expected to deliver 350 SWU per year.

An AC100 lead cascade started operation in March 2010 with plans to increase to 40-50 machines later in 2010. By the end of 2010, the company had spent $1.95 billion and required "additional capital beyond the $2 billion in Department of Energy loan guarantee programme funding that it has applied for and USEC's internally generated cash flow." It said that another $2.8 billion was required from "financial closing of a loan guarantee", despite the benefit of using existing infrastructure.

The full plant was expected to commence commercial operation in 2010, and achieve full 3.8 million SWU annual capacity at the end of 2012. However, by July 2009 the project was suspended due to the DOE refusing to award a $2 billion loan guarantee, and asking USEC to withdraw its application. USEC refused to do this, and in July 2010, it resubmitted its loan guarantee application to the DOE, pointing out that $200 million commitment by Toshiba and Babcock & Wilcox in May 2010 to support the ACP also strengthened its credentials³. Some of this was paid, but a stand-off then ensued pending a loan guarantee, which was never issued.

The schedule in February 2011 was: begin commercial operation in May 2014; reach 1 million SWU/yr capacity in August 2015; and achieve annual capacity of about 3.5 million SWU in September 2017. In March 2012 USEC said its immediate programme, dependent on government funding after May, was to “support building, installing and operating a 120-machine cascade and related support systems that will be replicated in 96 identical cascades in a full commercial plant.” This demonstration cascade was completed in April 2013 and became operational in October 2013. USEC then obtained funds from the DOE under a cooperative research, development and demonstration (RD&D) programme, 80% of the $10 million per month programme costs being funded by the DOE, and extended to September 2014.

In November 2013 USEC said: "Despite the technical progress being made by the RD&D programme, if funding is not in place at the end of the RD&D program or if USEC determines there is no longer a viable path to commercialization of ACP, we could make a decision to demobilize or terminate the project in the near term”. USEC said that its “prospects for adequate liquidity in 2014 are uncertain.”

With USEC unable to proceed with the project, the intellectual property and the demonstration cascade reverted to the DOE under the terms of the cooperative agreement. In March 2014 USEC applied for Chapter 11 reorganizational bankruptcy, a measure to enable it to restructure its debt while it continued operations. In its petition to the court, USEC listed assets of about $70 million and liabilities of $1.07 billion. The company emerged from bankruptcy at the end of September 2014 with the new name of Centrus Energy.

In April 2014 the DOE said that in the light of commercial deployment of ACP being not viable at that time, its Oak Ridge National Laboratory (ORNL) in Tennessee was taking over management of the project. Oak Ridge is where the ACP was originally developed in the 1980s.

The American Centrifuge Technology Demonstration and Operations Agreement between Centrus and ORNL ran to September 2015, backed by $97.2 million federal funding. It was then extended again for 12 months, with 60% reduced funding. Centrus said that this funding “excludes continued operations of America’s only operating cascade of advanced uranium enrichment centrifuges in Piketon,” which will therefore close down "after an investment of more than $2.6 billion in the project." In February 2016 Centrus announced that it had “completed operations of its demonstration cascade” of the ACP at Piketon but would maintain its NRC licence. In March 2016 Centrus signed the US Centrifuge Technology Advancement contract with UT-Battelle, as operator of ORNL. The contract ran for six months. In September 2016 it signed a further contract with UT-Battelle "for maintaining and advancing gas centrifuge uranium enrichment technology" over 12 months.

The DOE’s National Nuclear Security Administration (NNSA) is effectively now the keeper of US enrichment technology. The USA can only use uranium for national security (defence) purposes that has been enriched by US-origin technology, hence excluding Urenco and SILEX. It is not clear whether the USA needs any more HEU, but a related question is replenishment of tritium in nuclear warheads. This has been sourced from TVA’s Watts Bar 1 since 2003.

In its October 2018 Report to Congress, the NNSA said it plans to re-establish a domestic uranium enrichment capability to produce unobligated LEU (to use in nuclear reactors to produce tritium). The NNSA’s FY 2026 Congressional Justification published in May 2025 states that the Defense Fuels Program would ensure the supply of unobligated enriched uranium for defence purposes by the ‘tritium need date’ in the early 2040s by “incrementally deploying an enrichment capability using one or more gas centrifuge technologies, including the smaller scale Domestic Uranium Enrichment Centrifuge Experiment (DUECE) being developed at Oak Ridge National Laboratory (ORNL), and the AC100 large centrifuge.”

See also later subsection on Defence enrichment needs.

Russian proposal

The contract signed in March 2011 between Techsnabexport (Tenex) and USEC for the delivery of uranium enrichment services during 2013-2022 gave effect to a memorandum of January 2010 on the establishment of a joint venture in the USA to build a uranium enrichment plant based on Russian centrifuge technology. Tenex and USEC were undertaking a feasibility study on this, as a possible alternative to USEC American Centrifuge project. The main contract was for the supply of 21 million SWU to USEC over 2013-2022, worth $2.8 billion.

Military surplus and other government stocks

As noted above (USEC, Centrus and SWU supply from Russia) almost half of the uranium used in US nuclear power plants from 1994 to 2013 came from Russian weapons-grade military uranium, downblended in Russia. Under this programme, by the end of 2013, 500 tonnes of highly enriched uranium (HEU) had been downblended into 15,259 tonnes of low-enriched uranium (LEU, average 4.4% U-235) for reactor fuel, representing some million SWU supplied for about $8 billion (paid by electricity consumers). The programme recycled about 20,000 warheads. According to Tenex, its total revenue was $17 billion, including hard currency gains and the cost of natural uranium component as contra supply.

On the US side, 174 tonnes of military high-enriched uranium was initially declared to be surplus and available for civil power generation. Much or all of this has been downblended by Nuclear Fuel Services (NFS, a subsidiary of BWXT) in Tennessee, and the first fuel fabricated from it was shipped to Tennessee Valley Authority (TVA) power plants.

In 2005, the DOE's National Nuclear Security Administration (NNSA) announced that it was committing 40 tonnes of off-specification HEU in the USA to its Blended Low Enriched Uranium (BLEU) programme, with the fuel produced going to TVA power plants. To February 2012 about 22.7 t HEU had produced 301 t LEU (4.95% U-235), using natural uranium for blending.

In June 2007, the NNSA awarded contracts to Wesdyne International and Nuclear Fuel Services (NFS) to downblend 17.4 tonnes of HEU from dismantled warheads to yield about 290 tonnes of LEU, 230 t of which was for the American Assured Fuel Supply programme.  This LEU is available for use in civilian reactors by nations in good standing with the International Atomic Energy Agency that have good nonproliferation credentials and are not pursuing uranium enrichment and reprocessing technologies, and now also to US utilities. The fuel – worth some $500 million – would be sold at the prevailing market price.

In June 2009, NNSA awarded a further contract ($209 million) to NFS and Wesdyne to downblend 12.1 tonnes of HEU, into some 220 tonnes of LEU by 2012. This batch of LEU was to provide fuel supply assurance for utilities which participate in the DOE's mixed oxide fuel programme utilizing surplus plutonium from US weapons (see below). To cover the cost of the project, Wesdyne would sell a small part of the LEU (about 60 t) on the market over a three-to-four-year period. (The scheme is consistent with international concerns to limit the spread of enrichment technology to countries without well-established nuclear fuel cycles. Russia had agreed to join the initiative.)

In March 2008, the DOE announced a policy for dealing with uranium which was surplus to defence needs. The 2008 inventory, totalling nearly 72,000 tonnes of natural uranium equivalent, was diminished by 2013, when the DOE issued a uranium inventory management plan update:

* Natural uranium exchanged under the 1993 agreement whereby Russian blended-down uranium is supplied to US utilities – effectively Russian-origin stocks

The DOE 2008 plan showed a total of 22,700 tonnes of its uranium entering global markets before the end of FY2017, but with no more than 10% of US annual requirements being delivered to the market in any one year – apart from an allocation for the first cores of newly built US reactors.^c This 10% guideline was removed in the 2013 revision to the plan. Transactions over 2008 to 2013 included: 4300 tU to USEC in payment for the clean-up of the Portsmouth enrichment plant, downblending of much HEU, 9082 t DU as UF₆ to Energy Northwest, disposal of much unsaleable off-spec material, and an addition of 47.6 tonnes LEU to inventory in 2012 from enrichment of Russian natural U.

In 2014 the DOE confirmed that it would release 2705 tonnes of natural uranium equivalent annually to 2021, the same amount as in 2013, equivalent to 14% of US demand. It is mostly held as UF₆. In 2016 it released 2100 tU – 1600 tU as natural UF₆ and 500 tU as LEU from downblending by NNSA contractors, Wesdyne, which was ongoing. Much of this was to pay for Paducah and Portsmouth clean-up.

The stock of natural UF₆ could provide 1927 tU/yr for 7.5 years. After that, re-enrich the 73,500 tU in tails (av 0.375%) from the US stock of 464,000 tU in DUF₆ (including 295,000 tU at Paducah). The natural U value produced (@ $50/lb) is $3.4 billion, and it needs 14.7 million SWU to process and reduce tails to 0.20%.

The 114,000 tU of high-assay tails was DOE’s largest remaining uranium asset in 2016.

In April 2016 NNSA awarded a $241.5 million contract to NFS to downblend the last 10.4 tonnes of surplus HEU by 2019 to LEU, this being 1966 tonnes of natural uranium equivalent. In September 2018 TVA awarded NFS a $505 million contract to downblend 20.2 tonnes of HEU over 2019 to 2025 for LEU reactor fuel for Watts Bar 1 (approx. 4.4% U-235). The fuel is intended for use in the NNSA's Production of Tritium in a Commercial Light Water Reactor (CLWR) programme, where tritium is produced in TVA reactors using tritium-producing burnable absorber rods. TVA has been producing tritium for NNSA at Watts Bar 1 since 2004.

See also information page on Military Warheads as a Source of Nuclear Fuel.

Using enrichment tails

The US government owns some 700,000 tonnes of depleted uranium tails from past enrichment. In 2005 a pilot programme treated over 8000 tonnes of these.

In 2011 ConverDyn and Urenco USA formed a partnership to offer its services to DOE in utilizing more of these DU tails. The Competitive American Tails Upgrade Partnership (CATUP) will use the two companies' conversion and enrichment facilities in conjunction with its own natural uranium inventories to provide low-enriched uranium to the market while reducing the DU stockpile. CATUP's natural uranium concentrates would be converted to UF₆ at ConverDyn's plant and exchanged for depleted UF₆ from the DOE stockpile. This would then be cleaned up by CATUP, enriched by Urenco and sold.

Plutonium disposition

In addition to the HEU surplus, the US government has declared 61.5 tonnes of weapons-grade plutonium to be excess to the needs of the US defence programme. Of this, the government agreed under the 2000 US-Russia Plutonium Management and Disposition Agreement to dispose of 34 tonnes by 2014, incorporating it (with depleted uranium) into mixed-oxide (MOX) fuel. A 2010 protocol to the agreement confirmed MOX for light water reactors as the sole disposal option for the USA, while Russia was to dispose of the material as MOX in fast reactors.

Much investment was made to construct the Mixed-Oxide (MOX) Fuel Fabrication Facility (MFFF, see below) to fulfil the US side of this agreement, but in 2018 it was abandoned in favour of a cheaper dilution and disposal approach, which was slated to continue until 2050.

However, in March 2026 the expected shortfall of HALEU for advanced reactors (see above) prompted the DOE to announce plans to restart HB-Line within the H-Canyon chemical separation plant at the Savannah River site. Restarting it would take several years.

DOE said HB-Line could accelerate the disposal of remaining surplus plutonium by 10-13 years, while easing the HALEU shortage and saving taxpayer money. HB-Line could also recover valuable isotopes from the surplus plutonium at the same time.

H-Canyon dates from 1955 and originally recovered uranium, neptunium and plutonium from used military and research reactor HEU fuel. It also recovered neptunium-237 and plutonium-238 from special irradiated targets and played a vital role up to 2008 in the production of the Pu-238 used to power numerous deep space exploration programmes. Between 1998 and 2018 it recovered HEU from degraded materials and used fuel, to recycle it as LEU as part of the DOE's environmental management programme, before being laid up in 2020.

Construction of the Mixed-Oxide (MOX) fuel Fabrication Facility (MFFF) was authorized by the NRC in 2005 and preliminary work started in 2007. The full contract of $2.7 billion was awarded by NNSA to a consortium of CB&I, and Areva MOX Services (now Orano) in 2008.

The MFFF was based on Orano's Melox plant in France but with a more complex design due to the nature of purpose-made weapons-grade plutonium as compared to reactor-grade plutonium found as a by-product in used nuclear fuel. Most MOX plants use reactor-grade plutonium oxide powder separated after reprocessing which comprises about one-third non-fissile plutonium isotopes, whereas MFFF was to use weapons-grade plutonium with more than 90% fissile isotopes. MFFF also required facilities for plutonium pits dismantlement, metallic plutonium conversion to oxide and a waste solidification building.

Construction began in 2007 under an onerous and complex regulatory regime set up by the DOE, and it was slated to start operation in October 2016 and produce the first MOX fuel assemblies in 2018, for both PWR and BWR power plants.

The MFFF was intended to manufacture roughly 150 MOX fuel assemblies per year, consuming 3.5t of plutonium. It would therefore work through its allocated 34t in about 10 years and be available for longer should the US government decide to convert more surplus plutonium into MOX.

The US already had some experience of using MOX. It had been used in some US nuclear power plants before 1977 and weapons grade plutonium had been made into MOX which was used at the Saxton prototype reactor in the mid-1969s. In preparation for more widespread use of MOX from MFFF in US nuclear power plants, the NNSA signed an agreement with TVA in February 2010 to evaluate the use of this MOX in its Sequoyah and Browns Ferry power plants, moving up to possibly 40% core load. Duke Energy used four mixed oxide test fuel assemblies incorporating this weapons-grade plutonium (fabricated in France) at Catawba 1.

The complexities of MFFF’s design saw cost and schedule estimates escalate from $2.7 billion to $4.9 billion and then to $7.7 billion while operation slipped to 2019, which jeopardized its legitimacy as under the initiating agreement with Russia which specified 2014 as the startup date. For its part, Russia built a new 60 t/yr MOX plant at Zheleznogorsk and expanded an older plant at Dmitrovgrad to similar capacity, both by the 2014 deadline.

In 2014, when 60% complete, the MFFF was excluded from the DOE’s budget request due to cost escalation. Work continued, but with reduced funding and the NRC extended the construction licence to 2025 to allow for this. The US announced it was considering alternatives.

In April 2014 NNSA issued a report comparing alternative options for plutonium disposal. It said that the total cost of building and operating the Savannah River MOX plant would be over $31 billion, while burning the MOX in fast reactors would cost $58 billion, and immobilizing it in ceramic or glass and storing it at the Idaho National Laboratory would cost over $36 billion. Diluting it and storing it at WIPP in New Mexico would cost only $16 billion.

In early 2016 the DOE sought to terminate the project and dilute the plutonium instead so that it could be disposed at the Waste Isolation Pilot Plant (WIPP) in New Mexico. The dilution process carried out at Savannah River involves opening a canister of plutonium oxide, transferring it to a new much larger container, adding a dry adulterant so that plutonium is less than 10% of the total content by weight, sealing the can and then turning and rotating it to thoroughly mix its contents. NNSA said it expected the diluted material would be in about 113,000 drums of 200 litres each and could be shipped from Savannah River to WIPP from 2026 to 2050.

However, Areva reminded the DOE that, according to their Plutonium Management and Disposition Agreement and 2011 amending protocol, the USA and Russia must change their surplus plutonium at the isotopic level to render the nuclear material unattractive for weapons use. "Though considerations have been raised about pursuing other non-proliferation options, none of the other methods change the plutonium's isotopic structure, so would require renegotiations with the Russians," Areva said. Russia made it plain that the dilution option would amount to a violation of the 2011 protocol, and in October 2016 the Russian President unilaterally cancelled the agreement with some acrimony.

In May 2018 the Energy Secretary announced a decision to terminate the MOX project and instead pursue dilution and disposal. The Energy Secretary said: "The department's independent cost estimate concluded that the remaining dilute and dispose lifecycle cost is $19.9 billion. The department estimated the remaining lifecycle cost of the MOX fuel programme to be $49.4 billion."

In May 2018 CB&I-Areva consortium was acquired by McDermott. NNSA formally terminated the contract in October 2018, and NRC terminated the permit for the plant in February 2019. The MFFF facility has been repurposed for other NNSA work.

See also information page on Military Warheads as a Source of Nuclear Fuel and re plutonium, under Non-proliferation in the US Nuclear Power Policy paper.

Yucca Mountain and federal responsibility from 1987

In 1975 Nevada state legislature passed a resolution “strongly urging” DOE’s predecessor agency (the Energy Research and Development Administration) to choose the Nevada Test Site – where Yucca Mountain is located – for the disposal of nuclear waste.

In 1983 Yucca Mountain was one of nine potential sites for used fuel disposal identified by the Department of Energy and was the highest ranked of five nominated for further consideration. In 1986 it was the best of three remaining sites, based on its geohydrology, geochemistry, rock characteristics, tectonics, meteorology, costs and socioeconomic impacts.

In 1987, the Nuclear Waste Policy Act was amended to specify Yucca Mountain as the one and only disposal site for used fuel in the USA. From this moment, strong opposition to Yucca Mountain grew in Nevada, spearheaded by Senator Harry Reid and his advisor on the topic, Gregory Jaczko.

In 2007 the DOE estimated that Yucca Mountain could operate from 2021 with some expansion from its nominal 70,000 tonne capacity and a lifespan of about 110 years until decommissioning. As well as used reactor fuel Yucca Mountain was intended to take 2333 t of naval and DOE used fuel and 4667 t of other high-level waste, all from 126 sites in 39 US states. The cost was put at about $96 billion (in 2007 dollars).

Studies by the Electric Power Research Institute show that the repository could potentially hold at least 260,000 t and possibly 570,000 t of used fuel and high-level waste.

In June 2008 the DOE submitted a construction licence application to the Nuclear Regulatory Commission (NRC).

However, in February 2009 the new administration of President Barack Obama attempted to stop the projectⁱ by defunding it in DOE budgets. Obama also appointed Jaczko as chairman of the Nuclear Regulatory Commission in May 2009 and he unilaterally directed NRC staff to stop work on the licence application.

In 2010 the Atomic Safety and Licensing Board (an independent adjudicatory panel within the NRC) ruled that the DOE could not withdraw its licence application without the consent of Congress and therefore the NRC was still obligated by the Nuclear Waste Policy Act to review the application. Work continued.

The next blow to Yucca Mountain was Congress removing NRC funding for the review from FY2012 onwards.

An appeal to the DC Circuit Court of Appeals was made by the National Association of Regulatory Utility Commissioners, two states and others who argued that the NRC ignored its statutory responsibility when it terminated its review. In its ruling the court said the case "raises significant questions about the scope of the Executive's authority to disregard federal statutes. The case arises out of a longstanding dispute about nuclear waste storage at Yucca Mountain in Nevada. The underlying policy debate is not our concern. The policy is for Congress and the President to establish as they see fit in enacting statutes," and for the president and executive agencies to implement, it said. The court said that its task is to ensure that agencies such as the NRC comply with the law. The Nuclear Energy Institute and the Nuclear Waste Strategy Coalition said that this ruling on the used fuel repository licensing case was a clear signal that the NRC was obligated to complete its evaluation of the DOE application and issue a final decision granting or denying the licence.

In August 2013 the court directed the NRC to use leftover funds from previous years to finish its review of the safety evaluation report (SER) for the repository. The NRC complied and requested the DOE to prepare a supplementary environmental impact statement (EIS) for it. Volume 3 of the SER, a technical evaluation, was published in October 2014 and confirmed that Yucca Mountain would meet all regulatory requirements for a repository. Volumes 2, 4 and 5 of the SER were published by January 2015, completing the NRC’s technical safety review of the DOE licence application for the repository and confirming that its requirements were met. The DOE published the NRC’s draft supplementary EIS in November 2015 confirming the suitability of the site, and in May 2016 the NRC issued its final supplement to the EIS, analysing potential impacts on groundwater and surface groundwater discharges and determining that all impacts would be negligible.

Commercial proposals for interim used fuel storage

Private Fuel Storage (PFS) was a consortium of eight utilities impatient with the DOE and applied in 1997 for a licence to store used fuel on a site in Utah for up to 40 years pending disposal. In February 2006, the NRC issued a 20-year licence for a 40,000 tonne centralized surface dry storage facility on land owned by the Skull Valley band of the Goshute Indiansⁿ. But ongoing state government opposition led to the Department of Interior then disapproving the Goshute-PFS lease and the use of public land as a transport corridor to the planned facility. This decision was appealed in 2007, and in July 2010 the "arbitrary and capricious" 2006 Department of Interior ruling was overturned by a District Court ruling (which was not challenged). However the Department's Bureau of Indian Affairs maintained its opposition and in December 2012 PFS sought the termination of the 2006 licence due to the cost of maintaining it and the fact that all the PFS members now have dry storage at their plant sites. Construction time would have been 24 to 36 months, using Holtec’s Hi-Storm ventilated dry storage system. In 2015 PFS commenced moves to dissolve the entity.

In the light of the 2006 Utah setback, the Nuclear Energy Institute started a search for communities willing to host interim storage sites for used fuel. It received several offers and by mid-2008 had reduced the possibilities to two, and discussions were proceeding. A commercially-operated facility on a 400 ha site was envisaged for each location, with the DOE paying rent for casks stored there. The concept was picked up in a bipartisan 2013 Senate bill but did not proceed.

In response to the January 2013 DOE strategy calling for a consent-based approach for setting up a consolidated interim storage facility (CISF), two proposals came forward, in New Mexico and Texas. The NRC proposed a 42-month review period for each application.

The first consolidated interim storage facility (CISF) proposal was from some Texas counties which expressed interest. In January 2015 local government unanimously approved a plan by Waste Control Specialists (WCS), which operates the Texas Compact facility (see Low-level waste, below), to develop a CISF for used nuclear fuel at the low-level radioactive waste disposal facility the company has operated in Andrews County since 2012. The resolution references a March 2014 analysis by the Texas Commission on Environmental Quality that concluded consolidated storage of used fuel and high-level radioactive waste in Texas would result in “considerable savings” to electricity consumers compared with storage at each nuclear power plant. WCS applied to the NRC in April 2016 to build and operate an above-ground used fuel dry storage container facility at the 60 km² Andrews County site, on the New Mexico border and very close to the Urenco USA enrichment plant at Eunice, south of Hobbs. This would be initially for used nuclear fuel from Texas' nuclear facilities, and decommissioned reactors across the USA, and was expected to operate from 2020.

In May 2015 WCS signed an agreement with TN Americas (storage and transport division of Areva Nuclear Materials, now Orano USA) for the design, development, construction, operation and maintenance of the Texas CISF. Areva would also support the transport of used fuel to it. NAC International then joined Areva in supporting the proposal and seeking a licence for it. The CISF would provide above-the-surface storage for used fuel which has already been placed in dry storage casks at the nuclear plant site. The NUHOMS storage casks from Areva TN can accommodate high-burnup fuel. They would be encased in an additional NRC-certified transport cask and transferred primarily by rail to the facility, where they would be removed from the cask and placed in storage, in horizontal vaults – the NUHOMS horizontal storage module. The CISF would have 40,000 tonne capacity, developed in increments of 5000tonnes. The proposal included opportunities for 20-year licence renewals after the initial 40-year licence period.

However, in April 2017 WCS asked the NRC to suspend its review of the application pending its purchase by EnergySolutions. Its then owner, Valhi, was not supportive and wrote off costs incurred on the project. In June 2017 a US court barred the WCS sale on antitrust grounds. WCS said that the NRC’s $7.5 million cost estimate for reviewing the application was much higher than expected. In January 2018 WCS was acquired by an investment affiliate of private equity firm JF Lehman & Company. In March 2018 WCS and Orano USA announced a joint venture – Interim Storage Partners (ISP) – to complete licensing the project. In May 2020 the NRC issued a draft environmental impact statement for consolidated interim storage of up to 5000 tonnes of spent fuel for 40 years and a final EIS in July 2021. In September 2021 it issued an operating licence “to receive, possess, transfer and store up to 5000 tonnes of spent fuel.” Seven expansion phases over 20 years are envisaged to take capacity to 40,000 tonnes. In November 2020 the state governor had urged the NRC to deny the licence.

Orano's TN Americas and NAC between them represent 62% of existing dry storage systems in the USA, including 78% of the used nuclear fuel stored at sites where there is no longer an operating reactor. The licence request references dry storage casks designed and certified by Orano USA subsidiary Transnuclear (Orano TN) and by NAC International. ISP references over 460 NAC casks and 1265 Orano TN casks storing spent fuel at both operating and decommissioned US nuclear plants as of June 2019.

In September 2021, the NRC issued a construction and operation licence to ISP.

However, the state of Texas appealed to the Fifth Circuit Court of Appeals that under the Nuclear Waste Policy Act the NRC did not have authority to licence used fuel storage separate from a DOE programme, also pointing to the fact that the Act does not give the NRC specific authority to license away-from-reactor used fuel storage. The court agreed and vacated the licence. An NRC appeal was rejected.

The second consolidated interim storage facility (CISF) proposal was from the Eddy-Lea Energy Alliance (ELEA), a consortium of Eddy and Lea counties and Carlsbad and Hobbs cities, which came forward with a proposal supported initially by the state governor. ELEA has a 4 km² site between Carlsbad and Hobbs in the southeast of New Mexico, about 15 km north of the Waste Isolation Pilot Plant (WIPP). Holtec signed an agreement with the Eddy-Lea Energy Alliance to establish a Hi-Store CISF there, requiring less than 25 ha. The project included the design, licensing, construction and operation of an interim storage facility using an enlarged version of Holtec’s Hi-Storm UMAX storage system, already deployed at two US nuclear power plant sites. This stores used fuel in sealed canisters inside ventilated vertical steel and concrete containers 5 m high and below ground, with 11-tonne lids. The containers are set up in a 7.7 m deep excavation and low-strength concrete grout is backfilled around them. Canisters can be retrieved in a few hours at any time. (Holtec said that only 13 ha would be required to store 75,000 tonnes of used fuel, as envisaged for the suspended Yucca Mountain repository.)

In March 2017 Holtec applied to the NRC for a licence for phase 1 of the Hi-Store CISF, which would eventually cover 120 ha and hold 10,000 canisters of spent fuel for 40 years. Phase 1 would be for 500 canisters, with 8680 tonnes of spent fuel and was hoped to be operational in 2020. Nineteen subsequent phases would be for the rest of the capacity. The CISF would use Holtec’s already-certified Hi-Storm UMAX system “designed to accept every canister currently loaded at every US nuclear power plant.” Holtec in August 2016 had applied for NRC licensing for using Orano’s NUHOMS canisters in the Hi-Storm UMAX facility.

In May 2023 the NRC issued a licence to build and operate the facility. However, the project was immediately blocked by the entry into force of a bill sponsored by the state governor barring the storage and disposal of high-level radioactive waste without explicit consent.

In addition, New Mexico (together with some other parties) also appealed to the Tenth Circuit Court of Appeals that under the Nuclear Waste Policy Act the NRC did not have authority to license used fuel storage separate from a DOE programme, also pointing to the fact that the Act does not give the NRC specific authority to license away-from-reactor used fuel storage. Following the precedent set in Texas by the Fifth Court, the licence was vacated. Holtec wound up the project in October 2025.

Related information

USA: Nuclear Power
USA: Nuclear Power Policy
International Framework for Nuclear Energy Cooperation
Military Warheads as a Source of Nuclear Fuel