In 2019, Russia and China signed a contract for the supply of HEU fuel for the first seven years of operation of the CFR-600 fast neutron reactor. The reactor, located at Xiapu, started operations in December 2023. The second reactor, which is under construction at the same site, is expected to begin operations in 2026.

Fast neutron reactors can produce high-quality weapon-grade plutonium in the uranium blanket surrounding the active zone. China has maintained that this material will be used exclusively in its peaceful program. And indeed, the terms of the fuel supply agreement appear to restrict the use of the plutonium produced when the reactor uses the HEU fuel supplied by Russia.

Two key documents that regulate the issue are the Russian Regulations on the Export and Import of Nuclear Materials, first issued by the Russian government in 2000 and last amended in February 2023, and the Agreement on Cooperation in the Area of Peaceful Use of Nuclear Energy, which Russia and China signed in 1996.

According to paragraph 6 of the Regulations, export of materials like HEU to nuclear weapon states is allowed only if

exported items and items produced based on them will not be used for the production of nuclear weapons and other nuclear explosive devices or for any military purposes.

The Regulations requires the recipient state to provide written guarantees of peaceful use, which is then assessed by Rosatom (paragraph 8).

Article 4 of the Agreement on Cooperation contains similar language:

The parties assume that the items of nuclear export transferred under this Agreement will not be used for the production of nuclear weapons and other nuclear explosive devices or for any military purposes.

These two documents clearly prohibit the use "for any military purpose" of the plutonium that would be produced by any reactor that uses HEU fuel supplied by Russia.

This post contains a summary of INFCIRC/549 reports by the countries that submit annual civilian plutonium declarations that reflect the status of civilian plutonium stocks as of 31 December 2022. The total amount of plutonium declared as civilian was about 370 tonnes, an increase of about 7 tonnes since the end of 2021. Only about 140 tons of this material are under international (IAEA or Euratom) safeguards. The other 230 tonnes are not safeguarded, but are covered by various obligations not to use the material for military purposes.

Japan (INFCIRC/549/Add.1-26) reported owning the total of 45.1 tons of plutonium, 9.2 tons of which is in Japan (the numbers in 2021 were 45.8 tons and 9.3 tons respectively). According to the Status Report on Plutonium Management in Japan - 2022 released in July 2023, out of the 35.9 tons of plutonium abroad, 21.757 tons are in the United Kingdom and 14.113 tons are in France.

Germany (INFCIRC/549/Add.2-26) reported having no separated plutonium in the country for the third year in a row. Germany does not report separated plutonium outside of the country. It is believed to be less than 1 ton.

Belgium (INFCIRC/549/Add.3-22) declared no separated plutonium in storage or at reprocessing plants and "not zero, but less than 50 kg" of separated plutonium in other categories. It reported that it had no foreign plutonium as of 31 December 2022.

Switzerland (INFCIRC/549/Add.4-27) reported having less than 2 kg of plutonium in the country (in the "located elsewhere" category). The number has not changed since 2016 (it was "less than 50 kg" in 2015).

France (INFCIRC/549/Add.5-27) reported having 106.2 tons of separated unirradiated plutonium in its custody. Of this amount, 14.33 tons belongs to foreign countries. It appears that almost all that plutonium - 14,113 kg - belongs to Japan. The amount of plutonium owned by France is 91.87 tons, an increase of 6.97 tons from previous year (84.9 tons).

The United States in its 2022 report (INFCIRC/549/Add.6-25) declared 49.2 tons of separated plutonium, of which 4.6 tons are in MOX fuel and 44.6 tons are "held elsewhere" (most of this material is believed to be in weapon components). The total amount was reported to be 49.4 tons in 2021. It is possible that the change reflects the process of disposal of some material. The amount described as "disposed as waste" was 4.5 tons in 2021 and 4.7 tons in 2022.

China has not has not submitted its 2017-2022 reports as of 25 March 2024. The last INFCIRC/549 report submitted to the IAEA showed 40.9 kg of separated plutonium as of 31 December 2016.

The United Kingdom (INFCIRC/549/Add.8-26) reported owning 116.4 tons of separated plutonium, a decrease from 116.5 in 2021. In addition to that, the United Kingdom stores 24.1 tons of foreign plutonium (of which 21.757 tons is owned by Japan).

Russia (INFCIRC/549/Add.9-25) reported owning 64.5 tons of civilian plutonium, an increase of 1 ton from 2021.

In addition to reporting plutonium stocks, some countries also submit data on their civilian HEU:

Germany reported 0.35 tonnes of HEU in research reactor fuel, 0.94 tonnes of HEU in irradiated research reactor fuel, and 0.01 tonnes in the category "HEU held elsewhere." None of the numbers have changed since 2020.

France declared 5312 kg of HEU (5313 kg in 2021), of which 3761 kg (3760 kg) is unirradiated material - 506 kg (804 kg) of HEU at fuel fabrication or reprocessing plants, 78 kg (60 kg) at civil reactor sites, 3177 kg (2896 kg) at various research facilities. Also declared are 1551 kg (1533 kg) of irradiated HEU - 40 kg (62 kg) at civil reactor sites and 1511 kg (1491 kg) in other locations.

The United Kingdom reported having 691 kg of HEU (734 kg in 2021). Of this amount, 554 kg is unirradiated HEU (598 in 2021): less than 1 kg of unirradiated HEU is stored at the enrichment plants, less than 1 kg is at civil reactor sites, 440 kg - at fuel fabrication facilities, and 114 kg - at other sites (420 kg and 178 kg respectively in 2020). Irradiated HEU is located at civil reactor sites (5 kg) and other sites (131 kg).

During a visit to the Civaux nuclear power plant on 18 March 2024, France's Minister of the Armed Forces unveiled a plan to use the plant to produce tritium for the French nuclear weapons program. Civaux is a civilian power plant that belongs to and is operated by Electricité de France. According to the report, the nuclear regulator, l'Autorité de Sûreté Nucléaire, is expected to issue an approval in September 2024. The first test assemblies will be loaded in the reactor during a scheduled refueling in 2025.

Until now, France has been producing tritium in reactors of the CEA (Commissariat à l'Energie Atomique, Atomic Energy Commission). The practice of producing tritium in power reactors has been used by the United States, where lithium targets are irradiated in the reactors of the Tennessee Valley Authority. It should be noted that according to the US policy, reactors that produce tritium cannot use uranium enriched at civilian facilities. Uranium for the reactors involved in tritium production is obtained by down-blending excess military HEU.

This is not the first time France has used a civilian nuclear reactor for the weapons program. France used its Phénix breeder reactor for producing plutonium for the nuclear-weapon program (see "Fast Breeder Reactor Programs: History and Status", p. 25)

On 4 March 2024, India announced the "commencement of core loading" of the Prototype Fast Breeder Reactor (PFBR).

The construction of the reactor began in 2004 and its operation has been delayed numerous times. In March 2023, the Department of Atomic Energy announced that the reactor will begin operations in 2024. Although core loading marks a significant step towards this goal, the date when the reactor is expected to achieve criticality has not yet been announced.

Hui Zhang

In 2023, the China National Nuclear Corporation (CNNC) completed and may have started operating two large new centrifuge enrichment plants (CEP).

CNNC is operating three large centrifuge enrichment facilities that produce LEU for civilian purposes: Lanzhou (Gansu province, Plant 504), Hanzhong (Shaanxi province, Plant 405), and Emeishan (Sichuan province, the Emeishan civilian facility of Plant 814) (for more detail see: Hui Zhang, "China's uranium enrichment and plutonium recycling 2020-2040: current practices and projected capacities," in China's Civil Nuclear Sector: Plowshares to Swords?). In 2023, CNNC added two new centrifuge enrichment plants (CEP) to these facilities: Emeishan CEP3 (i.e. Project 3 at Emeishan) and Lanzhou CEP5.

Emeishan CEP3

Satellite imagery taken in February 2015 shows early construction activities at Emeishan CEP3, including preparation for pad construction. The images also reveal that construction was suspended from at least February 2016. However, according to local government documents, construction resumed and continued from 2019 through 2022. Satellite images show that major construction work was completed around November 2021. According to Chinese sources, Emeishan CEP3 started trial production as early as January 2023, and was expected to begin normal operation around the end of 2023. Unlike the other two facilities at Emeishan, CEP1 and CEP2, which are equipped with first-generation centrifuges, CEP3 is believed to be using second-generation centrifuges. These have been operating at the Hanzhong plant 405 since 2017.

20231215-Emeishan.png Emeishan CEP3 is believed to have a capacity of about 1.5-2 million SWU/year. Therefore, when combined with Emeishan CEP1 and Emeishan CEP2, which have a combined capacity of 2.2 million SWU/year, the total estimated capacity of the Emeishan enrichment facility has risen to approximately 4 million SWU/year. The image above, taken in May 2022, shows the layout of the facility (29.677314, 103.534625, Credit: CNES/Airbus, Google Earth).

Lanzhou CEP5

Construction activity at the main building of the Project 5 plant is visible in satellite images taken in late 2014. Other images also indicate that construction appeared to have been suspended in late 2015, but it resumed in the second half of 2022. The construction was apparently completed in early 2023, and according to Chinese sources, the Project 5 plant began initial operations around fall 2023. It is expected to start normal operations as early as the end of 2023.

20231215-Lanzhou.png As with Emeishan CEP3, Lanzhou CEP5 uses second-generation centrifuges and is estimated to have the total estimated capacity of 1.5-2 million SWU/year. Since the capacity of the other four plants at Lanzhou is about 2.6 million SWU/year, operations of Project 5 plant will bring the total capacity of Lanzhou plant to about 4.4 million SWU/year. The image above, taken in November 2022, shows the layout of the Lanzhou facility (36.148139, 103.523472, Credit: Maxar, Google Earth).

The table below describes individual plants at Emeishan and Lanzhou.

China's enrichment capacity

Taking into account the Hanzhong plant 405, which has a capacity of about 2.7 million SWU/year, China's total enrichment capacity is about 11 million SWU/year. This would be sufficient to provide enrichment services for the reactors with the total installed capacity of 70 GWe, which, according to the 14th Five-Year Plan (2021-2025), China is set up to achieve by 2025. Most of these reactors are light-water reactors, which are assumed to require about 130 tonne-SWU/year/GWe.

If China plans to bring its installed nuclear capacity to about 100 GWe by 2030, this will require an enrichment capacity of about 13 million SWU/year. It is anticipated that China will install about 2 million SWU/year or more during next Five-Year Plan (2026-2030). This expansion means that China will significantly increase its enrichment capacity. However, currently, it will primarily focus on servicing domestic demand.

Hui Zhang

20231215-CFR-600.jpgAccording to Chinese sources, China started up its first CFR-600 breeder reactor, running it at low power as of mid-2023. As of October 2023, the reactor had not yet been connected to the grid. The exact timeline for when it might begin generating electricity is unknown.

The reactor is the first of the two breeder reactors of this type built at Xiapu, Fujian province. The image above shows the reactor site (26.803567° 120.154710°) as of 23 December 2022.

Construction of the first unit began in 2017. At the time it was expected to begin operations in 2023, so the project is progressing according to the original plan. The initial core of the reactor is loaded with HEU fuel supplied by Russia. The first batch of fuel was delivered in 2022.

Construction of a second CFR-600 reactor at the same site started in 2020. This unit is expected to become operational in 2026.

Tatsujiro Suzuki

According to a Reuters report of 19 October 2023, Japan Nuclear Fuel Ltd (JNFL) still hopes to finish construction of Japan's long-delayed Rokkasho reprocessing plant in the first half of FY 2024 (i.e. during April-September 2024). Already more than 25 years behind schedule, there are reasons to believe that this new announcement is just another wishful plan that will end with another postponement.

One indication of further possible delays is that on September 28, 2023, Naohiro Masuda, president of JNFL, stated that the safety review of the reprocessing plant by Japan's Nuclear Regulation Authority will be difficult to complete by the end of the year 2023. He nevertheless insisted that the company could still meet completion target set as "by the end of the first half of FY 2024" (September 2024).

Here is a partial history of past key developments:

  • 1993 - Construction started,
  • 1997 - Initial target for completion,
  • 2006-2008 Hot tests conducted, revealing technical problems with the vitrification process,
  • 2011 - Fukushima Dai-ichi nuclear plant accident,
  • 2012 - New safety regulation standards introduced,
  • 2022 - Completion target date postponed to early in the first half of FY2024 (by June 2024) (26th postponement).

Why so many postponements?

There seem to be several underlying reasons for the postponements. First, JNFL lacks relevant expertise to manage such a technologically complex and hazardous project. It is owned by nine nuclear utilities plus all other major companies associated with nuclear power in Japan. Most of its senior executives are from shareholding companies (especially utility companies) and are not necessarily experts in the field of reprocessing.

Second, the technologies in the plant came from different companies and institutions. The management of the project is therefore technically complex.

Third, the post-Fukushima-accident nuclear facility safety licensing review process is much more stringent than before. For example, the Nuclear Regulation Authority told JNFL at their November 25, 2023 meeting:

JNFL should immediately make improvements because it is clear that JNFL does not understand the contents of the permit well enough to confirm the adequacy of the design of the facilities on site and has not visited the site.

Fourth, the financial costs to JNFL of postponement are covered by the utilities' customers because the utilities must pay a "reprocessing fee" every year, based on the spent fuel generated during that year, whether or not the reprocessing plant operates. The system by which the Nuclear Reprocessing Organization of Japan decides the reprocessing fee, is not transparent.

Fifth, the project lacks independent oversight. Even though JNFL's estimate of the cost of building and operating the Rokkasho plant has increased several-fold, no independent analysis has been done by a third party. One reason is that some of the shareholders are contractors themselves and have no incentive to scrutinize the reasons for the cost increases or the indefinite extension of the construction project.

After so many postponements, there is reason to wonder whether the plant will ever operate, but the government and utilities continue to insist that the plant will operate soon.

Even if Rokkasho were to operate, it may suffer from the same kinds of problems that marked Britain's light-water reactor spent fuel reprocessing experience, see the IPFM report Endless Trouble: Britain's Thermal Oxide Reprocessing Plant (THORP) by Martin Forwood, Gordon MacKerron and William Walker.

Why does Japan's commitment to reprocessing continue?

There are four reasons:

Spent fuel management. Currently, most of Japan's spent nuclear fuel is stored in nuclear power plant cooling pools. But the pool capacities are limited and the 3000-ton-capacity Rokkasho spent fuel pool is also almost full. The nuclear utilities must therefore start operating the Rokkasho plant unless they can create additional spent fuel storage capacity, either on- or off-site. The Mutsu spent fuel storage facility is a candidate, but due to the concern that spent fuel could stay there forever, Mutsu city refuses to accept spent fuel unless the Rokkasho reprocessing plant begins to operate. The Rokkasho plant design capacity is 800 tons of spent fuel per year.

Legal and institutional commitments. Under Japan's nuclear regulations, utilities must specify a "final disposal method" for spent fuel. The law on regulation of nuclear materials and nuclear reactors states that "when applying for reactor licensing, operators must specify the final disposal method of spent fuel" (Article 23.2.8). In addition, there was a clause that "disposal method" should be consistent with implementation of the government policy, which specified reprocessing as the disposal method. Although that clause was deleted in the 2012 revision of the law after the Fukushima accident, the Law on Final Disposal of High-Level Radioactive Waste still bans direct disposal of spent fuel. In addition, the 2016 Law on Reprocessing Fees legally requires utilities to submit reprocessing fees for all spent fuel generated every year since they stated in their applications that "final disposal method" for their spent fuel would be reprocessing.

Commitments to hosting communities. The nuclear utilities committed - albeit tacitly - to the communities hosting nuclear power plants that they would remove the spent fuel to reprocessing plants, since that was the national policy. Separately, JNFL signed an agreement with Rokkasho village and Aomori prefecture that says that if the Rokkasho reprocessing plant faces "severe difficulties," measures will be considered including the return of spent fuel stored at Rokkasho to the nuclear power plants.

Local governments hosting nuclear power plants were not involved in this deal, however. They could therefore just refuse to receive spent fuel from Aomori.

In fact, after the Fukushima accident, when the government was considering amending the nuclear fuel cycle policy to include a "direct disposal option" for spent fuel in a deep underground repository, the Rokkasho village parliament (at the behind the scenes suggestion by the then JNFL president, Yoshihiko Kawai), issued a strong statement asking for "maintenance of the current nuclear fuel cycle policy."

The statement continued that, if Japan's fuel cycle policy changed, Rokkasho would: i) refuse to accept further waste from the reprocessing of Japan's spent fuel in the UK and France, ii) require the removal of reprocessing waste and spent fuel stored in Rokkasho, iii) no longer accept spent fuel, and should receive compensation for the damages caused by the change of the policy. This is an example how "local opposition" is often raised by the utilities and the government as a barrier to changing the reprocessing policy.

Institutional and bureaucratic inertia. Bureaucrats, who, in Japan, rotate to new positions every two or three years, are reluctant to take the risk of changing existing policies. They therefore tend to stick with past commitments. Institutional inertia becomes stronger as a project becomes bigger. The Rokkasho reprocessing project is one of the largest projects ever in Japan. Changing the project is therefore very difficult.

Will Japan's "plutonium capping policy" have any real impact?

In 2018, Japan's Atomic Energy Commission announced a new policy on "Basic Principles on Utilization of Plutonium" (see also the IPFM post). In the new policy, JAEC proposed that: 1) Japan would reduce its stockpile of separated plutonium starting with a commitment not to increase it, 2) Reprocessing would take place only when a credible plan to use the separated plutonium existed.

This policy, in conjunction with the new Reprocessing Fee Law, gives the government legal authority to control the pace of reprocessing. However, it is not clear how the "capping policy" will be implemented. It is not a legally binding document, and no regulation has been introduced to control reprocessing. Utilities must submit specific plans for plutonium use to the JAEC for its review before reprocessing of their fuel begins. But the JAEC can only give advice to the government about the credibility of these plans. There is therefore reason for concern that this policy may not be sustained. (A similar "paper rule" has existed since August 2003).

A way out

A way out of this situation would be:

Find additional spent fuel storage capacity, either on- or off-site. Local communities may be more willing to accept on-site dry cask storage of spent fuel, if they are told that it is safer than spent fuel pool storage. For example, Saga Prefecture and Genkai-town, which host Kyushu Electric's Genkai Nuclear Power Plant have agreed to host a dry cask storage that is to start accepting spent fuel in FY 2027. Host communities may want guarantees that spent fuel will be removed after a specified storage period. Such a guarantee could be given by the central government.

Amend the law on final disposal of high-level radioactive waste to allow direct disposal of the spent fuel in a deep underground repository. This would provide more flexibility in spent fuel management and make it easier for local communities to host interim spent fuel storage.

Amend the Reprocessing Fee Law to allow Reprocessing Fund to be used by the government to implement shutdown of the Rokkasho reprocessing plant including payment for JNFL's financial debt and for dry cask interim storage. This would enable the government to end the Rokkasho reprocessing plant project.

Additional resources

Tadahiro Katsuta and Tatsujiro Suzuki, Japan's Spent Fuel and Plutonium Management Challenges, Report of the International Panel on Fissile Material, September 2006.

Masafumi Takubo and Frank von Hippel, Ending reprocessing in Japan: An alternative approach to managing Japan's spent nuclear fuel and separated plutonium, Report of the International Panel on Fissile Material, November 2013.

Plutonium Separation in Nuclear Power Programs. Status, Problems, and Prospects of Civilian Reprocessing Around the World, Report of the International Panel on Fissile Material, July 2015.

Frank N. von Hippel and Masafumi Takubo, Banning Plutonium Separation, Report of the International Panel on Fissile Material, July 2022.

The US company Centrus began operating a centrifuge cascade that will produce high-assay low enriched uranium (HALEU). The cascade, comprised of 16 centrifuges, is deployed at the American Centrifuge Plant in Piketon, Ohio. It is expected to produce about 20 kg of HALEU by the end of the year. The capacity of the current cascade is estimated to be sufficient to produce about 900 kg of HALEU a year. Centrus plans to build a cascade of 120 centrifuges that will be producing about 6000 kg of HALEU a year. This plan, however, depends on the availability of funding. Operations of the current cascade were supported by the Department of Energy.

The government of Japan released The Status Report of Plutonium Management in Japan - 2022, which details its plutonium holdings. According to the report,

As of the end of 2022, the total amount of separated plutonium both managed within and outside of Japan was approximately 45.1 tons, approximately 9.3 tons of which was held domestically and the rest of approximately 35.9 tons was held abroad.

The amount of domestic storage, 9.3 tons, hasn't changed since 2021. While about 0.6 tons were loaded into the Takahama Unit 4 reactor, approximately the same amount of plutonium fabricated into MOX fuel was transferred from France.

Of the plutonium stored abroad, 14,113 kg are stored in France (14,760 in 2021) and 21,757 kg - in the United Kingdom (21,780 in 2021). The reprocessing of Japan's spent fuel held in France had been completed by the end of 2017.

In 2021, Japan reported having a total of 45.8 tons of separated plutonium, of which 9.3 tons were held domestically.

URENCO USA (UUSA) is planning to deploy additional centrifuge cascades at its facility in the United States. According to the company, the addition

will provide an additional capacity of around 700 tonnes of SWU per year, a 15 per cent increase at UUSA, with the first new cascades online in 2025.

UUSA operates an enrichment facility in Eunice, New Mexico since 2010. The plant is licensed to produce LEU with enrichment of up to 5% (although UUSA has been exploring an option of producing HALEU with enrichment of up to 20%). The current capacity of the Eunice plant is reported to be 4900 tSWU/year. The original plan called for the capacity of 5700 tSWU/year.

URENCO is also considering expanding the capacity of the enrichment facility in Gronau, Germany.