Japan's Nuclear Restart Math Gets Harder as Rokkasho Cost Hits ¥15.98 Trillion
Reprocessing cost overruns and a 10.4% capacity-utilisation drop at KEPCO undercut the energy-security case for restarting Japan's reactors.
Japan Nuclear Fuel Limited's Rokkasho spent-fuel reprocessing plant will now cost ¥15.98 trillion to complete, up ¥360 billion from the previous year's estimate and the highest figure yet recorded for the project3. Construction costs alone climbed ¥180 billion to ¥3.92 trillion, while the co-located MOX fuel plant rose ¥80 billion to ¥2.68 trillion3. The numbers landed in a Japan NRG Weekly published on 2026-06-29, but they matter more now because the government has tied the stable and continuous use of nuclear power to securing roughly 6,000 tonnes of spent-fuel storage capacity by around 20307.
The arithmetic is uncomfortable for anyone who has framed Japan's restart programme as a clean energy-security fix. Reprocessing recovers 95-97% reusable uranium and plutonium while shrinking the volume and hazard of high-level waste, but the upfront bill keeps climbing3. Every yen added at Rokkasho is a yen not spent on the 14 reactors Japan plans to rebuild by 2050, a programme that would add 16 GW of generation capacity if all units are completed2.
Those additions are not trivial in scale. Japan currently generates between 60% and 70% of its electricity from imported hydrocarbons, which is the core of its energy-security problem2. The country already operates 60 nuclear reactors with 125 million kW of installed capacity, and China has overtaken the United States as the world's largest nuclear nation after building 34 GW over ten years2. Japan's planned 16 GW would be a meaningful but partial offset against its hydrocarbon dependence.
Operating performance is the more immediate problem. KEPCO guided to a FY2026 nuclear capacity-utilisation rate of 70.5%, a 10.4% year-on-year drop, citing extended regular inspections at Takahama as Units 1 and 2 prepare for internal structure replacement and Units 3 and 4 undergo steam generator replacement4. That guidance, reported in Japan NRG Weekly on 2026-04-20, is not a fuel-supply story. It is a maintenance and outage story, and it points to the operational drag that comes with ageing fleets.
Fuel-cycle economics are not the only constraint. The DOE's National Nuclear Security Administration recently described Japan's transfer of 1.7 tonnes of HALEU as the largest single international shipment of uranium to the US in the agency's history1. Leaving aside the diplomatic symbolism, the shipment underscores how tightly nuclear fuel logistics are now interwoven with alliance politics. Centrus is one of three US companies supported through a $2.7 billion DOE programme to expand domestic enrichment capacity6.
The uranium market has noticed the broader revival. Uranium ETF URA closed at $41.65 on 2026-09-20, down 2.94%, while coal ETF COAL sat at $25.00, down 3.10% [live prices]. The moves are small, and single-session ETF prints say little about medium-term fuel-cycle demand. But the direction is worth noting against a background of heavy government support for enrichment capacity in the US6.
Japan is not alone in treating nuclear as a strategic hedge. India's nuclear energy mission targets 100 GW of capacity by 2047, and the recent India-Australia summit produced a pact expected to strengthen fuel security and long-term development5. India and Australia are not Japan, and their fuel-cycle arrangements differ, but the shared logic is the same: import-dependent economies are using nuclear to reduce exposure to hydrocarbon markets.
That logic has limits. If a major accident occurs, the long-term effects touch energy, food, water, logistics, communications and healthcare, and land where people can live7. The note that security must be defined across all of those dimensions, rather than just fuel imports, was made in a piece published on 2026-09-17. It is the argument that has quietly shaped Japanese public opinion since Fukushima, and it explains why restart timetables have repeatedly slipped.
The harder question for a gas trader is what Japan's nuclear programme means for LNG demand. Japan's hydrocarbon share is 60-70%, and the 16 GW of planned nuclear additions would, if delivered, reduce the call on imported coal and gas2. But the near-term picture is dominated by KEPCO's utilisation guidance and by the Rokkasho cost line, both of which push the fuel-cycle and restart timetable out rather than forward4,3.
Tokyo's next move is the 6,000-tonne storage target around 2030, which the government itself calls an important policy issue for stable nuclear use7. If Rokkasho's cost keeps rising and storage capacity is not secured, the restart programme stalls on spent-fuel logistics rather than on reactor safety. Watch the next Japan NRG Weekly and the FY2026 utilisation prints from KEPCO's peers for the first sign of whether 70.5% is a floor or a marker of further decline.