Uranium Supply vs Demand 2026–2027: How Big Is the Deficit?
60-second answer: World mines produced 60,213 tonnes of uranium (tU) in 2024, about 90% of reactor requirements, per the World Nuclear Association. That implies a 2024 gap of roughly 6,700 tU, or about 17 Mlbs U₃O₈ (estimated from WNA's rounded 90% figure). The OECD-NEA/IAEA Red Book (2026 edition) puts the gap smaller, at about 2,600 tU (~6.7 Mlbs), because it counts higher production and lower requirements. Secondary supply, mostly inventories, filled the difference. WNA expects requirements to rise from 68,920 tU in 2025 to about 86,000 tU in 2030 in its reference scenario. Whether the gap widens depends on how fast mines ramp against that.
How big is the uranium supply gap?
The 2024 gap sits between about 2,600 tU and 6,700 tU, depending on whose numbers you use. Both bodies agree mines fell short of reactor needs. They disagree on the size because they measure requirements and production differently.
| Line, 2024 | WNA | Red Book 2026 (OECD-NEA/IAEA) |
|---|---|---|
| World mine production | 60,213 tU (real) | 61,924 tU (real) |
| Reactor requirements | ~66,900 tU (estimated: 60,213 ÷ 90%) | ~64,500 tU (418 reactors, 378 GWe, as of Jan 1, 2025) |
| Implied gap | ~6,700 tU (~17.4 Mlbs U₃O₈) | ~2,600 tU (~6.7 Mlbs U₃O₈) |
Sources: WNA World Uranium Mining Production, updated October 23, 2026; NEA/IAEA Red Book press release, September 14, 2026. Gaps are our arithmetic. Because WNA publishes the share as a rounded 90%, its implied gap could sit anywhere from about 6,300 to 7,100 tU.
The gap has narrowed. Mines covered 74% of requirements in 2020 and 90% in 2024 (per WNA), as Canadian output recovered from 3,885 tU to 14,309 tU over the same years.
Demand: how much uranium do reactors need?
Reactors need about 68,920 tU a year as of 2025, roughly 179 Mlbs U₃O₈, per WNA's reactor requirements table, which draws on the September 2025 Nuclear Fuel Report. That covers 443 operable reactors with 405,667 MWe of net capacity.
The forward path is a projection. In the Nuclear Fuel Report 2025 reference scenario, WNA expects requirements of about 86,000 tU in 2030 and just over 150,000 tU in 2040. Its lower scenario reaches just over 107,000 tU by 2040, and its upper scenario just over 204,000 tU.
The implied math: reaching 86,000 tU by 2030 requires growth of about 4.5% a year from 2025. Reaching 150,000 tU by 2040 requires about 5.3% a year for fifteen years. That is arithmetic on WNA's targets, not our forecast.
Three sources add demand on top of today's fleet.
- Restarts bring shut reactors back. Palisades (800 MW) and Crane (835 MW) together are about 1.6 GW, which our 170 tU per GWe-year screening constant puts at roughly 270 tU a year (modeled). See how restarts and uprates affect demand.
- Uprates raise an operating reactor's licensed power, usually by a few percent per plant.
- New builds add the most. WNA counts 79 reactors under construction worldwide, 38 of them in China. Track them on the reactor tracker.
AI-era power deals are a fourth headline, though most do not add demand by themselves. Our nuclear-for-AI deal book tracked 33.3 GW of announced capacity on October 28, 2026, which we model at 12.4 Mlbs a year of incremental U₃O₈. Only 1.6 Mlbs a year of that is from firm, signed deals (modeled). A power contract with an operating plant changes who buys the electricity, not how much fuel the reactor burns.
Primary supply: how much uranium is mined per year?
Mines produced 60,213 tU in 2024, about 156.5 Mlbs U₃O₈, per WNA. The Red Book counts 61,924 tU for the same year, the highest since 2016. Kazakhstan supplied 39%, Canada 24% and Namibia 12% (per WNA). See uranium production by country and the largest uranium mines in the world.
WNA has not yet published a 2025 world total. The two largest producers have reported:
- Kazatomprom produced 25,839 tU in 2025 (100% basis), mid-range of guidance it had already cut by about 18% in August 2024.
- Cameco produced 21.0 Mlbs (its share) in 2025, above revised guidance.
The 2026 guidance points in two directions. Kazatomprom guides 27,500–29,000 tU (100% basis) for 2026, but in August 2025 it cut its nominal licensed level from 32,777 tU to 29,697 tU, about 10% lower, saying the market was not strong enough to justify a return to full levels. Cameco guides 19.5–21.5 Mlbs (its share). Both figures are on the guidance tracker, and quarterly actuals are on the uranium production tracker.
Producer guidance is a soft number. Of 14 producer guidance periods we tracked on October 28, 2026:
- 4 landed above the range
- 3 landed in range
- 0 landed below
- 3 were withdrawn before they could be scored
- 4 are pending or unscored
No scored period has missed low. Meanwhile, three were withdrawn outright, including Peninsula Energy's Lance guidance for both 2025 and 2026. Withdrawn guidance is a miss the scoreboard cannot count. The current tally is on the guidance tracker.
New mines are slow. The Red Book found that no new uranium mining project began production in 2023–2024, and puts typical lead times for new projects at 15 to 20 years. WNA's fuel report warns that top producing mines are expected to deplete in the 2030s, per World Nuclear News.
Secondary supply and inventories: what fills the gap
Secondary supply is uranium that reaches reactors without being freshly mined. It comes from utility and supplier inventories, underfeeding (enrichers running extra centrifuge work so they need less natural uranium feed), downblended weapons material, government stockpiles and reprocessed fuel. The secondary supply guide breaks down each source.
No public survey measures secondary supply as a world flow. What exists is regional:
- US commercial inventories were 170.0 Mlbs U₃O₈ equivalent at end-2025 (preliminary), up from 156.0 Mlbs at end-2024, per EIA's Uranium Marketing Annual. Inventories built during 2025 rather than drawing down.
- Russia, China, Japan and India publish no inventory survey.
Financial vehicles also hold pounds off the market. The Sprott Physical Uranium Trust (SPUT) held 81,697,348 lb as of November 5, 2026 (per Sprott). It had added 250,000 lb between June 22, 2026, when our tracking began, and October 28, 2026 (per our SPUT tracker), and bought nothing in the 30 days to October 28 because it traded below NAV. Its units traded 13.40% below NAV on November 5, 2026 (Sprott). The full ledger of holders is on the uranium inventories page.
Two rules apply to these numbers. Regional surveys illustrate; they do not sum to a world total, because EIA and the Euratom Supply Agency use different definitions. And pools overlap: a supplier's stock can include pounds held for a utility.
The utility coverage gap is a separate question
The coverage gap measures how much uranium utilities still need to buy, not how much exists. US utilities had 186.3 Mlbs U₃O₈e of requirements uncovered for 2026–2035 as of end-2025, per EIA's Uranium Marketing Annual. Coverage falls from 82% in 2028 to 6% by 2034. EU utilities, per the Euratom Supply Agency, are at most 96% covered in 2028, falling to 36% by 2034.
Think of it like a homeowner who has fixed the mortgage rate for three years but not for the seven after. The house still exists. The open question is the price of the next contract. Read the utility contracting coverage gap or track it on the utility coverage tracker.
Subtracting coverage percentages from mine output mixes two different questions. Keep them apart.
What would close the gap, and the risks
No one can credibly forecast how the balance resolves. Three scenarios cover the range, each with its own mechanism.
- Tightening. Requirements track WNA's reference path while Kazakh output stays below nominal levels and new mines slip. Inventories draw down and utilities contract harder into the uncovered years.
- Balanced. Restarts and expansions such as Langer Heinrich, which delivered 4.82 Mlbs in FY2026, just above its raised range, keep pace with demand growth, and secondary supply covers a narrowing gap.
- Loosening. Reactor builds and restarts slip, Kazatomprom returns to full licensed levels, or inventories keep building as US stocks did in 2025.
The risks cut both ways. On the supply side, sulphuric acid availability has already cut Kazakh output once and conditions its 2026 guidance. On the demand side, restart schedules have proven soft, and announced data-center capacity is mostly not firm. The price angle on each scenario is on the uranium thesis page.
The gap is real, and it is smaller than it was in 2020.
Frequently asked questions
Is there a uranium shortage? Not a physical shortage at reactors today. Mines produced about 90% of reactor requirements in 2024 (per WNA), and inventories covered the rest. The question is whether mine supply grows fast enough as requirements rise.
How big is the uranium supply deficit? For 2024, roughly 6,700 tU (~17 Mlbs U₃O₈) on WNA figures, or about 2,600 tU (~6.7 Mlbs) on Red Book figures. Both are estimates from published production and requirements, not a measured flow.
How much uranium do reactors need per year? About 68,920 tU in 2025, roughly 179 Mlbs U₃O₈, per the World Nuclear Association. The Red Book puts current requirements at about 64,500 tU for the 418 reactors operating at the start of 2025.
How much uranium is mined per year? 60,213 tU in 2024 per WNA, about 156.5 Mlbs U₃O₈, or 61,924 tU per the Red Book. Kazakhstan produced 39% of the WNA total.
What fills the gap between mine supply and demand? Secondary supply: utility and supplier inventories, enricher underfeeding, downblended weapons material, government stockpiles and reprocessed fuel. No body publishes a measured world total for these flows, so the gap is inferred from production and requirements.
Informational only, not investment advice. Every figure links to its primary source so you can check it yourself.