The core thesis: China is the only country approving nuclear at industrial scale.
57 reactors operating today. 26 under construction. 40+ planned through 2035.
At 100 GWe — the government's own target — China alone adds ~36M lbs/yr of new demand vs today.
That is nearly the entire current US annual requirement.
U₃O₈ Demand Forecast 2024–2035 (M lbs/yr)
Bear — 85 GWe by 2035 (delays)
| Year |
GWe (Base) |
Bear (M lbs) |
Base (M lbs) |
Bull (M lbs) |
YoY Δ Base |
Key Driver |
| 2024 | 56.8 | 39 | 40 | 41 | — | Baseline |
| 2025A | ~67.8 | 46 | 48 | 49 | +8M | ~10 units completed — record year; surpassed US demand |
| 2026 | 66.0 | 43 | 46 | 48 | +3M | Construction pipeline maturing |
| 2027 | 71.5 | 46 | 50 | 53 | +4M | 6+ units complete |
| 2028 | 77.0 | 49 | 54 | 58 | +4M | Ramp to 80 GWe |
| 2029 | 82.0 | 52 | 57 | 62 | +3M | Planned approvals online |
| 2030 | 87.0 | 55 | 61 | 67 | +4M | 2025-batch completions |
| 2031 | 91.0 | 57 | 64 | 70 | +3M | Near-100 GWe threshold |
| 2032 | 94.0 | 59 | 67 | 73 | +3M | Steady-state additions |
| 2033 | 96.5 | 61 | 70 | 76 | +3M | — |
| 2034 | 98.5 | 62 | 72 | 79 | +2M | Approaching target |
| 2035 | 100.0 | 63 | 76 | 85 | +4M | Government target |
Methodology: ~200 tU/GWe/yr at 85% capacity factor, converted at 2.6 lbs U₃O₈/tU.
Bear assumes 15% construction delays and 80 GWe by 2030. Bull assumes accelerated approvals post-2025.
Demand includes fresh fuel only — does not model strategic reserve building which adds incremental upside.
Sources: IAEA PRIS, WNA Nuclear Fuel Report 2023, analyst estimates.
Operating Fleet — ~67 Reactors · ~67.8 GWe (Q1 2026 · includes 2025 completions)
| Operator | Design | Units | GWe | Status |
|---|
| CNNC | CNP-600, CNP-1000, HPR-1000 | ~40 | ~39.7 | Operating |
| CGN | ACPR-1000, EPR, HPR-1000 | ~23 | ~24.7 | Operating |
| SPIC | AP-1000 | 4 | 3.4 | Operating |
| Total Operating | Operator split est. — verify IAEA PRIS | ~67 | ~67.8 | |
Under Construction — 26 Reactors · ~29 GWe
| Site / Units | Design | GWe | Est. Complete | Demand Add (M lbs/yr) |
|---|
| Sanmen 3&4 / Haiyang 3&4 | CAP-1000 (AP-1000 variant) | 4.4 | 2026–2027 | +2.0 |
| Fangchenggang 3&4 | HPR-1000 (Hualong One) | 2.2 | 2026 | +1.0 |
| Lufeng 1&2 | HPR-1000 | 2.2 | 2027 | +1.0 |
| Taipingling 1&2 | HPR-1000 | 2.2 | 2027–2028 | +1.0 |
| Xudapu 1&2 | HPR-1000 | 2.2 | 2027–2028 | +1.0 |
| Zhangzhou 1&2 | HPR-1000 | 2.2 | 2026–2027 | +1.0 |
| Sanао 1&2 / Huizhou 1&2 | HPR-1000 | 4.4 | 2028–2029 | +2.0 |
| Additional 10 units (various) | HPR-1000, CAP-1000 | 11.0 | 2029–2032 | +5.0 |
| Total Under Construction | | ~29.0 | | +14M lbs/yr |
6–8 units completing per year through 2030. China approves reactors in annual batches — typically 6–10 units per year since 2022 restart of approvals post-Fukushima pause.
HPR-1000 (Hualong One) is the standard design: 1,100 MWe, 60-year design life, being exported to Pakistan and Argentina.
Each completed unit adds ~0.5M lbs U₃O₈ demand per year at full power.
China produces <10% of what it consumes domestically.
~2,400 tU/yr domestic output (mostly Xinjiang) vs ~15,000 tU/yr current requirements.
The gap is filled by long-term contracts and spot purchases — overwhelmingly from Kazakhstan, with Russia, Namibia, and Australia as secondary sources.
Import Sources — Estimated 2024 Mix
| Country / Source | Est. tU/yr | Share | Security |
|---|
| 🇰🇿 Kazakhstan (KAP) | 5,400 | 43% | Concentrated risk |
| 🇷🇺 Russia (TVEL/Rosatom) | 1,900 | 15% | Geopolitical risk |
| 🇳🇦 Namibia (Rössing/Husab) | 1,500 | 12% | CGN-owned (Husab) |
| 🇦🇺 Australia | 1,250 | 10% | LT contracts |
| 🇺🇿 Uzbekistan | 750 | 6% | Growing |
| 🇳🇪 Niger | 500 | 4% | Disrupted (2023 coup) |
| 🇨🇳 Domestic (Xinjiang) | 1,100 | 8% | In-situ leach, expanding |
| Other / spot | 250 | 2% | — |
| Total | ~12,650 tU | 100% | |
Strategic Context
| Factor | Detail |
|---|
| Strategic reserve | Estimated 3–5 years forward cover. Not disclosed publicly. CNNC builds reserves opportunistically on spot price weakness. |
| Overseas equity | CGN owns 90% of Husab (Namibia, 3,500 tU/yr capacity). CNNC has equity in Kazakh JVs. Covers ~15–20% of needs via owned production. |
| Kazakhstan dependency | 43% concentration is a structural vulnerability. Nazarbayev-era contracts locked in at below-market prices — now rolling to market terms as LT deals expire. |
| Enrichment self-sufficiency | China operates its own enrichment (CNEIC). Less exposed to SWU bottleneck than Western utilities. Can run higher tails assay to extract more U from feed. |
| 2035 supply gap | At 76M lbs demand vs ~10M lbs domestic production, China needs to import ~66M lbs/yr by 2035 — vs ~36M lbs today. A ~30M lbs/yr import increase in 9 years. |
The import increase alone (30M lbs/yr by 2035) equals roughly 1.5× Canada's entire annual uranium production.
This demand cannot be met by Kazakhstan alone without meaningful spot price discovery.
China vs World — Demand Share Over Time
| Year |
World Demand (M lbs) |
China Demand (M lbs) |
China Share |
US Demand (M lbs) |
Note |
| 2010 | 175 | 12 | 7% | 44 | Pre-Fukushima |
| 2015 | 170 | 22 | 13% | 43 | China expansion begins |
| 2020 | 162 | 32 | 20% | 41 | COVID demand dip globally |
| 2024 | ~195 | 40 | 21% | 40 | China ≈ US demand |
| 2025A | ~205 | ~48 | 23% | 40 | China overtook US — 10 units completed |
| 2027 | ~220 | 55 | 25% | 40 | China further ahead |
| 2030 | ~240 | 61 | 25% | 41 | 1 in 4 lbs goes to China |
| 2035 | ~275 | 76 | 28% | 42 | Nearly 1 in 3 lbs |
China surpassed the US as the world's largest uranium consumer in 2025 — ahead of prior model estimates of 2027.
At ~48M lbs/yr China now exceeds US demand (~40M lbs). By 2035, nearly 1 in 3 lbs of global uranium demand comes from China.
Unlike Western utilities which hold inventory and have contract flexibility, China builds strategic reserves on top of operational demand — meaning their actual purchasing exceeds consumption in any given year.
Incremental Demand 2024→2035 by Region (Base Case)
| Region | 2024 (M lbs) | 2035 (M lbs) | Delta | % of New Demand |
|---|
| China | 40 | 76 | +36 | 45% |
| South/East Asia excl. China | 18 | 28 | +10 | 13% |
| Europe | 45 | 50 | +5 | 6% |
| USA | 40 | 42 | +2 | 3% |
| Rest of World | 52 | 79 | +27 | 34% |
| World Total | 195 | 275 | +80 | 100% |
World demand estimates: WNA Nuclear Fuel Report 2023, analyst adjusted for 2025–2026 actuals.
Why China Changes the Uranium Market Structurally
| Factor | Analysis |
|---|
| Scale of build-out |
China is commissioning 6–8 reactors per year — more than the rest of the world combined. Each reactor needs ~400–500 tonnes UO₂ fuel loaded for first core, creating a front-loaded demand spike as units come online. |
| Strategic reserve building |
China purchases uranium for strategic reserves separately from operational demand. These purchases are opaque, opportunistic, and price-insensitive — they absorb spot market supply without generating visible demand data. Estimated at 3–5 years forward cover. |
| Supply concentration risk |
43% of Chinese imports come from Kazakhstan. As LT contracts signed at 2015–2020 prices expire, KAP will reprice to market. China has no choice but to pay — they have no domestic alternative. This repricing dynamic is bullish for spot. |
| No demand elasticity |
Nuclear fuel is <5% of a reactor's operating cost. Chinese utilities do not reduce consumption when uranium prices rise. Demand is fully inelastic — only supply responds to price signals. |
| Hualong One exports |
CNNC/CGN are exporting HPR-1000 (Hualong One) to Pakistan (2 operating), Argentina (under construction), and multiple MOU pipeline countries. Each export unit extends Chinese-sourced uranium demand internationally. |
| Bear case — delays |
Construction delays are the primary bear risk. Chinese reactors average 62 months to build vs 72-month global average — but regulatory bottlenecks, supply chain issues, or policy change could push timelines right. Bear case assumes 20% of planned capacity delayed by 2+ years. |
Bottom line: China is not a marginal buyer. By 2035 it sources 28% of global uranium demand
from a market where new mine supply takes 10–15 years to develop.
The mines being built today are not sufficient to meet 2035 demand even in the bear scenario.
China's build-out is the structural demand floor that no bear thesis can dismiss.
Key uncertainties to monitor: Annual reactor approval batch size (NRC equivalent announcements, typically March–April each year) ·
KAP contract renewal terms as existing deals roll off 2026–2028 ·
CGN/CNNC spot purchasing activity (watch COMEX and physical broker reports) ·
Hualong One export pipeline (each new export = ~0.5M lbs/yr incremental demand).