Two Forces Are Stacking on Uranium Demand. Supply Can’t Match Either

Eighty-one reactors are under construction worldwide as of the end of September 2026, the most aggressive concurrent build pace since the 1980s, and uranium spot prices have responded in step, reaching US$89.45 per pound while a pipeline of first-of-a-kind small modular reactor permits builds quietly behind them.

Two structurally different forces are now stacking onto a market that has not yet attracted a matching wave of new mine capacity. The first is a state-sponsored Asian reactor wave, concentrated in China and India, already drawing fuel and set to keep doing so for decades. The second is a technology-sector pivot toward nuclear as data-centre baseload, moving from memoranda of understanding toward construction permits and early project finance.

Neither force substitutes for the other. They accumulate on the same constrained supply stack. Here is what that stacking actually means for uranium demand, where the supply side can realistically respond, and which parts of the bull case rest on firm ground versus hope.

Asia’s reactor wave is the structural floor under uranium demand

The number worth sitting with is the acceleration. Just over two years ago, in mid-2024, the World Nuclear Industry Status Report counted 59 reactors under construction globally. By end-2024, the World Nuclear Association put the figure at 62 units. As of 30 September 2026, that count stands at 81 reactors, representing 80,488 MWe of net capacity.

That jump from 59 to 81 in a little over two years is not a forecast. It is concrete already poured and fuel-loading schedules already set in motion, because reactors begin drawing uranium several years before they ever produce power.

The WNA reactor construction data underpinning these figures confirms that most units under construction or planned are concentrated in Asia, a regional weighting that locks in decades of fuel procurement obligations regardless of how Western SMR timelines evolve.

China sits at the centre of this. At the mid-2024 cut-off, it accounted for roughly 46% of global construction, 27 of the 59 units. In 2026 alone, Chinese developers broke ground on seven large Generation III+ pressurised water reactors, each in the 1,116 to 1,161 MWe range.

Each of those seven units is a fuel commitment measured in decades. What this tells you is that the demand floor under uranium is not a policy ambition awaiting approval. It is a procurement obligation attached to projects whose foundations are already in the ground.

India’s 100 GW ambition and the gap between aspiration and supply contracts

India adds a second, more complex layer. On 1 March 2026, construction started on Kaiga 5 and Kaiga 6, each a 630 MWe pressurised heavy water reactor, marking a meaningful expansion of the country’s domestic fleet.

Two signals point to how seriously India is planning its fuel supply:

• Regulatory reform has modified domestic liability legislation to reduce risk for foreign contractors, opening the door to international participation in new builds.
• Indian utilities signed supply agreements with Kazatomprom and Cameco totalling more than 45 million pounds of uranium between late 2023 and early 2024.
• Current capacity sits at roughly 8-9 GWe, against a stated government target of 100 GWe by 2047.

The contracts are the concrete signal; the 100 GWe target is the aspirational one. That gap between current capacity and a 2047 ambition is where analysts draw a line, a line this analysis returns to when mapping firm demand against hope.

What the tech sector’s nuclear pivot actually means for uranium: signal or substance?

The milestones that matter arrived in the second half of 2026. The US Nuclear Regulatory Commission issued the Tennessee Valley Authority a construction permit for a GE Vernova Hitachi BWRX-300 small modular reactor at the Clinch River site near Oak Ridge, Tennessee, around 29 September 2026, following a 14-month review. It was the first US construction permit for that design.

Months earlier, in June 2026, the Darlington SMR-1 BWRX-300 project in Ontario, advanced by Ontario Power Generation, recorded its construction start, explicitly positioned to supply regional data-centre and industrial loads.

Why is the technology sector turning toward nuclear at all? Four structural reasons drive the interest, and they operate independently of one another.

1. 24/7 reliability for AI workloads that cannot tolerate the multi-day gaps that renewables plus storage struggle to cover without extreme overbuild.
2. Decarbonisation targets, where firm low-carbon baseload satisfies net-zero commitments without fossil backup.
3. Grid congestion near data-centre clusters, where transmission build-out lags demand and nuclear can anchor co-located supply.
4. Cost visibility from decade-spanning power purchase agreements, attractive against volatile wholesale prices.

The operational case is specific. Compact reactor designs are engineered to run for long stretches before refuelling.

The operational case for data-centre nuclear rests heavily on SMR technology fundamentals, particularly the engineering characteristics that allow compact designs to run for extended intervals without refuelling, matching the multi-decade planning horizons that large cloud providers require.

Next-generation SMRs are designed to operate continuously for 10 to 20 years before requiring core material replacement, a feature that aligns with the multi-decade planning horizons of data-centre hosts far better than short-cycle renewable contracts.

Here is where optimism needs sobriety alongside it. Most publicly reported tech-nuclear arrangements remain memoranda of understanding and feasibility studies rather than binding, project-finance-ready agreements. The NRC accepted GEH’s topical report for review and posted its final safety evaluation in February 2026, a step toward design certification but not certification itself, which the regulator has indicated is still 18-24 months away.

What the construction permit and the Canadian build start actually deliver is credibility, not immediate uranium volume. They mean cloud providers can now point to real projects clearing real regulatory hurdles when building internal investment cases, shifting the conversation from speculative to plannable.

The supply side cannot absorb both demand engines without structural change

Against those two demand stories sits a supply reality that moves far more slowly. The price signal is already flashing.

Uranium traded at US$89.45 per pound on 30 September 2026, up roughly 9% year-on-year, and well above the US$82-83 per pound peak recorded by Sprott Asset Management in December 2025.

A price near US$90 is market participants telling you they already see structural tightness. What they are not yet seeing is a commensurate response in new mine capacity.

The uranium supply gap is already visible in the mismatch between the reactor construction acceleration documented here and a mining sector that largely exhausted its pipeline of shovel-ready Tier-1 projects during the decade-long post-Fukushima bear market.

The reasons are built into the industry. New mines and mill expansions carry long lead times and heavy capital intensity, and the bear market that ran from roughly 2011 to 2020 left the sector with few shovel-ready Tier-1 projects. Primary supply is also concentrated in Kazakhstan, Canada, and a handful of other jurisdictions, adding geopolitical and logistics risk to any supply response.

Uranium specialists tend to frame demand as overlapping layers that stack rather than substitute for one another.

In most institutional modelling, SMR fuel demand is formally counted only from around 2030-2031 onward, treated as an option on future demand rather than a present volume driver. A further constraint sits underneath that: many advanced SMR designs require high-assay low-enriched uranium (HALEU), for which commercial-scale enrichment and fabrication capacity outside Russia is not yet established.

The read you should take is this. The market is already pricing structural tightness before the SMR layer arrives. Any combined conventional-plus-SMR demand surge would need supply from projects that do not yet exist at current cost structures.

Where the uranium bull case rests on firm ground and where it leans on hope

The honest version of the uranium thesis separates what is locked in from what is contingent. Both are real. They simply carry different weights.

• Firm-floor demand: concrete poured on 81 reactors under construction, Asian conventional fleets already contracting fuel, India’s 45-million-pound supply agreements signed and dated.
• Aspirational ceiling: India’s 100 GWe target for 2047 against an 8-9 GWe base, and large-scale tech-sector SMR deployment arriving before 2035.

The floor does not depend on anything going right from here. The ceiling depends on several things going right in sequence.

Three risk vectors sit under the SMR demand layer. Regulatory timing comes first: NRC design certification for the BWRX-300 remains 18-24 months out, and full licensing is a multi-year process beyond that. Fuel-cycle readiness comes second: without commercial HALEU supply outside Russia, licensed reactors could still be constrained. Financing comes third: first-of-a-kind units carry technology and construction risk that lifts the cost of capital, a particular drag in an elevated-rate environment.

The GAO’s examination of HALEU supply chain gaps found that commercial-scale enrichment capacity outside Russia remains largely unbuilt, a structural constraint that could delay licensed SMR deployments by years even after regulatory milestones are cleared.

Base case versus optimistic scenario: what the timing of SMR certification changes

The base case treats SMR fuel demand as a 2031-onward layer and asks a single question: is Asian conventional demand alone enough to hold price above the level that incentivises new mines? On current evidence, it plausibly is.

The optimistic case needs three things to break favourably. NRC design certification completing inside the 18-24 month window. HALEU supply chains progressing outside Russia. And at least one major tech-company agreement converting from memorandum to binding PPA. Each would pull SMR uranium demand visibility forward, and each is where you should concentrate your scrutiny rather than treating the whole thesis as one undifferentiated bet.

What the convergence of these two demand engines means for how the market prices uranium from here

The spot price near US$89.45 per pound reflects one of these two engines, not both. It prices the Asian floor, the 81 reactors and the fuel contracting already underway. It does not yet price the SMR ceiling, which institutional models still count from 2030-2031. The distance between those two numbers is where the investment debate actually lives.

The structural case for sustained demand is not speculative when anchored to construction counts and signed contracts. What remains speculative is the magnitude and timing of the SMR increment, and that timing turns on three specific catalysts worth tracking:

• NRC design certification for the BWRX-300, currently 18-24 months out from late 2026.
• HALEU supply-chain development outside Russia, the fuel-cycle bottleneck that could gate deployment even after licensing.
• Tech-company agreements converting from MOU to binding PPA, the signal that fuel-cycle contracting, which must begin years ahead of first fuel load, is genuinely under way.

Any move materially above current spot in the next 24-36 months will likely be a function of how quickly those late-2026 regulatory milestones convert into real procurement signals.

For investors wanting to translate the demand-stacking analysis into a structured framework for capital allocation, our dedicated guide to the nuclear investment paradigm shift examines how institutional portfolios are pricing the gap between current spot and the incentive price required to bring new mine supply online.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and financial projections are subject to market conditions and various risk factors. Forward-looking statements regarding SMR deployment and uranium demand are speculative and subject to change based on regulatory, financing, and market developments.