186 Million Pounds Unfilled: Inside the Uranium Procurement Crunch

US utilities purchased 16% less uranium in 2025 than they did in 2024, even as prices climbed and geopolitical risk mounted. For a commodity that requires years of lead time to move from mine to reactor, that is not a rational response to a tightening market. It is a structural complacency problem playing out in slow motion.

The most recent EIA data quantify the cost of that complacency. US utilities have left approximately 186 million pounds of uranium requirements unfilled through 2035, with the uncovered share rising from roughly 40% by 2030 to nearly 70% by 2031 and virtually 100% by 2035. The nuclear fleet that burns through that fuel cannot be switched off, substituted, or powered by something else. The demand is non-discretionary. The contracting is not keeping up.

Here is the structural case from the EIA data forward: why the gap exists, why utilities keep deferring despite knowing it is there, and what happens mechanically when they finally start competing for the same finite supply. By the time you reach the end, you will have a clear view of where the uranium procurement story stands in mid-2026 and what the setup implies for the years ahead.

What the EIA data actually show about the 2030 coverage cliff

Start with the baseline. At the end of 2024, EIA data showed US operators had left 184 million pounds of uranium requirements unfilled out of maximum anticipated requirements of 418 million pounds for the 2024-2034 period. That is roughly 44% uncovered.

The 2025 survey data, compiled by Bank of America and others from the same EIA framework, show the situation has deteriorated. Maximum deliveries under contract for 2026-2035 total just 174 million pounds against approximately 360 million pounds of anticipated requirements. That leaves around 186 million pounds unfilled, an overall coverage rate of approximately 48%, down from roughly 56% the year before.

Those aggregate numbers obscure the real story. The coverage gap is not evenly distributed across the decade. It is front-loaded with relative comfort and back-loaded with a cliff. Year-by-year contracted volumes tell the countdown:

• 2026: approximately 42 million pounds contracted
• 2030: approximately 17.9 million pounds
• 2032: approximately 9 million pounds
• 2033: approximately 3.2 million pounds

By 2030, roughly 40% of expected demand is uncovered. By 2031, that figure jumps to nearly 70%. By 2035, virtually nothing is under contract.

These are not projections. They are the contracted volumes already locked in EIA data. The coverage cliff is not something that might develop. It is already recorded and widening with each survey period. For anyone holding or considering uranium exposure, this is the structural demand gap that will need to be filled by a fleet with no alternative fuel source.

The uranium supply shortage reaching into 2035 reflects more than utility under-contracting: primary mine production has spent most of the past decade below replacement levels, with the 2014-2020 price collapse deferring or cancelling projects whose output would now be flowing into a market badly in need of incremental pounds.

Why utilities keep deferring despite knowing the gap exists

The numbers are public. Every utility procurement department with an EIA subscription can see the same coverage trajectory. And yet the behavioural response has been to slow down, not accelerate.

Over the 2025 survey period, US utilities committed to term contracts covering only approximately 12.9 million pounds of uranium, set against roughly 186 million pounds of unfilled requirements running through 2035. Taken together, their total purchases were 16% lower in 2025 than in 2024, even though the average price paid hit a record high.

That looks irrational from the outside. From inside a utility procurement department, it has its own logic. Utilities point to projected new mine supply timelines, particularly in Kazakhstan and Canada, as evidence that material will be available when they need it. The assumption is a familiar one: because uranium has historically always found its way to market, the same will hold true going forward.

This institutional posture has a long pedigree. US utilities have not yet developed strong conviction around new reactor construction, which suppresses demand-side urgency. If you do not expect to need dramatically more fuel than you currently burn, the gap feels theoretical rather than operational. Projected mine production becomes a reason to wait, not a risk to hedge against.

The problem is that this behaviour is internally consistent but collectively self-defeating. Each individual utility’s decision to defer is rational if supply materialises on schedule. But the historical record does not support that assumption, and when multiple under-covered utilities recognise this simultaneously, they face the same finite uncommitted supply at the same time. For an investor, the divergence between a documented, widening gap and a procurement sector moving in the opposite direction is the signal to watch. It does not mean the market is wrong about supply. It means the correction, when it comes, is likely to be sharper and more compressed than the gap alone would suggest.

The fuel cycle trap: why 2030 is closer than it looks on the calendar

Most readers will look at 2030 and count four years of runway. The uranium fuel cycle says otherwise.

Uranium purchased as U₃O₈ (triuranium octoxide, the yellowcake concentrate that comes out of a mine) cannot be loaded into a reactor. It must pass through a sequential chain of processing steps, each with its own contracting lead time and scheduling constraint:

The uranium fuel cycle begins with raw yellowcake concentrate and passes through conversion, enrichment, and fuel fabrication before a single assembly can be loaded into a reactor, a sequence that makes uranium uniquely resistant to just-in-time procurement strategies.

1. Mining and milling to produce U₃O₈ concentrate
2. Conversion to uranium hexafluoride (UF₆), the gaseous form required for enrichment
3. Enrichment to low-enriched uranium (LEU), which raises the concentration of the fissile isotope U-235 to reactor-usable levels (typically 3-5%)
4. Fuel fabrication into finished fuel assemblies, purpose-built for a specific reactor design

Under normal conditions, moving from a uranium purchase to a delivered fuel assembly requires approximately two to three years. When conversion or enrichment capacity is constrained, it takes longer.

Orano’s nuclear fuel cycle overview confirms that approximately two years elapse from the point ore is mined to the delivery of finished fuel assemblies, a lead time that makes the mid-2020s the last practical window for utilities to contract fuel needed at reactors by 2030.

That mechanical reality reframes every number in the previous section. Fuel needed in 2030 should have been under meaningful contracting consideration by the mid-2020s. Utilities that have not yet moved are not buying four years of decision time. They are buying a fraction of that, and the fraction is compressing.

How the enrichment constraint shifts procurement pressure toward raw uranium

For several years, the mid-stream bottleneck, conversion and enrichment, dominated the fuel-cycle discussion. Russia controls approximately 40-45% of global enrichment capacity and roughly 20% of global conversion capacity. The US has enacted legislation banning Russian enriched uranium imports by 2028, concentrating demand on Western enrichers whose capacity is expanding but not yet fully at scale.

Orano has outlined plans to raise its enrichment capacity from approximately 7.5 million to 10 million separative work units (SWU, the industry measure of enrichment effort), lifting its market share from 12% to 16% and targeting replacement of roughly 10% of Russian supply. Orano is also planning a new conversion facility in the United States. Urenco USA currently operates at approximately 4.3 million SWU per year, representing roughly one-third of current US fleet enrichment requirements.

Mid-stream constraints have eased materially over the past three to four years, though they have not been eliminated entirely. As Western enrichment and conversion capacity has come online, the acute pricing pressure that burdened utilities on those services has started to abate.

The shift matters directly for the uranium supply thesis: as mid-stream capacity gradually fills in, the tightest constraint moves to the front of the fuel cycle, concentrating procurement pressure on raw U₃O₈ availability. That is precisely the segment where the EIA coverage data record the most significant shortfall.

What happens when utilities finally start competing for the same supply

The conditions for a procurement scramble are now visible in the data. Four earlier sections have established the individual components. Assembled together, they form a specific trigger scenario.

Kazakhstan is the single largest source of primary uranium globally, and it is the production source most commonly cited by utilities as the supply they are relying on. Bank of America flags that Kazakh production is exposed to sulphuric acid supply constraints, a material input risk that could limit the pace or reliability of planned expansion. Forward supply models also depend on projects in Canada and Africa, all of which carry permitting, financing, and technical schedule slippage risk that the mining industry’s historical record suggests is the norm rather than the exception.

The specific conditions that make a disorderly procurement scramble structurally probable:

• Approximately 186 million pounds of US utility requirements remain unfilled through 2035
• Demand is non-discretionary; reactors cannot operate without fuel and cannot switch to an alternative
• Spot market volumes in uranium are thin relative to total consumption, meaning even modest demand surges move prices disproportionately
• Execution risk is concentrated in a single jurisdiction (Kazakhstan) that supplies the largest share of global primary production
• The fuel-cycle decision window is compressed, leaving limited time to respond once a supply signal arrives

The market trigger does not require a global supply crisis. A single high-profile project delay, in a market where multiple utilities are simultaneously under-covered, is sufficient to generate concurrent demand for the same limited uncommitted supply. When demand is inelastic and buyers must secure fuel regardless of price, price discovery tends to overshoot the volume of the shortfall by a wide margin.

By 2031, approximately 70% of US utility uranium requirements are uncovered. That is the year the coverage cliff moves from uncomfortable to acute, and it is less than five years away.

For a uranium investor, the core insight is that the distance between a gradual coverage deterioration and a disorderly price event is not a volume threshold. It is a catalyst event, and the conditions for that catalyst are structural rather than speculative.

For investors exploring how specific supply-chain disruption events have historically triggered disproportionate price moves in thin uranium spot markets, our dedicated guide to uranium market disruption scenarios examines case studies from Fukushima, the Cigar Lake flood, and Kazakh production curtailments.

What mid-2026 actually tells you about the uranium setup ahead

Four analytical threads run through this piece, and each one reinforces the others rather than standing alone.

The quantified gap is real and sits in the EIA data, not in a forecast model. The behavioural deferral is documented in the 2025 survey and has a logical but fragile internal rationale. The fuel-cycle timing is mechanical and non-negotiable; no amount of urgency can compress a two-to-three-year processing pipeline. And the supply-side execution risk skews asymmetrically toward disappointment, because mining projects rarely arrive on schedule or at nameplate volumes.

Two variables will determine whether the procurement crunch materialises as a disorderly scramble or a more orderly repricing:

• The pace of utility contracting acceleration in 2026-2027, which will signal whether procurement departments are beginning to act on the data they already have
• On-schedule delivery of projected Kazakh and Canadian production, which will determine whether the supply utilities are counting on actually arrives

The 2028 Russia import ban provides a structural forcing function that cannot be deferred. Western enrichment and conversion capacity must be in place by that date, and the mid-stream expansions from Orano and Urenco will either meet that deadline or create additional procurement pressure.

These structural market forces extend beyond the EIA coverage data alone: the convergence of the 2028 import ban, accelerating AI-driven data centre electricity demand, and compressed Western enrichment timelines means the procurement crunch the article describes is reinforced by demand-side pressures that were largely absent from prior uranium cycles.

The uranium supply thesis does not require new information to be right. It is a structural condition already visible in public EIA data, whose resolution depends on whether utilities move early or wait for the catalyst that forces their hand. The procurement history of US utilities leans toward reaction over anticipation, which suggests that when the acceleration does come, it will arrive suddenly rather than building gradually.

What you are looking at in mid-2026 is a market where the demand is locked, the contracting is behind, the clock is mechanical, and the supply carries execution risk concentrated in a single jurisdiction. The question is not whether the gap closes. The question is how.

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. Financial projections are subject to market conditions and various risk factors.