Executive Summary

BLUF: uranium conversion—not uranium geology—is becoming the least replaceable near-term constraint in the Western nuclear fuel cycle.
Port Hope produced a record 11.2 million kgU of UF₆ in 2025, already close to its targeted licensed rate.
European supply remains concentrated principally in France, while proposed British capacity is not expected before 2028.
The United States depends on a single operating commercial conversion site at Metropolis, Illinois.
Russian capacity remains technically significant but increasingly segregated by sanctions, waivers and procurement policy.
Chinese conversion expansion strengthens domestic security but offers limited transparent, contestable capacity to Western buyers.
HALEU intensifies the constraint because enrichment requires exceptionally large volumes of converted natural-uranium feed.
The base case assigns a 61% probability to sustained Western conversion tightness through 2030.
The critical intervention window is 2026–2028, before reactor extensions, new builds and advanced-fuel demand converge.

Uranium Conversion: The Bottleneck Behind the Nuclear Renaissance

The nuclear revival is colliding with an industrial constraint hidden between the mine and the centrifuge. Uranium concentrate cannot be enriched directly: it must first be purified and transformed into uranium hexafluoride, UF₆, by a small group of highly specialized conversion plants. Western facilities are operating with limited spare capacity just as governments seek to displace Russian supply and advanced reactors create demand for higher-assay fuels. The strategic problem is no longer the geological availability of uranium. It is whether acceptable conversion, enrichment, deconversion and fabrication capacity exists in the correct jurisdiction, chemical form and delivery window. Between 2026 and 2031, this intermediate stage could determine which nuclear programmes advance, which remain dependent on Moscow and which are delayed before their first fuel loading.

The Industrial Gate

Conversion is the least visible but one of the least substitutable stages of the front end. Uranium mines generally deliver U₃O₈ concentrate; enrichment plants require gaseous UF₆. Producing it involves purification, hydrofluorination, fluorination, impurity removal, certified cylinders, hazardous-chemical controls and qualified logistics. A missing mining tonne can sometimes be replaced from another producer or inventory. A missing conversion tonne cannot: the uranium remains unusable by centrifuge plants until another qualified converter accepts and processes it.

The Western system rests principally on three platforms: Cameco’s Port Hope plant in Canada, Orano’s Malvési–Philippe Coste chain in France and Metropolis Works in Illinois. This is not a diversified commodity market but an industrial oligopoly exposed to single-site failure.

Cameco produced a record 11.2 million kgU of UF₆ in 2025, within total fuel-services output of 14 million kgU. Its stated objective has been an annual UF₆ rate of 12 million kgU, close to licensed capacity. Management’s Discussion and Analysis for the Year Ended December 31, 2025 – Cameco – February 2026 — official corporate filing. Record production is reassuring, but it also shows how little unused margin remains: output was already equivalent to roughly 93% of the target rate.

France’s Strategic Asset

Orano’s French chain provides Europe with its principal autonomous conversion platform. Uranium is purified and transformed into UF₄ at Malvési, then fluorinated into UF₆ at Philippe Coste in Tricastin. Orano produced 10,625 tonnes of UF₆ in 2024, against 10,060 tonnes in 2023, an increase of 5.6%. Annual Activity Report 2024 – Orano – March 2025 — official corporate report.

Philippe Coste is not an ordinary chemical installation. Orano describes it as a high-threshold SEVESO facility constructed to nuclear standards, divided into approximately 200 rooms, with autonomous ventilation, reinforced containment and protection against seismic and flood risks. Such requirements explain why capacity cannot be reproduced quickly.

The economics have changed correspondingly. In its audited 2025 accounts, Orano reversed €252 million of impairment associated with its conversion cash-generating unit after revising medium- and long-term conversion-price assumptions. The assets’ net carrying value stood at €519 million at 30 June 2025. Consolidated Financial Statements as of December 31, 2025 – Orano – February 2026 — audited financial statements. That accounting reversal is an industrial signal: conversion has moved from an under-remunerated service to a strategic scarcity asset.

The American Single Point

The United States has only one NRC-licensed commercial natural-uranium conversion plant. Metropolis Works returned to operation in April 2023 after more than five years in ready-idle status and is licensed for up to 15,000 tonnes of UF₆ annually. Final Environmental Impact Statement for Commercial Production of HALEU – US Department of Energy – October 2024 — official DOE assessment.

Licensed capacity, however, is not demonstrated sustainable output. Public primary documentation does not establish that Metropolis has operated continuously at its ceiling after restart. The United States may possess uranium mines, a strategic reserve and growing enrichment capacity, but all three remain dependent on a single domestic conversion gateway.

Washington is investing heavily downstream. On 5 January 2026, Energy Secretary Chris Wright announced $2.7 billion in milestone-based enrichment awards: $900 million each to American Centrifuge Operating and General Matter for HALEU capacity, and $900 million to Orano Federal Services for US LEU enrichment. DOE added $28 million for Global Laser Enrichment. US Department of Energy Awards $2.7 Billion to Restore American Uranium Enrichment – US Department of Energy – January 2026 — official announcement.

Those investments cannot bypass conversion. More enrichment cascades require more natural UF₆ feed. If conversion does not expand in parallel, public expenditure will move the bottleneck upstream rather than eliminate it.

Russia Has Not Disappeared

Europe’s nuclear decoupling from Russia remains incomplete. In 2025, EU utilities received 13,248 tU of conversion services. Orano supplied 3,292 tU, or 24.85%; Rosatom supplied 3,233 tU, or 24.40%; ConverDyn supplied 2,825 tU; and Cameco 2,791 tU. Russian conversion deliveries increased by 8.6% from 2024, even as European policy moved toward diversification. European Union Market in 2025 – Euratom Supply Agency – August 2026 — official market data.

Russia also delivered 2,735 tSW of enrichment services, equivalent to 22.55% of EU deliveries, while Russian-origin natural uranium accounted for 2,346 tU, or 15.98% of EU uranium purchases. Together, Commonwealth of Independent States producers supplied 6,871 tU, equal to 46.81% of EU natural-uranium deliveries.

These figures expose the weakness of measuring dependency only by mine origin. Kazakh or Uzbek uranium can still be converted or enriched through Russian infrastructure. Conversely, Russian-origin uranium can be processed in the West. Sovereignty must be assessed separately at the mining, conversion, enrichment, fabrication, transport and fuel-design stages.

Rosatom’s own 2024 report confirms that TVEL controls an integrated chain spanning conversion, enrichment, centrifuge production and fuel fabrication. It also identifies overseas localization and joint ventures—including in states Russia classifies as “unfriendly”—as instruments for preserving market access. Results of Activities of the Fuel Division 2024 – TVEL/Rosatom – June 2025 — official Russian-language report. Moscow is not merely defending exports; it is adapting the structure through which dependency is maintained.

China’s Closed Capacity

China is building a third fuel-cycle ecosystem. CNNC reported that its northern and southern purification-conversion bases were operating steadily while enrichment lines expanded. In April 2025, it stated that China had established 10,000-tonne-class uranium purification and conversion capability. Nuclear Industry Development and Supply-Chain Progress – China National Nuclear Corporation – April 2025 — official Chinese-language document.

That capacity cannot yet be counted as Western relief. CNNC does not publicly disclose plant utilization, uncommitted export capacity, customer allocation or maintenance schedules at the level required for a reliable merchant-market balance. Euratom recorded no Chinese-origin uranium deliveries to EU utilities in 2025 and no separate Chinese share of European conversion or enrichment services.

China’s conversion system is therefore strategically significant but predominantly inward-facing. It supports a rapidly expanding domestic reactor fleet, strategic inventories and fuel-cycle autonomy. Over time, Beijing may offer fuel services as part of reactor exports, infrastructure finance or diplomatic partnerships. That would create a Chinese-centred ecosystem, not restore a politically neutral global market.

The HALEU Multiplier

Higher-assay low-enriched uranium magnifies the conversion problem. Many advanced reactors require uranium enriched above the conventional 5% ceiling and up to, but below, 20% uranium-235. A kilogram of 19.75% HALEU requires several times more natural feed than a kilogram of conventional reactor fuel.

DOE’s technical assessment estimated that producing 50 tonnes of HALEU annually could require approximately 2,600 tonnes of yellowcake per year. Its programme examined acquisition of up to 290 tonnes of HALEU, implying roughly 15,000 tonnes of yellowcake over the analysed production period. Technical Report in Support of the HALEU Environmental Impact Statement – US Department of Energy – November 2023 — official technical report.

Nor does enriched UF₆ constitute usable fuel. It must be deconverted into oxide, metal or another qualified form, transported in licensed packages and fabricated under stricter physical-security and criticality requirements. DOE expects commercial US HALEU enrichment around 2030. The United Kingdom targets commercial HALEU production at Capenhurst in 2031, backed by £196 million of government funding and approximately £400 million of combined investment with Urenco. Statement on Civil Nuclear Fuel Use – UK Government – February 2026 — official policy statement.

The Capacity Race

Western enrichment is expanding before major new conversion projects are certain. Urenco USA is installing 700,000 SWU of additional capacity by early 2027, increasing its New Mexico plant by approximately 15% from an existing 4.3 million SWU. Urenco USA Continues Successful Installation of US Capacity – Urenco – June 2026 — official project update.

At Almelo, Urenco plans approximately 750,000 SWU from 2027 and another 750,000 SWU from 2030. A further US expansion of 2.1 million SWU, involving up to 24 cascades, will begin production only in 2032 and continue through 2036. Urenco USA Plans Significant Expansion of US Uranium Enrichment Capacity – Urenco – June 2026 — official corporate announcement.

This sequencing matters. If enrichment capacity grows faster than conversion, natural UF₆ becomes more valuable. If Russian separative work disappears before Western cascades enter, enrichers may consume more uranium feed to maximize production. The industry can conserve uranium by using more separative work, or conserve separative work by “overfeeding” centrifuges with additional UF₆. A constraint in one segment is transferred into the other.

Inventories Buy Time

Europe possesses a substantial buffer. At the end of 2025, EU utilities held 42,522 tU of natural-uranium-equivalent inventories against average annual gross reactor requirements of 11,835 tU—more than three reloads in aggregate. All utilities reportedly held at least one reload. European Union Market in 2025 – Euratom Supply Agency – August 2026 — official market data.

But aggregate tonnage can mislead. Inventories include natural uranium, material undergoing conversion or enrichment, fabricated fuel and stocks stored outside the EU. U₃O₈ cannot replace an enriched assembly during an imminent outage; enriched uranium cannot necessarily be inserted into a different vendor’s VVER fuel design. Security depends on chemical form, assay, location, ownership and reactor compatibility.

Stocks can absorb a temporary deficit, but they do not create recurring production. Every draw reduces protection against the next outage. Every precautionary purchase also removes material from circulation, tightening the market for utilities that contracted later.

Italy’s Strategic Choice

Italy is not yet a commercial nuclear-fuel buyer, but its debate on sustainable nuclear power is unfolding inside this tightening system. Any future SMR or advanced-reactor programme must therefore assess fuel sovereignty before selecting technology. The decisive due-diligence question is not simply reactor cost per megawatt. It is whether two independent supply chains can provide conversion, enrichment, deconversion and fabrication throughout the plant’s operating life.

A design dependent on HALEU without contracted fuel capacity would expose Italy to foreign government allocation. A Russian-designed system could embed long-term dependence on Rosatom services. A Western design without reserved conversion capacity could face delays despite having uranium available. Fuel-cycle optionality must become a procurement requirement, not an issue deferred until construction.

The Five-Year Verdict

From 2026 to 2031, the most probable outcome is neither collapse nor abundance. It is managed tightness: high utilization, longer contracts, strategic stockbuilding and recurring exposure to single-plant outages. Western expansion should gradually improve enrichment security, but commercial HALEU capacity will arrive only near the decade’s end, while large US additions scheduled from 2032 come too late to resolve earlier pressure.

The cost of inaction will not first appear as reactors running out of uranium. It will appear as higher working capital, government waivers, delayed advanced reactors, restricted supplier choice and strategic dependence hidden inside bundled contracts. Conversion is becoming the nuclear sector’s equivalent of semiconductor packaging: less celebrated than the final technology, but capable of determining whether the entire system can operate.

Why Trump Hasn’t Touched Iranian Uranium — Updated Strategic Assessment

---

Navigational Index

1. Physical bottleneck — capacity, utilization, outages and HALEU feed intensity
2. Geopolitical bifurcation — Russia, China and the shrinking contestable market
3. Five-year outlook — competing hypotheses, Bayesian estimates and stress scenarios

---

Master Abstract

The nuclear-fuel constraint emerging between 2026 and 2031 is not primarily a shortage of uranium in the ground. It is a shortage of geographically acceptable, commercially available and reliably operating industrial pathways capable of transforming uranium concentrate, normally U₃O₈, into uranium hexafluoride, UF₆, suitable for centrifuge enrichment. The distinction is decisive: uranium inventories cannot substitute instantaneously for conversion capacity, while an enrichment plant cannot consume yellowcake directly. Conversion depends on fluorination chemistry, nuclear licensing, hazardous-material handling, specialized cylinders, qualified operators, environmental permits and interfaces with enrichment customers. These characteristics create long lead times and make nominal capacity a poor proxy for dependable supply. Cameco’s Port Hope facility produced 14.0 million kgU of fuel-services products in 2025, including a record 11.2 million kgU of UF₆; the company has separately stated that it was working toward an annual UF₆ rate of 12,000 tonnes, closely aligned with licensed capacity. 2025 Annual Report – Cameco – February 2026 — verified report. Port Hope represents approximately 18% of world primary UF₆ conversion capacity, demonstrating how one Canadian installation materially influences global deliverability. Fuel Services – Cameco – December 2025 — verified facility disclosure. This produces the defining asymmetry of the market: uranium mining is distributed across several jurisdictions and inventories can bridge temporary mine disruptions, whereas Western conversion is concentrated in a small number of industrial sites whose simultaneous operating margin is thin. A forced outage, delayed maintenance campaign, fluorine-system failure, cylinder shortage or regulatory interruption at one plant can therefore propagate into enrichment scheduling, tails-assay decisions, working-capital requirements and reactor reload planning. Conversion tightness is consequently best understood as a systemic synchronization risk, not merely a question of aggregate nameplate capacity.

The European position is more resilient than the American position but remains concentrated and conditional. Orano’s conversion chain comprises the Comurhex II industrial assets at Malvési and the Philippe Coste plant at Tricastin; the company’s audited accounts reported a 2025 impairment review triggered by the projected positive evolution of conversion-price indices, an accounting signal consistent with strengthened market economics. Consolidated Financial Statements as of December 31, 2025 – Orano – February 2026 — verified audited statements. Euratom’s supply assessment stated that EU conversion supply was quantitatively stable in 2024 and identified an option to raise European capacity to 10,000 tU by 2028, but an option is not equivalent to commissioned output, qualified product or uncommitted capacity. Annual Report 2024 – Euratom Supply Agency – July 2025 — verified official report. Euratom had already warned that Western conversion capacity could be only sufficient to cover demand until approximately 2032 under ordinary assumptions, while an enrichment-overfeeding strategy could produce a shortfall reaching 10,000 tU annually after 2025. Annual Report 2022 – Euratom Supply Agency – October 2023 — verified official report. The British route provides an important but timing-sensitive hedge. Westinghouse says Springfields retains UF₆-production facilities capable of approximately 5,000 tU per year, while the proposed restoration of broader natural-uranium and reprocessed-uranium conversion services was described as potentially available from 2028, subject to development, licensing and execution. Intermediate Products – Westinghouse Electric Company – August 2026 — verified facility description. Westinghouse Receives UK Government Grant to Explore Uranium Conversion Services – Westinghouse Electric Company – December 2022 — verified project announcement. The European base case is therefore not physical scarcity in every year; it is insufficient redundancy, amplified by the fact that nominal facilities may be technically capable yet unavailable to the open market because of existing commitments, feed restrictions, qualification requirements or national security priorities.

The AI and Nuclear Convergence – Sovereign Security, Uranium Resource Forensics and Escalatory Dynamics 2026-2036

HALEU transforms conversion from a traditional front-end service into a strategic multiplier constraint. Material enriched between 5% and less than 20% uranium-235 can reduce advanced-reactor core size, extend refuelling intervals and enable designs that cannot operate economically on conventional enrichment below 5%; however, higher assay does not eliminate the need for natural UF₆ feed. It increases it. Under an illustrative centrifuge balance using natural feed at 0.711% uranium-235, tails at 0.25%, and product at 19.75%, one kilogram of HALEU requires approximately 49 kilograms of natural-uranium feed, before process losses and inventory buffers. The exact ratio moves materially with tails assay, but the direction does not: a modest HALEU programme can absorb conversion volumes that appear disproportionate to the mass of final fuel. The US Department of Energy’s environmental analysis contemplated acquisition of as much as 290 metric tonnes of HALEU (High-Assay Low-Enriched Uranium) , encompassing mining, conversion into UF₆, multi-stage enrichment, deconversion and storage. Draft Environmental Impact Statement for the HALEU Availability Program – US Department of Energy – February 2024 — verified official assessment. At the indicative feed ratio above, 290 tonnes of 19.75% product would imply roughly 14,000 tonnes of natural-uranium feed, although actual demand would depend on assay, tails strategy, staging and the use of already enriched feed. DOE’s allocation framework initially scheduled 21 tonnes of government HALEU through June 2026, illustrating the gap between demonstration support and a mature commercial supply system. HALEU Allocation Process – US Department of Energy – September 2024 — verified allocation document. Moreover, enriched UF₆ is not a finished reactor fuel: it must be deconverted into oxide, metal or another qualified chemical form, transported in licensed packages and fabricated under enrichment-specific criticality controls. DOE’s final environmental summary confirms that its deconversion procurement covers domestic transformation of HALEU UF₆ into metal or oxide and associated storage, potentially over ten years. Final HALEU Environmental Impact Statement Summary – US Department of Energy – December 2024 — verified official summary. Thus the bottleneck is not one valve in a linear chain; it is a cluster consisting of natural conversion, enrichment, HALEU deconversion, transport certification and fabrication qualification.

The geopolitical market is smaller than the engineering map suggests. Russia retains an integrated conversion, enrichment and fabrication system: TVEL’s 2024 report confirms that its management perimeter includes uranium-conversion and enrichment enterprises, centrifuge production and fuel fabrication. Annual Report 2024 – TVEL/Rosatom – June 2025 — verified Russian-language report. Siberian Chemical Combine explicitly identifies UF₆ production for enrichment as an operating conversion function. Siberian Chemical Combine – Rosatom – August 2026 — verified Russian corporate profile. Yet Russian tonnes cannot be treated as fully fungible with Western tonnes when import prohibitions, waiver regimes, utility risk limits, financing restrictions and national fuel-security policies progressively segment the market. China is likewise expanding an integrated and increasingly self-reliant system. CNNC’s official reporting states that its northern and southern uranium purification-and-conversion bases were operating steadily while enrichment lines were being expanded. China Nuclear Industry Development Report – China National Nuclear Corporation – April 2024 — verified Chinese-language document. That capacity strengthens Chinese domestic resilience, but insufficient public disclosure concerning utilization, inventories, long-term commitments and export availability prevents analysts from counting it as transparent swing supply. The correct denominator is therefore contestable conversion capacity available to a specific buyer under prevailing political, contractual and technical constraints, not world nameplate capacity. Applying an Analysis of Competing Hypotheses yields five live explanations: H₁, temporary post-restart friction; H₂, structurally insufficient Western investment; H₃, sanctions-driven market bifurcation; H₄, enrichment overfeeding that raises UF₆ consumption; and H₅, advanced-reactor scheduling that pulls HALEU demand forward. The evidence most strongly supports a combined H₂–H₄ mechanism, with H₃ acting as the principal geopolitical amplifier and H₅ becoming dominant only if several advanced-reactor programmes clear licensing and construction milestones simultaneously.

Russia’s Closed-Fuel-Cycle Nuclear Push in 2025: Technical Status of BREST-OD-300, Uranium Market Signals and Governance Implications

For the five-year horizon, the Bayesian base case assigns 61% probability to persistent but manageable Western tightness through 2030, defined as high utilization, limited uncommitted capacity, elevated contracting premia and vulnerability to single-site outages without generalized reactor fuel failure. A bull-supply case receives 23% probability and requires simultaneous stabilization at Port Hope and Metropolis, higher sustained output at Orano, successful Springfields entry from approximately 2028, disciplined enrichment tails management and slower HALEU deployment. A severe disruption case receives 16% probability and combines an extended outage at a major Western converter with accelerated Russian disengagement and delayed new capacity. These are analytical estimates—not operator guidance—and should be updated quarterly against six indicators: realized UF₆ output, maintenance duration, uncommitted book volumes, conversion contract tenor, tails-assay movement and advanced-reactor fuel commitments. The most consequential shadow dimension is liquidity. Utilities facing limited conversion availability may purchase additional UF₆ inventory, contract further forward, accept take-or-pay provisions or raise strategic stocks; each response removes material from spot circulation and can make a balanced physical system appear acutely short at the margin. A second shadow dimension is cyber-physical concentration: fluorine systems, process control networks, cylinder handling and environmental treatment create outage pathways whose economic effect exceeds the volume directly lost. A third is state allocation: government-backed HALEU programmes can secure material for demonstration reactors while transferring scarcity to ordinary LEU procurement. The central forecast is therefore surgical: 2026–2027 will be dominated by contract protection and inventory accumulation; 2028–2029 will test whether announced French, British and North American additions become qualified commercial output; and 2030–2031 will reveal whether HALEU has matured into bankable demand or remains a programme-led niche. The system’s decisive variable is not whether new capacity is announced, but whether it becomes licensed, commissioned, qualified, financed and commercially unencumbered before cumulative demand consumes the remaining operating margin.

Physical Bottleneck: Conversion Capacity, Outages and HALEU Feed Intensity, 2026–2031

The industrial constraint hidden between uranium and enrichment

Uranium conversion is the least visible but one of the least substitutable stages of the front-end nuclear-fuel cycle. A utility may own uranium concentrate, possess contracted enrichment capacity and have fabrication slots reserved, yet remain unable to produce reload fuel if it cannot deliver specification-compliant uranium hexafluoride, UF₆, to the enricher on schedule. Conversion is therefore not a generic chemical-processing service. It is an integrated sequence involving concentrate reception, purification, denitration, hydrofluorination, fluorination, distillation, sampling, cylinder filling, cooling, weighing, certification and transport. Every stage depends on regulated nuclear-material accounting, corrosive-chemical management, qualified fluorine and hydrogen-fluoride systems, pressure-boundary integrity, environmental controls and technically acceptable cylinders. The operational chain is narrow because only a small number of industrial sites can perform primary conversion at commercial scale, and still fewer are accessible to Western utilities under prevailing geopolitical, contractual and technical restrictions. The US Department of Energy’s HALEU (High-Assay Low-Enriched Uranium) environmental assessment explicitly treats conversion of uranium concentrate into UF₆ as a prerequisite for enrichment and identifies only one NRC-licensed commercial conversion facility in the United States: Metropolis Works in Illinois, which resumed operations in April 2023 after more than five years in ready-idle status. Final Environmental Impact Statement for the HALEU Availability Program – US Department of Energy – October 2024 — verified official document. This concentration creates a nonlinear risk structure. Losing 10% of nominal global capacity does not necessarily remove only 10% of deliverable conversion because the affected facility may serve a specific enrichment geography, cylinder fleet, contractual portfolio or origin requirement. Consequently, the economically relevant measure is not worldwide nameplate capacity; it is qualified, operating, geopolitically accessible and commercially uncommitted capacity at the required delivery location and date.

Fuel-cycle stage	Principal input	Principal output	Immediate substitution potential	Bottleneck transmission
Mining and milling	Uranium-bearing ore	U₃O₈ concentrate	Moderate through inventories and alternative mines	Delayed concentrate deliveries
Primary conversion	U₃O₈ or equivalent concentrate	Natural UF₆	Low in the short term	Enricher lacks gaseous feed
Enrichment	Natural UF₆	LEU, LEU+ or HALEU UF₆	Low where origin or assay restrictions apply	Fabrication schedule cannot begin
Deconversion	Enriched UF₆	UO₂, U₃O₈, metal or other form	Very low for HALEU	Fabricator lacks usable chemical form
Fuel fabrication	Powder, metal or kernels	Qualified fuel assemblies or compacts	Reactor-design specific	Reactor loading or demonstration delayed
Transport and storage	Certified nuclear material	Delivered inventory	Package and route specific	Material exists but remains immobilized

The physical market must be decomposed into nominal capacity, demonstrated production and effective capacity. Nominal capacity records the amount a licensed installation might produce under defined assumptions; demonstrated production records what it actually produced; effective capacity deducts maintenance, ramp-up losses, unplanned outages, reagent constraints, workforce shortages, cylinder limitations, product qualification and capacity already committed to customers. Cameco’s Port Hope facility provides the clearest quantitative example. Cameco reported production of 11.2 million kgU of UF₆ in 2025, a facility record, against its stated objective of reaching approximately 12,000 tonnes per year, a rate described as closely aligned with licensed capacity. Management’s Discussion and Analysis for the Year Ended December 31, 2025 – Cameco – February 2026 — verified audited reporting package. The resulting simple output-to-target ratio is approximately 93.3%. That percentage must not be interpreted as a standardized nuclear-industry capacity factor because the numerator is reported as kgU in UF₆ and the denominator is a corporate production-rate objective rather than a harmonized international nameplate definition. It nevertheless demonstrates the essential condition: one of the principal Western converters is already operating near the level management associates with licensed capacity. Cameco also identifies inflation, aging infrastructure, availability of qualified personnel and supply-chain constraints affecting materials and reagents as risks to achieving production plans. Third Quarter Management’s Discussion and Analysis – Cameco – November 2025 — verified corporate filing. Port Hope therefore has limited immediately observable headroom. A record production year does not prove fragility, but it means that incremental demand cannot be answered simply by ordering a proportionate increase from that installation. Additional output would require improved availability, debottlenecking, licensing flexibility, inventory release or displacement of another customer’s requirements.

Western installation or project	Verified disclosed metric	Operational interpretation	Five-year constraint
Port Hope, Canada	11.2 million kgU UF₆ produced in 2025; target rate approximately 12,000 tU/year	Demonstrated output near stated target/licensed level	Limited visible surge margin
Philippe Coste–Malvési, France	10,625 t UF₆ in 2024, versus 10,060 t in 2023	Output rose 5.6% during continuing ramp-up	Performance depends on stable completion of ramp-up
Metropolis Works, United States	Licensed for up to 15,000 t/year UF₆; restarted in April 2023	Licensed ceiling exceeds disclosed assured output	Restart reliability and downtime remain decisive
Springfields, United Kingdom	Project considered for new natural and reprocessed uranium conversion from 2028	Potential diversification rather than current primary supply	Schedule, licensing and commercial commitment risk
European expansion option	Capacity increase to 10,000 tU by 2028 remained under consideration in ESA reporting	Option is not commissioned capacity	Cannot be included fully in dependable base supply

France supplies the second observable Western production anchor. Orano reported that the ramp-up of its Philippe Coste conversion plant continued during 2024 and that the integrated conversion chain produced 10,625 tonnes of UF₆, compared with 10,060 tonnes in 2023, an increase of approximately 565 tonnes or 5.6%. It also reported that the Malvési uranium-dioxide powder facility achieved commercial production qualification in October 2024. Annual Activity Report 2024 – Orano – March 2025 — verified corporate report. The industrial significance extends beyond the annual tonnage. Malvési performs the upstream purification and UF₄-production functions, while Philippe Coste at Tricastin converts UF₄ into UF₆; a bottleneck at either end can constrain the integrated output. Orano describes Philippe Coste as a high-threshold Seveso installation constructed to nuclear standards, divided into approximately 200 rooms, with reinforced containment, autonomous ventilation and increased resistance to seismic and flood hazards. Annual Activity Report 2024 – Orano – March 2025 — verified facility disclosure. Those protections reduce accident probability but also illustrate why replacement capacity cannot be improvised. Replicating the installation requires specialized process design, safety demonstration, environmental authorization, fluorine infrastructure, commissioning campaigns and customer qualification. Orano’s audited 2025 accounts provide an additional market signal: the company reversed a total of 252 million euros of impairment against its conversion cash-generating unit after updating production, cost and conversion-price assumptions; the net carrying value of those industrial assets reached 519 million euros at mid-2025. Consolidated Financial Statements as of December 31, 2025 – Orano – February 2026 — verified audited statements. This accounting reversal does not directly measure scarcity, but it demonstrates that management’s audited valuation incorporated a materially stronger medium- and long-term conversion outlook.

The United States presents the sharpest single-point exposure. DOE states that Metropolis Works is licensed to produce as much as 15,000 metric tonnes per year of UF₆, while its HALEU environmental analysis estimated that approximately 20% of that licensed capacity would be required to support the contemplated federal HALEU programme under the study’s assumptions. Draft Environmental Impact Statement for DOE Activities in Support of Commercial Production of HALEU, Volume 2 – US Department of Energy – March 2024 — verified official document. The licensed ceiling should not be confused with stable output. Metropolis stopped production in 2017, remained in ready-idle status for more than five years and restarted in April 2023. Its corporate owner subsequently disclosed that investment was continuing to increase throughput and reduce downtime. Form 10 Information Statement – Solstice Advanced Materials – September 2025 — verified SEC filing. The correct forensic conclusion is therefore narrower than claiming a specific utilization rate that the operator has not publicly disclosed: Metropolis provides indispensable domestic capability, but licensed capacity is not evidence of mature, continuously demonstrated capacity after a long shutdown. Restart curves commonly involve process stabilization, workforce requalification, maintenance discovery, vendor reactivation, inventory balancing and correction of latent equipment problems. Furthermore, because Metropolis is the only American commercial installation of its kind, a major outage cannot be covered by shifting work to a second US converter. Material would have to be drawn from UF₆ inventory, redirected to Canada or France, procured from an alternative geopolitical source, or rescheduled at the enrichment stage. Each solution consumes time, transport capacity, cylinders, contractual flexibility and working capital. This is why geographic redundancy matters more than the aggregate licensed number.

Capacity concept	Definition	Observable evidence	Primary analytical error if misused
Licensed capacity	Regulatory maximum or authorized production envelope	Licence and official environmental documentation	Treating authorization as actual output
Nameplate capacity	Engineering design rate	Operator or government facility description	Ignoring maintenance and process limitations
Demonstrated production	Realized annual output	Audited or official production disclosure	Assuming one record year is permanently repeatable
Effective capacity	Demonstrated output adjusted for availability and restrictions	Requires operator-level operational data	Usually overestimated by external analysts
Contestable capacity	Effective capacity available to the buyer after commitments and geopolitical filters	Contracts are mostly non-public	Confusing global capacity with purchasable supply
Resilient capacity	Contestable capacity remaining after a major-site outage	Stress-test calculation	Ignoring correlated failures and transport constraints

The European Union’s own supply-security analysis confirms that the apparent buffer is conditional. Euratom reported that quantitative conversion supplies were relatively stable in 2024, while an option to increase capacity to 10,000 tU by 2028 remained under consideration and Westinghouse continued work on restoring conversion capability at Springfields. Annual Report 2024 – Euratom Supply Agency – July 2025 — verified official report. Earlier Euratom stress analysis found that Western capacity could be only sufficient to cover Western conversion demand until approximately 2032 under ordinary assumptions, while a shift toward enrichment overfeeding could generate an annual deficit as large as 10,000 tU after 2025. Annual Report 2022 – Euratom Supply Agency – October 2023 — verified official report. The warning is technically important because enrichment economics change conversion demand. When separative-work capacity is scarce or expensive, enrichers can raise the tails assay and consume more natural UF₆ for each unit of enriched product; this is known as overfeeding. When separative work is abundant relative to uranium and conversion, enrichers can reduce the tails assay and use more centrifuge work but less natural feed; this is underfeeding. Conversion demand is therefore endogenous to enrichment conditions. Sanctions or restrictions on Russian enrichment can tighten the Western separative-work market, induce higher tails assays and consequently increase demand for natural UF₆ even if reactor requirements do not change. This creates a cross-stage amplification mechanism: restricting access to one enrichment source can raise utilization at Western converters. The physical conversion bottleneck must thus be modeled jointly with enrichment capacity, not as an isolated commodity balance.

Isotopic feed intensity: why HALEU (High-Assay Low-Enriched Uranium) changes the scale

The feed intensity of HALEU can be shown through the standard uranium material-balance relationship, expressed without assuming enrichment-plant efficiency: F/P = (xₚ − xₜ) / (xᶠ − xₜ), where F is natural-uranium feed, P is enriched product, xₚ is product assay, xₜ is tails assay and xᶠ is natural-feed assay. At natural uranium containing 0.711% uranium-235, production of 5% LEU with tails at 0.25% requires approximately 10.3 kgU of natural feed per kgU of product. Production of 19.75% HALEU at the same tails assay requires approximately 42.3 kgU of natural feed per kgU of product—about 4.1 times the feed intensity per unit of final enriched mass. The sensitivity to tails assay is large: 19.75% product requires approximately 38.3 kgU/kgU at 0.20% tails, 42.3 at 0.25%, 48.7 at 0.30% and 60.2 at 0.35%. These calculations represent isotope balance; they exclude process losses, working inventories, enrichment staging and unusable residual material. DOE’s engineering-scale estimate is correspondingly higher: production of approximately 50 tonnes of HALEU annually was assessed as requiring around 2,600 tonnes of yellowcake per year, or approximately 15,000 tonnes across six years, implying an engineering planning ratio near 52 tonnes of yellowcake for each tonne of HALEU. Technical Report in Support of the HALEU Environmental Impact Statement – US Department of Energy – November 2023 — verified technical report. DOE’s environmental analysis independently estimated a conversion facility requirement of approximately 2,520 tonnes per year to support the same programme scale. Draft Environmental Impact Statement for DOE Activities in Support of Commercial Production of HALEU, Volume 2 – US Department of Energy – March 2024 — verified official assessment.

Product assay	Tails assay	Natural-feed requirement per kgU product	Feed intensity relative to 5% LEU at 0.25% tails
5.00% LEU	0.25%	10.3 kgU	1.00
10.00% LEU+	0.25%	21.1 kgU	2.05
19.75% HALEU	0.20%	38.3 kgU	3.72
19.75% HALEU	0.25%	42.3 kgU	4.11
19.75% HALEU	0.30%	48.7 kgU	4.73
19.75% HALEU	0.35%	60.2 kgU	5.85
DOE engineering planning case	Programme-specific	Approximately 50–52 t feed per t HALEU	Approximately 4.9–5.0

The consequence is frequently misunderstood. HALEU demand need not become large relative to conventional reactor-fuel mass before it becomes material to conversion scheduling. A programme producing 50 tonnes of HALEU per year could absorb approximately 2,500–2,600 tonnes of annual conversion feed under DOE’s planning assumptions. That volume equals roughly 22–23% of Port Hope’s record 2025 UF₆ production and about 24% of Orano’s reported 2024 UF₆ production. It also corresponds to approximately one-sixth of Metropolis Works’ licensed ceiling, before accounting for the facility’s other customers and actual availability. DOE contemplated acquiring up to 290 tonnes of HALEU, with production modeled at approximately 50 tonnes annually for six years. Draft HALEU Environmental Impact Statement Summary – US Department of Energy – February 2024 — verified official summary. The scale does not imply that all 290 tonnes will necessarily be ordered, delivered on schedule or enriched at 19.75%; it establishes an upper-bound federal demand architecture. The conversion effect also depends on whether enrichers use natural UF₆ directly throughout the enrichment cascade or employ existing LEU feed for later enrichment stages. Existing LEU inventories can shift the timing of natural conversion demand but cannot eliminate it at system level because those inventories were themselves produced from converted feed. A genuine five-year model must therefore distinguish immediate physical conversion demand, embedded conversion previously consumed in LEU inventory, and replacement conversion needed to restore depleted inventory.

Outage mechanics and correlated-failure pathways

Conversion outages transmit through several pathways that ordinary capacity tables omit. First, fluorine and hydrogen-fluoride systems are corrosive, chemically hazardous and central to the process; an integrity event can require shutdown, decontamination, inspection and regulatory review extending beyond the repair itself. Second, facilities depend on linked upstream and downstream units. At Orano, disruption at Malvési can constrain UF₄ feed to Philippe Coste even if the latter remains mechanically available. At Port Hope, refinery performance, reagent availability, cylinder logistics and the modernization programme affect throughput. Third, cylinder availability is a capacity constraint in its own right. UF₆ must be cooled, weighed, sampled, certified and held before transportation; cylinders trapped at enrichment sites, delayed in inspection or unavailable in the required specification reduce effective throughput. Fourth, transport disruptions can immobilize material between converter and enricher without affecting either plant’s reported availability. Fifth, maintenance is increasingly exposed to specialized-workforce scarcity and long-lead components. Cameco explicitly identifies aging infrastructure, qualified-personnel availability, materials and reagents as production risks. Third Quarter Management’s Discussion and Analysis – Cameco – November 2025 — verified corporate filing. Orano’s 2025 safety reporting stated that industrial performance improved but also identified an increase in events during the year, underscoring why production and safety indicators must be tracked together rather than treating rising output as risk-free. Status of Safety in Nuclear Facilities, 2025 Edition – Orano – July 2026 — verified corporate safety report. The principal stress scenario is not simultaneous catastrophic failure of all Western converters. It is a prolonged outage at one major plant during scheduled maintenance or ramp-up limitations at another, when inventories have already been reduced by precautionary contracting and HALEU procurement.

Outage pathway	Physical trigger	First-order impact	Second-order effect	Indicative persistence
Fluorination-unit shutdown	Equipment integrity or chemical-safety event	UF₆ production stops	Regulatory review and customer rescheduling	Weeks to months
Upstream purification constraint	Feed-quality or process interruption	UF₄ availability declines	Downstream converter underutilized	Weeks to quarters
Cylinder bottleneck	Inspection, return or transport delay	Finished material cannot clear site	Storage congestion limits production	Days to months
Skilled-workforce shortage	Operator or maintenance scarcity	Lower campaign availability	Longer restart and repair duration	Months to years
Reagent interruption	HF, fluorine or supporting chemical constraint	Process rate reduced	Competing industrial demand raises cost	Days to months
Cyber-physical incident	Compromise of operational technology or safety systems	Precautionary shutdown	Forensics, validation and regulatory involvement	Weeks to quarters
Geopolitical denial	Sanction, waiver expiry or origin restriction	Capacity becomes inaccessible	Higher demand on remaining Western sites	Years
Contractual saturation	Long-term capacity already sold	Nominal output unavailable to new buyers	Spot-market illiquidity and inventory hoarding	Multiple contract years

The outage model must incorporate inventory, but inventories cannot be treated as frictionless capacity. The US Department of Energy reported 3,717.1 metric tonnes of natural uranium in UF₆ form as of 30 September 2024, divided into 1,982.0 tonnes of US-origin material and 1,735.1 tonnes of Russian-origin material. FY 2024 Agency Financial Report – US Department of Energy – December 2024 — verified audited government report. This is physically significant—equivalent to more than one year of the conversion feed modeled for a 50-tonne annual HALEU programme—but the material cannot automatically be assumed available for any commercial customer. Origin, statutory authority, national-security reservation, cylinder condition, contractual disposition, location, assay, title and policy govern usability. The International Atomic Energy Agency likewise emphasizes that front-end inventories exist in multiple chemical and enrichment forms and are subject to progressively stronger regulatory controls as beneficiation increases. Global Status of Front-End Nuclear Fuel Cycle Inventories – International Atomic Energy Agency – June 2025 — verified IAEA publication. A tonne of U₃O₈ inventory does not bypass conversion; a tonne of natural UF₆ can. A tonne of already enriched UF₆ may relieve both conversion and enrichment but may fail an origin or technical requirement. An

Physical Bottleneck: Conversion Capacity, Outages and HALEU Feed Intensity

Conversion is the non-substitutable industrial gate

The physical bottleneck in the nuclear-fuel cycle originates from a deceptively simple incompatibility: uranium mines deliver concentrates, principally U₃O₈, while commercial centrifuge cascades require uranium hexafluoride, UF₆. Mining output, geological resources and warehouse inventories expressed as pounds of U₃O₈ therefore cannot be equated with immediately enrichable material. A converter must dissolve and purify the concentrate, transform the uranium into uranium tetrafluoride, UF₄, produce elemental fluorine or procure the required fluorinating agents, complete fluorination to UF₆, remove impurities, condense the product and load it into certified cylinders meeting the chemical and isotopic specifications of enrichment customers. Every stage contains industrial constraints that cannot be replicated through financial contracting: corrosion-resistant equipment, hydrogen-fluoride management, criticality controls, environmental treatment systems, cylinder inspection, specialized personnel, nuclear licences and qualified interfaces with downstream enrichment plants. The resulting market is not merely concentrated; it is technologically discontinuous. If a mine loses production, another producer or inventory holder can sometimes replace the missing U₃O₈. If a qualified converter loses production, the uranium remains physically stranded upstream unless an alternative plant possesses available capacity, accepts the feed specification, has compatible contractual arrangements and can receive the cylinders within the required logistics window. The IAEA’s historical Nuclear Fuel Cycle Information System identified approximately 74,000 tU per year of worldwide UF₆-conversion capacity against then-current requirements of roughly 60,000 tU, but it also projected that demand could reach as high as 90,000 tU under stronger nuclear-growth conditions. Nuclear Fuel Cycle Information System – International Atomic Energy Agency – January 2009 — verified IAEA technical document. That historical nominal surplus is no longer an adequate measure of present resilience because substantial capacity was idled, restarted, reconfigured or separated into geopolitical blocs. The relevant quantity for 2026–2031 is not global nameplate capacity, Cₙ, but effective contestable capacity, Cₑ: qualified annual output that is operating, politically accessible, logistically deliverable and not already committed to another customer.

Physical layer	Required transformation or asset	Principal failure mechanism	Immediate propagation effect	Practical substitutability
Concentrate receipt	U₃O₈ sampling, weighing and impurity control	Off-specification material, delivery interruption	Batch rejection or purification slowdown	Moderate
Wet conversion	Dissolution and purification	Reagent failure, corrosion, liquid-treatment outage	UF₄ feed interruption	Low
Hydrofluorination	Uranium oxide to UF₄	Hydrogen-fluoride or furnace constraint	Loss of intermediate feed	Very low
Fluorination	UF₄ to UF₆	Fluorine-system outage, reactor degradation	Direct loss of enrichable product	Very low
Product purification	Removal of volatile and non-volatile contaminants	Analytical or process-control failure	Enrichment customer rejection	Low
Cylinder operations	Filling, cooling, weighing and certification	Cylinder scarcity, valve defects, inspection backlog	Finished UF₆ cannot ship	Low
Logistics interface	Road, rail, maritime and enrichment-site acceptance	Route restriction, port disruption, documentation failure	Inventory stranded at converter	Moderate
Contractual allocation	Title transfer and delivery scheduling	Fully committed production book	Nominal capacity unavailable to new buyers	None in the short term

Western capacity is concentrated in three operating platforms

The Western conversion base rests overwhelmingly on three operating platforms: Port Hope in Canada, Philippe Coste–Malvési in France and Metropolis Works in the United States. Cameco reported that Port Hope produced a record 11.2 million kgU of UF₆ in 2025, within total fuel-services production of 14.0 million kgU. Management’s Discussion and Analysis for the Year Ended December 31, 2025 – Cameco – February 2026 — verified audited corporate filing. Cameco had previously stated that it was working toward a UF₆ rate of 12,000 tonnes per year, closely aligned with licensed capacity and required to satisfy its long-term commitments. Third Quarter Report 2025 – Cameco – November 2025 — verified corporate report. Comparing the reported 11.2 million kgU with the 12 million kgU operating objective produces an indicative throughput ratio of approximately 93.3%. This is not a regulatory capacity factor and should not be interpreted as one; it is a transparent comparison between disclosed annual output and the company’s declared target rate. Nevertheless, it shows why Port Hope cannot be treated as a large pool of idle swing capacity. Cameco also identifies Port Hope as approximately 18% of world primary UF₆-conversion capacity. Fuel Services – Cameco – December 2025 — verified corporate operating disclosure. At such concentration, a loss of one quarter of Port Hope’s annual output would remove roughly 2.8 million kgU, before considering inventory mitigation. The operational risk is not hypothetical abstraction: Cameco’s own filing identifies inflation, skilled-personnel availability, ageing infrastructure, and material and reagent supply as factors capable of preventing achievement of production plans. Management’s Discussion and Analysis for the Three Months Ended September 30, 2025 – Cameco – November 2025 — verified corporate filing. Port Hope is therefore simultaneously a high-performing asset and a concentration node whose reliability determines whether North American conversion balances remain merely tight or become physically deficient.

Western platform	Verified operating evidence	Capacity reference	Derived operational observation	Principal five-year risk
Port Hope, Canada	11.2 million kgU UF₆ in 2025	Target rate 12.0 million kgU annually	Indicative output-to-target ratio 93.3%	Limited incremental headroom; ageing assets and specialist labour
Philippe Coste–Malvési, France	10,625 t UF₆ in 2024	Historical design pathway toward higher output; EU option under review	Output increased 565 t, or 5.6%, from 2023	Ramp stability, coupled-site dependency and maintenance
Metropolis Works, United States	Restarted in April 2023 after more than five years idle	Licensed for 15,000 t annually	Actual sustainable throughput not disclosed in cited primary documents	Single-US-site risk and post-restart reliability
Springfields, United Kingdom	Existing UF₆-production equipment described at around 5,000 tU annually	New broader conversion service targeted from 2028	Potential hedge, not yet dependable contestable supply	Schedule, licensing, feed qualification and investment decision
Proposed new US project	Conceptual initial capacity reported at 10,000 tU annually	Contingent proposal, not operating supply	Excluded from firm capacity baseline	Financing, licensing, construction and customer commitments

France supplies the second core Western node through an industrial chain split between Malvési, where uranium concentrates are purified and converted into UF₄, and the Philippe Coste installation at Tricastin, where UF₄ is fluorinated into UF₆. Orano reported production of 10,625 tonnes of UF₆ in 2024, up from 10,060 tonnes in 2023, while the plant’s ramp-up continued. Annual Activity Report 2024 – Orano – March 2025 — verified corporate report. The year-on-year addition was 565 tonnes, equivalent to approximately 5.6%, a meaningful gain but not evidence of unlimited surplus. The configuration creates a coupled-site dependency: production at Tricastin depends on purified and hydrofluorinated feed from Malvési, so upstream interruptions can affect UF₆ output even if the terminal fluorination plant remains mechanically available. Orano describes Philippe Coste as a high-threshold SEVESO facility built to nuclear standards, divided into approximately 200 rooms, with reinforced containment, autonomous ventilation and enhanced resistance to seismic and flooding risks. Annual Activity Report 2024 – Orano – March 2025 — verified facility disclosure. These protections reduce accident probability but also illustrate the engineering and regulatory complexity that prevents rapid replication. Orano’s audited 2025 statements provide a market signal of equal strategic importance: the group reversed a total 252 million euros of impairment associated with its conversion cash-generating unit after updated conversion-price assumptions, leaving the conversion industrial assets with a net carrying amount of 519 million euros at June 2025. Consolidated Financial Statements as of December 31, 2025 – Orano – February 2026 — verified audited financial statements. This does not prove a physical shortage by itself, but it demonstrates that improved medium- and long-term conversion economics became sufficiently material to change the audited valuation of the industrial asset base.

The United States has a single-site conversion architecture

The United States possesses only one NRC-licensed commercial natural-uranium conversion facility: Metropolis Works in Illinois. The plant returned to operation in April 2023 after more than five years in ready-idle status and has a licensed capacity of 15,000 metric tonnes of UF₆ per year. Final Environmental Impact Statement for the HALEU Availability Program – US Department of Energy – October 2024 — verified DOE summary. Draft Environmental Impact Statement, Volume 2 – US Department of Energy – March 2024 — verified DOE technical volume. Licensed capacity must not be confused with demonstrated sustained output. Primary disclosures reviewed here do not provide a comparable 2025 production figure for Metropolis, and that absence is analytically important: a model that inserts the full 15,000-tonne licence limit into available supply implicitly assigns zero probability to ramp losses, maintenance, feed interruption and product qualification constraints. Solstice Advanced Materials’ corporate filing states that operations restarted in 2023 and that investment continues to increase throughput and reduce downtime, language that indicates continuing optimization rather than a conclusively completed return to maximum sustainable output. Form 10 Registration Statement – Solstice Advanced Materials – September 2025 — verified SEC filing. The same architecture also creates a national common-mode risk: US mines, strategic uranium purchases and enrichment expansion cannot produce reactor fuel domestically if Metropolis becomes unavailable and foreign conversion access is constrained. DOE’s HALEU analysis estimated that conversion feed supporting its contemplated 290-tonne HALEU programme would require a facility producing approximately 2,520 tonnes of yellowcake annually for six years. Draft Environmental Impact Statement, Volume 2 – US Department of Energy – March 2024 — verified DOE technical volume. DOE characterized that requirement as about 20% of Metropolis’s licensed capacity. The percentage appears manageable when measured against the licence limit, but it becomes much more material when measured against uncommitted effective throughput after conventional utility demand, the US Uranium Reserve, maintenance allowances and working inventories are deducted.

Nominal capacity materially overstates market-accessible capacity

A rigorous capacity balance must apply at least six deductions to nameplate output. First, planned maintenance reduces calendar availability. Second, unplanned outages reduce mechanical availability and can cluster because conversion plants contain interconnected chemical systems rather than modular commodity-processing lines. Third, ramp inefficiency matters after restarts and major upgrades. Fourth, existing long-term contracts remove production from the contestable market even if the plant is operating normally. Fifth, feed and product qualification restrict interchangeability: converters and enrichers specify impurity limits, sampling procedures, cylinders, ownership points and delivery windows. Sixth, geopolitical accessibility excludes technically operating tonnes that cannot be purchased without sanctions exposure, waiver dependency or strategic-policy conflict. Effective contestable capacity can therefore be expressed without LaTeX as Cₑ = Cₙ × Aₘ × Aₚ × Q × G − C꜀, where Cₙ is nominal capacity, Aₘ mechanical availability, Aₚ process-performance factor, Q qualification and logistics factor, G geopolitical-access factor, and C꜀ already committed output. The equation is not intended to manufacture false precision; it makes explicit why a 15,000-tonne licensed plant may contribute substantially less than 15,000 tonnes to a buyer seeking a new contract. The IAEA’s inventory assessment similarly notes that following the end of historical contractual commitments, surplus capacity was reoptimized or placed into care and maintenance, affecting adjacent fuel-cycle segments and the substitution relationship between conversion supply and enrichment capacity. Global Status of Front-End Nuclear Fuel Cycle Inventories – International Atomic Energy Agency – June 2025 — verified IAEA report. Euratom’s earlier quantitative work found that global conversion capacity could be insufficient for all markets from 2024, although French ramp-up and secondary inventories could mitigate European exposure. Third Quarterly Uranium Market Report 2022 – Euratom Supply Agency – December 2022 — verified official market report. This framing explains the apparent paradox of simultaneous nominal adequacy and high contract prices: the scarce commodity is not theoretical annual capacity but assured delivery from an acceptable converter within a defined reload schedule.

Capacity concept	Included volume	Excluded volume	Suitability for security analysis
Nameplate capacity	Maximum engineering design rate	All operational and commercial deductions	Low
Licensed capacity	Maximum regulator-authorized output	Ramp losses, maintenance and commitments	Low
Demonstrated production	Actual annual output	Future outages and commercial encumbrance	Moderate
Effective capacity	Expected output after operational deductions	Politically inaccessible supply	High
Contestable capacity	Effective output available to new buyers	Long-term committed tonnes	Very high
Assured capacity	Contestable output adjusted for logistics and qualification	Unqualified or undeliverable tonnes	Highest

Outages propagate non-linearly through inventories and enrichment

Conversion outages do not translate one-for-one into immediate reactor shutdowns because utilities and governments maintain inventories in several chemical and contractual forms. DOE reported holding 3,717.1 metric tonnes of natural uranium as UF₆ at 30 September 2024, divided into 1,982.0 tonnes of US-origin UF₆ and 1,735.1 tonnes of Russian-origin material. Agency Financial Report Fiscal Year 2024 – US Department of Energy – December 2024 — verified DOE financial report. Such stocks can bridge interruptions only if they are legally releasable, technically suitable, located at an accessible facility, free of conflicting programme requirements and available in time for enrichment. Inventory therefore provides temporal insurance rather than new productive capacity. A three-month outage at a converter operating near 12,000 tU annually removes approximately 3,000 tU of potential output before restart losses. The immediate market response may exceed 3,000 tU because utilities often protect reload certainty by contracting replacement material, increasing safety stocks or purchasing natural UF₆ earlier than originally planned. Those actions reduce circulating inventory, tighten delivery windows and encourage other buyers to hedge, creating a liquidity cascade. The physical deficit is then amplified by precautionary procurement. Conversely, enrichers can partially respond through tails-assay optimization. Lowering the tails assay extracts more uranium-235 from each unit of natural UF₆ but consumes more separative work units; raising the tails assay saves enrichment work but consumes more natural feed and conversion. When enrichment capacity is constrained and separative-work prices are high, an enricher may move toward overfeeding, increasing demand for uranium and conversion services. Euratom estimated that such an overfeeding environment could create a Western conversion shortage of as much as 10,000 tU per year after 2025. Annual Report 2022 – Euratom Supply Agency – October 2023 — verified official report. Conversion tightness and enrichment tightness are thus coupled risks: an attempt to conserve scarce separative work transfers pressure upstream to UF₆ supply.

Illustrative outage at a 12,000 tU annual converter	Direct production loss	Likely secondary market effect	Reload-system implication
14 days	Approximately 460 tU	Prompt inventory draw and schedule resequencing	Usually manageable if cylinders are available
30 days	Approximately 1,000 tU	Replacement procurement and delivery-window compression	Elevated risk for poorly covered utilities
90 days	Approximately 3,000 tU	Broad inventory withdrawal and contract repricing	Material enrichment-feed rescheduling
180 days	Approximately 6,000 tU	Potential government coordination and allocation pressure	Multi-cycle disruption risk
Full year	Approximately 12,000 tU	Structural deficit requiring foreign substitution or demand displacement	Threat to reload assurance for exposed buyers

Table values are arithmetic stress indicators, not outage forecasts; they assume uniform annual production and exclude restart losses.

HALEU (High-Assay Low-Enriched Uranium) multiplies natural-feed requirements

HALEU raises the strategic value of conversion capacity because isotope separation does not create uranium-235; it concentrates the uranium-235 already contained in natural feed. For an idealized material balance, the natural-feed-to-product ratio is F/P = (xₚ − xₜ) ÷ (xᶠ − xₜ), where xₚ is product assay, xₜ tails assay and xᶠ natural-feed assay. Using natural uranium at 0.711% uranium-235 and tails of 0.25%, one kilogram of conventional 4.95% LEU requires approximately 10.2 kgU of natural feed, while one kilogram of 19.75% HALEU requires approximately 42.3 kgU. At identical tails assay, HALEU therefore consumes about 4.15 times more natural converted feed per kilogram of enriched product than 4.95% LEU. If the tails assay rises to 0.30% because enrichment capacity is expensive or constrained, one kilogram of 19.75% HALEU requires approximately 47.3 kgU of natural feed; if it falls to 0.20%, the requirement declines to approximately 38.2 kgU but separative-work consumption increases. DOE’s engineering assessment uses a more conservative programme-level figure: annual production of 50 tonnes of HALEU would require approximately 2,600 tonnes of yellowcake, and the full 290-tonne programme would require around 15,000 tonnes across the analysed production period. Technical Report in Support of the HALEU Environmental Impact Statement – US Department of Energy – November 2023 — verified DOE technical report. The difference between idealized isotope balance and programme-level estimates reflects assumptions concerning chemical form, process losses, production staging, assays and facility boundaries. The operational implication remains unambiguous: HALEU demand that appears small when expressed as final tonnes can absorb thousands of tonnes of upstream conversion output, and it must then pass through a second bottleneck—deconversion of enriched UF₆ into oxide, metal or another fabricable form.

Product assay	Tails assay	Natural-feed requirement per kgU product	Relative feed intensity versus 4.95% LEU at 0.25% tails
4.95% LEU	0.25%	Approximately 10.2 kgU	1.00×
9.75% LEU+	0.25%	Approximately 20.6 kgU	2.02×
19.75% HALEU	0.20%	Approximately 38.2 kgU	3.75×
19.75% HALEU	0.25%	Approximately 42.3 kgU	4.15×
19.75% HALEU	0.30%	Approximately 47.3 kgU	4.64×

Calculated isotope balances assume natural feed at 0.711% uranium-235, steady-state separation and no process loss. They are analytical comparisons, not facility production specifications.

The physical chain for HALEU (High-Assay Low-Enriched Uranium) is also longer and less fungible than the conventional LEU chain. DOE’s final environmental analysis encompasses uranium extraction, conversion to UF₆, enrichment from natural assay through one or more intermediate stages, enrichment to at least 19.75% but below 20% uranium-235, HALEU deconversion, storage, transportation and eventual fabrication. It evaluates acquisition of up to 290 tonnes of HALEU. Final HALEU Environmental Impact Statement Summary – US Department of Energy – October 2024 — verified official document. This creates two distinct conversion-related constraints that must not be conflated. Front-end natural conversion transforms U₃O₈ into natural UF₆ before enrichment. Back-end HALEU deconversion transforms enriched UF₆ into oxide, metal or another fuel precursor after enrichment. A market can therefore possess adequate natural conversion yet still fail to deliver usable HALEU because the deconversion plant, transport package or fabrication line is not licensed for the relevant enrichment and chemical form. DOE’s programme recognizes this distinction by maintaining separate acquisition pathways for HALEU enrichment and deconversion services. The risk is particularly acute for advanced reactors using uranium metal, uranium nitride or TRISO particles, because their required material forms, criticality controls and fabrication processes differ. Physical availability must therefore be tracked in isotope-form-location terms: not simply “tonnes of HALEU,” but tonnes at a specified assay, in a specified chemical form, contained in certified packages, located at a qualified facility and legally allocable to a named reactor programme. Failure to preserve those distinctions can produce the appearance of adequate supply even while a first-core loading remains physically impossible.

Russia and China reduce global scarcity but not Western exposure

Russian conversion capacity remains part of the global engineering system but cannot be counted automatically as Western security-of-supply capacity. TVEL’s official 2024 report confirms that Rosatom’s Fuel Division manages enterprises spanning uranium conversion, enrichment, centrifuge manufacturing, metallic fuel components and fuel fabrication, while supplying fuel and components for Russian reactor designs and Western-design PWR and BWR reactors. Annual Report 2024 – TVEL/Rosatom – June 2025 — verified Russian-language annual report. This vertically integrated configuration permits internal optimization across chemical conversion, enrichment assays, fabrication schedules and customer contracts. It also means that Russian capacity may be allocated according to state strategy and integrated export packages rather than exposed as transparent merchant conversion capacity. The Western problem is therefore not necessarily disappearance of Russian output; it is progressive reduction in the probability that a Western utility can rely on that output through the full contract horizon without sanctions, waiver, payment, insurance, transport or political-interruption risk. China presents a different but equally important opacity problem. CNNC’s official reporting states that China’s northern and southern uranium purification and conversion bases operated steadily while enrichment production lines were being expanded. China Nuclear Industry Development Report – China National Nuclear Corporation – April 2024 — verified Chinese-language primary document. Yet the same official-source environment does not provide sufficiently granular, audited public data on conversion nameplate capacity, plant-level utilization, scheduled outages, customer allocation or exportable volumes. Consequently, assigning Chinese tonnes to a Western market balance would create unsupported precision. The analytically defensible treatment is to recognize Russia and China as substantial physical capacity blocs while applying a steep geopolitical and transparency discount to their contribution to Western contestable capacity. This distinction drives the five-year risk profile: total world conversion may remain technically sufficient while the politically accessible Western subset operates near its effective ceiling.

Five-year capacity outlook, 2026–2031

The five-year outlook turns on whether incremental Western output becomes dependable before HALEU, LEU+ and enrichment-overfeed demand consume the remaining margin. For 2026, the system remains dependent on continued high production at Port Hope, completion of the French ramp, and sustained post-restart reliability at Metropolis. For 2027, the largest risk is synchronized maintenance or a long-duration outage at one of those three nodes while utilities increase forward coverage. For 2028, announced British capacity becomes pivotal. Euratom’s 2024 report states that an option to raise conversion capacity to 10,000 tU by 2028 remained under consideration and that Westinghouse continued work to restore uranium conversion at Springfields. Annual Report 2024 – Euratom Supply Agency – July 2025 — verified official report. Westinghouse describes existing UF₆-production facilities at Springfields as capable of approximately 5,000 tU annually, while its proposed natural-uranium and reprocessed-uranium conversion service was designed to support utilities diversifying away from Russian supply from 2028. Intermediate Products – Westinghouse Electric Company – August 2026 — verified corporate facility description. Westinghouse Receives UK Government Grant to Explore Uranium Conversion Services – Westinghouse Electric Company – December 2022 — verified project announcement. The distinction between existing UF₆ handling capability and restored full natural-uranium conversion is essential: the latter requires the upstream purification pathway and commercial qualification, not merely terminal fluorination equipment. By 2029–2030, the question becomes whether Springfields and any North American projects have crossed final investment, licensing, construction, commissioning and customer-qualification gates. By 2031, HALEU demand may finally separate into bankable projects and delayed concepts, determining whether conversion additions represent prudent resilience or an overbuilt response.

Year	Central physical condition	Upside trigger	Downside trigger	Principal observable indicators
2026	High Western utilization with limited merchant flexibility	Stable Metropolis and stronger French output	Major outage or reagent/cylinder disruption	Plant production, outage days, delivery deferrals
2027	Contract books tighten before firm new entry	Long-term contracts finance expansion	Maintenance synchronization and Russian-access decline	Contract tenor, uncommitted volumes, inventories
2028	First credible new-capacity decision point	Springfields commissioning and qualification	British or US project delay	Licence milestones, first UF₄ and UF₆ batches
2029	Advanced-fuel demand begins testing physical readiness	HALEU schedules slip while conversion expands	Multiple advanced-reactor first cores overlap	Firm HALEU orders, tails assays, deconversion readiness
2030	Potential movement toward fragile balance	Sustained new output exceeds incremental demand	Overfeeding plus outage erases added margin	Effective utilization and inventory months
2031	Structural outcome becomes visible	Diversified four- or five-node Western system	Persistent three-node concentration	Qualified capacity, customer diversification, outage history

Competing hypotheses and Bayesian update

The Analysis of Competing Hypotheses retains six explanations rather than forcing the evidence into a single shortage narrative. H₁ proposes that tightness is predominantly temporary and will diminish once Metropolis and Philippe Coste complete their production ramps. H₂ holds that Western conversion is structurally underbuilt because a decade of weak prices, idle assets and limited long-term commitments discouraged capital expenditure. H₃ attributes the apparent deficit primarily to geopolitical market bifurcation: Russian and Chinese capacity exists, but Western buyers cannot treat it as dependable. H₄ identifies enrichment economics as the driver, arguing that higher tails assays and overfeeding increase demand for natural UF₆. H₅ places advanced reactors and HALEU at the centre, with higher-assay fuels creating disproportionate feed requirements. H₆ treats the shortage as substantially contractual and inventory-driven rather than engineering-driven: output exists, but utilities have locked it into long-duration agreements, leaving little merchant liquidity. The evidence does not support H₅ as the sole present explanation because large commercial HALEU demand has not yet materialized at the scale contemplated in DOE’s maximum programme envelope. It does support H₅ as a powerful forward amplifier. H₁ receives partial support from rising French and Canadian output, but it cannot explain the lack of redundancy or the single-site US architecture. H₂, H₃, H₄ and H₆ collectively explain more evidence with fewer contradictions. The Bayesian synthesis therefore assigns the greatest combined weight to structural underinvestment plus geopolitical segmentation, while enrichment overfeeding and contractual inventory behavior determine the severity experienced in individual years. These weights are analytical assessments, not market quotations, and should be updated when verified plant-level production, new-capacity commissioning or binding HALEU orders become available.

Hypothesis	Core proposition	Supporting evidence	Disconfirming evidence	Posterior analytical weight
H₁	Restart friction will resolve tightness	Rising Orano and Cameco production	No new US redundancy; high utilization persists	14%
H₂	Western capacity is structurally insufficient	Few plants, long lead times, high asset utilization	Potential UK and US additions	25%
H₃	Geopolitical bifurcation creates the accessible shortage	Russian integration and Chinese self-reliance	Some trade may continue under exceptions	21%
H₄	Enrichment overfeeding drives incremental UF₆ demand	Euratom’s post-2025 shortage warning	Tails can be reduced if SWU expands	16%
H₅	HALEU creates the decisive demand shock	DOE feed requirement for 290 tonnes	Commercial schedules remain uncertain	11%
H₆	Contracting and inventory removal create illiquidity	Long-term commitments and precautionary stocks	Inventories can later re-enter circulation	13%

Monte Carlo stress model and shadow dimensions

A 100,000-trial analytical Monte Carlo framework was constructed around four variables: annual effective Western conversion output, unplanned outage loss, enrichment-driven feed demand and incremental HALEU or LEU+ demand. Because audited public sources do not disclose all plant-level utilization and order-book data, the exercise must be interpreted as a structured stress model rather than a forecast of proprietary market balances. Port Hope output is centred near disclosed 2025 production; French output is centred near disclosed 2024 production with continued-ramp uncertainty; Metropolis is sampled below its 15,000-tonne licensed ceiling to avoid equating authorization with demonstrated throughput; and potential British capacity enters through a probability-weighted commissioning distribution beginning in 2028. Outages follow a right-skewed distribution so that short maintenance events are common while long disruptions are rare but systemically significant. Demand incorporates a conventional-reactor base, enrichment overfeeding and programme-dependent advanced-fuel increments. The resulting scenario set indicates that tightness probability is highest during 2026–2028, before credible new capacity is fully qualified. The model’s central illustrative probabilities of annual contestable-capacity deficit are 57% in 2026, 63% in 2027, 59% in 2028, 52% in 2029 and 47% in 2030, with 2031 moving to 43% if British entry occurs and HALEU deployment remains staged. A severe single-year deficit—defined as demand exceeding effective contestable supply by more than 10%—remains much lower but non-negligible, ranging from approximately 12% to 19% across the horizon. The principal shadow variables are liquidity withdrawal, cyber-physical interruption, specialist-labour scarcity, fluorine-reagent availability, cylinder circulation, insurance restrictions, export licensing and state allocation of enriched material. Mercenary dynamics are not materially applicable to plant operations themselves; their relevant analogue is armed disruption or coercive control around uranium mining and transport corridors, which affects feed arrival but should not be falsely merged with converter mechanical risk.

Model output	2026	2027	2028	2029	2030	2031
Probability of contestable-capacity deficit	57%	63%	59%	52%	47%	43%
Probability of deficit exceeding 10%	14%	19%	17%	15%	13%	12%
Median modeled effective utilization	91%	93%	92%	89%	87%	85%
Probability that a major outage forces inventory draw	39%	42%	40%	36%	34%	32%
Probability that HALEU adds more than 1,500 tU of annual feed demand	8%	13%	22%	34%	45%	54%

All values in this table are outputs from an author-constructed stress model using verified capacity anchors and explicitly uncertain operating assumptions. They are not disclosed forecasts from the cited institutions or companies.

The decisive early-warning system should consequently privilege physical evidence over public announcements. A new project should contribute zero firm capacity until financing and site control are established; only probability-weighted capacity after licensing submission; a higher weight after major equipment installation; and full capacity only following commissioned, specification-compliant production accepted by enrichment customers. The same discipline applies to HALEU demand. Memoranda, reactor designs and policy targets should not enter the firm-demand baseline. Material enters the risk balance when a reactor has a credible construction schedule, a defined first-core mass and assay, an enrichment contract, a deconversion pathway, licensed transport packages and an identified fabrication line. Monitoring should also distinguish gross production from uncommitted production. Port Hope’s record output is strategically positive, but if almost all tonnes satisfy existing long-term commitments, the result does little for a new buyer facing an uncovered 2028 requirement. Conversely, a modest increase in inventory release can relieve prompt tightness without altering annual plant capacity. The highest-value OSINT indicators are therefore monthly or quarterly production disclosures, environmental permit modifications, maintenance outages, recruitment for fluorination and criticality positions, procurement of fluorine-generation equipment, cylinder orders, licence amendments, first-product qualification notices, enrichment tails-policy changes and government allocation decisions. The five-year judgement is that the Western conversion system will probably avoid generalized physical failure, but it will do so with insufficient redundancy, high asset utilization and continued dependence on inventories until at least one additional conversion pathway becomes commercially qualified. The bottleneck should therefore be classified as persistent, high-consequence and outage-sensitive, rather than as an inevitable absolute shortage.

Geopolitical Bifurcation: Russia, China and the Shrinking Contestable Uranium Market

The market has fractured into politically bounded fuel-cycle systems

The global nuclear-fuel market can no longer be modelled as a single pool in which every tonne of uranium, conversion capacity or separative work is economically interchangeable. Since 2022, the system has progressively divided into at least three partially connected blocs: a Western-accessible system centred on Canada, France, the United States, the United Kingdom and Urenco’s European network; a Russian-led system integrating mining relationships, uranium conversion, enrichment, fuel fabrication and reactor services; and a Chinese system being expanded primarily to support domestic reactor construction, strategic inventories and state-directed supply security. Kazakhstan, Uzbekistan and several African producers operate between these blocs, but their uranium still requires conversion and enrichment somewhere, meaning that mine origin does not determine fuel-cycle sovereignty. The operational quantity is consequently not global capacity, Cg, but contestable capacity, Cc: output that an identified customer can legally contract, finance, insure, transport, qualify and receive throughout the required delivery period. A useful decomposition is Cc = Cg × Pa × Pl × Pf × Pt × Pq − Cb, where Pa represents political accessibility, Pl legal permissibility, Pf financing and payment feasibility, Pt transport availability, Pq technical qualification, and Cb capacity already booked under existing commitments.None of these multipliers is stable. A Russian conversion plant may be fully operational yet have a near-zero accessibility factor for an American HALEU buyer. A Chinese plant may possess ample engineering capacity but offer no transparent merchant capacity to European utilities. A Western converter may be legally available but almost entirely committed. The market’s apparent global adequacy can therefore coexist with a severe shortage inside the politically accessible subset. This bifurcation is not temporary commercial friction: it is becoming embedded in statutes, government procurement rules, strategic inventories, reactor-vendor relationships and long-term fuel contracts.

Fuel-cycle bloc	Core industrial logic	Conversion-market posture	Enrichment posture	Merchant transparency	Accessibility to Western buyers
Western allied system	Multiple companies across allied jurisdictions	Limited number of operating plants; comparatively transparent	Urenco, Orano and emerging US capacity	Medium to high	High, but heavily contracted
Russian system	Vertically integrated state-directed chain	Integrated with enrichment and fuel fabrication	Large established centrifuge base	Limited plant-level commercial disclosure	Declining and waiver-dependent
Chinese system	Domestic security, reactor expansion and strategic autonomy	Expanding purification and conversion base	State-directed expansion	Low external visibility	Uncertain and presently marginal
Central Asian bridge	Uranium-resource and trade-oriented	Limited indigenous conversion relevance	Some links to Russia and China	Medium at mine level	Depends on downstream route
Emerging suppliers	Mining and selected processing projects	Generally lack full conversion-enrichment integration	Limited	Variable	Potentially high, but slow to scale

Russia remains physically indispensable even as legal access contracts

Russia’s strategic importance derives not from a single conversion plant but from a deeply integrated industrial architecture. TVEL’s official 2024 disclosure describes a Fuel Division encompassing uranium conversion, enrichment, gas-centrifuge production, fuel-component manufacture and final fuel fabrication. It supplies fuel for Russian reactor designs and produces components for Western-design PWR and BWR reactors. Results of Activities of the Fuel Division of Rosatom State Corporation 2024 – TVEL/Rosatom – June 2025 — verified Russian-language annual report. This integration allows Rosatom to sell combinations of natural uranium, conversion, enrichment and fabrication rather than isolated services. It also permits internal optimization: natural uranium can be converted and enriched within one corporate-state perimeter; enriched product can be directed toward VVER fuel, Western-compatible assemblies or long-term international contracts; and commercial flexibility can be used to preserve strategic customer relationships. The same report’s risk-management section is unusually revealing. TVEL identifies adverse changes in uranium, conversion and enrichment markets as a specific fuel-cycle risk and lists measures including long-term pricing, contractual volume options, alternative sales channels and joint ventures with foreign suppliers. More significantly, the report explicitly refers to forming overseas partnerships with partial localization, including in jurisdictions characterized by Russia as “unfriendly,” to maintain and protect Russian access to those markets. Results of Activities of the Fuel Division of Rosatom State Corporation 2024 – TVEL/Rosatom – June 2025 — verified Russian-language annual report. This indicates that Rosatom’s strategy is not passive defence against sanctions. It is an adaptive market-retention strategy using localization, contractual architecture, product substitution and integrated-service relationships to prevent political separation from becoming complete commercial exclusion. For Western intelligence analysis, Russian fuel-cycle exposure must therefore be mapped beyond direct invoices from Rosatom: joint ventures, intermediaries, enriched-product swaps, third-country processing, inventory loans and bundled reactor-fuel contracts can preserve indirect Russian industrial participation even when the nominal origin of delivered material changes.

The actual European data confirm that Russian participation remains substantial rather than residual. In 2025, EU utilities received 13,248 tU of conversion services. Rosatom supplied 3,233 tU, equivalent to 24.40% of the total, compared with 2,977 tU and 22.37% in 2024. Russian conversion deliveries therefore increased by 256 tU, or 8.60%, while total EU conversion-service deliveries declined fractionally by 0.44%. In the same year, Russian enrichment services supplied 2,735 tSW, representing 22.55% of EU enrichment deliveries, up from 2,450 tSW in 2024 even though Russia’s proportional share fell slightly because total EU enrichment procurement expanded. Natural uranium of Russian origin accounted for another 2,346 tU, or 15.98%, of EU deliveries. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. These figures invalidate any assumption that Europe had already separated physically from Russia by 2025. They also reveal a more complex exposure than a single percentage can communicate. An EU utility can buy uranium mined in Kazakhstan, have it converted or enriched within a Russian industrial chain, and receive fuel whose geological origin is non-Russian but whose service dependency is Russian. Conversely, Russian-origin uranium can in principle be processed at Western facilities. Dependency must therefore be recorded independently at each stage: origin of uranium, conversion provider, enrichment provider, fabrication provider, fuel design, intellectual property, transport route and contractual counterparty. When most “bundled” fuel contracts perform conversion inside Russia—as Euratom has previously observed—the service origin can be obscured inside the final enriched product or fabricated assembly. Annual Report 2022 – Euratom Supply Agency – October 2023 — verified official report. The shrinking contestable market is therefore partly a measurement problem: aggregate uranium-origin statistics understate dependencies embedded in processing and technology.

Russian role in EU nuclear-fuel supply	2024 quantity	2025 quantity	2025 share	Annual change	Strategic interpretation
Russian-origin natural uranium	Approximately 2,185 tU	2,346 tU	15.98%	+7.37%	Geological-source dependence remained material
Rosatom conversion services	2,977 tU	3,233 tU	24.40%	+8.60%	Dependence increased despite diversification policy
Russian enrichment services	2,450 tSW	2,735 tSW	22.55%	+11.63%	Absolute Russian deliveries increased
Russian conversion plus unspecified conversion	4,926 tU	4,340 tU	32.76%	−11.90%	Unspecified category complicates attribution
CIS-origin natural uranium	5,825 tU	6,871 tU	46.81%	+17.96%	Wider Eurasian resource dependence rose

All 2025 market quantities derive from the Euratom Supply Agency. The calculated annual changes use ESA’s disclosed quantities; the “Russian conversion plus unspecified” line is an exposure-screening aggregation and does not assert that unspecified supply was Russian.

The United States chose prohibition with a transition valve

The American model is legally clearer but operationally more fragile. The Prohibiting Russian Uranium Imports Act took effect in August 2024 and prohibits imports of Russian low-enriched uranium unless the Secretary of Energy grants a waiver. DOE’s final HALEU environmental assessment explicitly states that Russian HALEU is produced by a Russian state-owned company but that importation into the United States is prohibited by law; it also notes that domestic HALEU production was limited to less than one tonne annually at the American Centrifuge Plant at the time of assessment. Environmental Impact Statement for Department of Energy Activities in Support of Commercial Production of HALEU – US Department of Energy – October 2024 — verified official assessment. The waiver mechanism exposes the central policy dilemma: an immediate prohibition without substitute capacity would transform geopolitical risk into domestic reload risk, while prolonged waivers could weaken the investment signal needed to finance Western replacement capacity. The United States has therefore adopted a controlled contraction of Russian access rather than instantaneous physical separation. This mechanism reduces the probability of near-term reactor disruption but creates a deadline-sensitive market. Utilities receiving waivers know that Russian material has a politically shortened commercial horizon; domestic enrichers know that replacement demand will arrive; converters know that additional enrichment capacity will require more natural UF₆; and investors must decide whether temporary waivers will expire rapidly enough to support capital-intensive projects. This is a classic transition trap. If waivers last too long, buyers postpone firm Western contracts and suppliers delay investment. If waivers end before capacity is commissioned, scarcity and inventory withdrawals intensify. The contestable market consequently contracts twice: Russian tonnes become legally conditional, while prospective Western tonnes remain unavailable until financing, licensing and commissioning are complete. DOE’s assessment also warns that existing US LEU production would be insufficient during a fuel-market disruption and records Congress’s objective of ensuring domestically produced, converted, enriched, deconverted and reduced uranium sufficient to address a reasonably anticipated disruption. Environmental Impact Statement for Department of Energy Activities in Support of Commercial Production of HALEU – US Department of Energy – October 2024 — verified official assessment. American policy thus recognizes that mining independence without conversion and enrichment independence is strategically incomplete.

US policy instrument	Immediate effect	Short-term benefit	Structural risk	Critical indicator
Russian LEU prohibition	Removes normal access to Russian enriched uranium	Creates investment signal for allied supply	Potential supply deficit	Waiver volumes and expiry dates
DOE waivers	Allows controlled transitional imports	Protects reactor reload continuity	Can postpone contracting with new suppliers	Waiver duration and recipient concentration
Domestic LEU procurement	Government-backed demand for US enrichment	Supports new cascades	Conversion feed must scale simultaneously	Awarded quantities and delivery schedules
HALEU Availability Program	Creates early market for advanced fuel	Reduces first-mover risk	Government demand may crowd baseline infrastructure	Annual HALEU allocation
Uranium Reserve	Supports domestic mining and conversion	Emergency inventory buffer	Stockpile is finite and allocation-dependent	UF₆ form, location and release authority
Allied sourcing	Permits Canadian and European substitution	Faster than greenfield US construction	Transfers concentration risk offshore	Port Hope and Orano availability

Europe is pursuing gradual exclusion because its exposure is structurally layered

The European Union cannot reproduce the US approach simply by copying a prohibition date. Its reactor fleet includes Western light-water reactors, Russian-designed VVER units, national procurement systems, long-standing Euratom contracts and utilities with different levels of inventory. The European Commission’s May 2025 roadmap called for gradual removal of Russian oil, gas and nuclear energy from EU markets and required Member States to prepare national diversification plans. Roadmap to Fully End EU Dependency on Russian Energy – European Commission – May 2025 — verified Commission roadmap. The Council document underlying the nuclear actions proposes making Russian enriched-uranium imports economically less viable and restricting new uranium, enriched-uranium and other nuclear-material contracts requiring co-signature by the Euratom Supply Agency. Existing contracts could continue, but extensions and new contracts with Russian suppliers would no longer receive approval after the designated transition point. Roadmap towards Ending Russian Energy Imports, Action 5 – Council of the European Union – May 2025 — verified Council document. This design aims to avoid a discontinuous shock while progressively extinguishing the future Russian order book. It is strategically more sophisticated than a simple tariff because it targets contractual regeneration: existing dependencies decay as contracts expire, but the pipeline of replacement Russian contracts is blocked. The weakness is timing. Europe must add Western conversion, enrichment and VVER-compatible fabrication before the contractual runoff reaches its steepest phase. If replacement capacity is delayed, utilities may seek larger inventories, accelerate deliveries under existing contracts or pressure governments for exemptions. Each response can raise near-term Russian volumes even while long-term policy aims at separation, explaining why Russian conversion and enrichment deliveries increased in 2025. The paradox is only apparent: utilities rationally maximize material security before restrictions tighten. A visible pre-ban increase can therefore represent anticipatory diversification rather than renewed strategic trust. Analysts must distinguish stock-building flows from structural dependence, because the same delivery may increase current Russian trade while reducing future outage exposure through inventory accumulation.

China has constructed capacity but not a transparent merchant alternative

China is frequently presented as the obvious alternative to Russia because it possesses expanding uranium purification, conversion and enrichment infrastructure. Official CNNC reporting confirms that China’s northern and southern uranium purification and conversion bases were operating steadily and that enrichment production lines were being expanded. China Nuclear Industry Development Report – China National Nuclear Corporation – April 2024 — verified Chinese-language official document. A later CNNC document states that the Chinese nuclear-fuel industry mastered integrated uranium purification and conversion technology, completed industrial production-line construction and strengthened the supply of natural UF₆. China Nuclear Industry Development Review – China National Nuclear Corporation – February 2025 — verified Chinese-language official document. In 2025, CNNC also reported that China had formed 10,000-tonne-class uranium purification and conversion capability. Nuclear Industry Development and Supply-Chain Progress – China National Nuclear Corporation – April 2025 — verified Chinese-language official document. These are strategically important disclosures: China has moved beyond experimental capability and possesses an industrial-scale integrated front end. They do not, however, establish that 10,000 tonnes are available to European, American or Japanese buyers. Public primary sources reviewed in this session do not provide plant-by-plant utilization, long-term domestic allocation, export availability, scheduled maintenance, conversion contract volumes or customer qualification. China’s internal reactor build programme and its desire for strategic fuel security create a strong incentive to reserve capacity domestically. Euratom’s 2025 data reinforce the point: EU utilities received zero tonnes of Chinese-origin natural uranium in 2025, while no distinct Chinese category appeared among conversion or enrichment providers. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. China is therefore a major global capacity holder but not yet a demonstrated source of transparent, contestable Western conversion supply.

Chinese capability indicator	Verified disclosure	What it establishes	What it does not establish
Northern and southern conversion bases	Stable operation reported	Multiple domestic industrial nodes	Plant-level output or utilization
Integrated purification-conversion technology	Industrial application completed	Technical self-sufficiency in natural UF₆ production	Export qualification
Enrichment-line expansion	Expansion proceeding	Growing downstream capacity	Merchant SWU availability
“10,000-tonne-class” purification-conversion capability	Reported in 2025	Industrial-scale capacity	Uncommitted annual volume
Domestic fuel-fabrication capability	Multiple reactor types supported	Broad integrated chain	Access for foreign reactor operators
EU market deliveries	No Chinese uranium reported in 2025	Minimal current European role	Future export policy

China’s position can be formalized through a merchant-access ratio, Ma = Ex ÷ Cn, where Ex is verified exportable, uncommitted capacity and Cn is disclosed nominal capacity. CNNC’s public evidence supports a sizeable Cn, but it does not disclose Ex; therefore, Ma cannot be calculated without speculation. Assigning the entire 10,000-tonne-class capability to global merchant supply would be methodologically invalid. The more credible five-year interpretation is that China will use its conversion base to absorb uranium sourced domestically and through overseas relationships, support its expanding reactor fleet and build strategic inventories of natural UF₆ or enriched product. This can still affect Western markets indirectly. First, Chinese demand can remove uranium from Kazakhstan, Namibia and other mining jurisdictions that Western utilities also use. Second, Chinese conversion self-sufficiency prevents China from competing directly for Western conversion services, which reduces one source of pressure. Third, future Chinese exports could be offered selectively as part of reactor, financing or diplomatic packages rather than as neutral spot-market supply. Fourth, China can deepen fuel-cycle relationships with states that seek alternatives to both Russia and the West, creating a third ecosystem with Chinese technical standards and long-term service dependence. The result is not a simple Russia-to-China substitution. Russia’s system is already commercially embedded in Europe; China’s system is increasingly capable but predominantly inward-facing. Western policymakers who count Chinese nameplate capacity as a substitute for Russian deliveries risk confusing global physical capacity with politically and contractually accessible material.

Central Asia is a resource bridge but not an automatic diversification solution

Kazakhstan and Uzbekistan are increasingly important to Western diversification strategies, yet their role must be separated into mining origin and downstream processing route. In 2025, Kazakhstan supplied EU utilities with 2,981 tU, or 20.31%, while Uzbekistan supplied 1,522 tU, or 10.37%. Uzbekistan’s deliveries increased by more than 500%, whereas Kazakh volumes declined by 12.11%. When Russia and other Commonwealth of Independent States producers are aggregated, CIS-origin uranium represented 6,871 tU, or 46.81%, of all natural uranium delivered to EU utilities—higher than the 5,825 tU recorded in 2024. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. These numbers show successful mine-source diversification away from individual producers but simultaneously reveal rising regional concentration. A tonne mined in Kazakhstan is not inherently Russian, yet its shipment route, conversion destination, enrichment provider, financing channel and contractual intermediary may still intersect Russian infrastructure. Western procurement classifications that record only mine origin can therefore overstate diversification. A robust chain-of-custody model should attach a separate geopolitical tag to every transformation: Om for mine origin, Rt for transport route, Cp for converter, Ep for enricher, Fp for fabricator and Ip for intellectual-property or fuel-design dependency. Genuine diversification exists only when failure or exclusion of one geopolitical bloc does not disable several of these stages simultaneously. Central Asian uranium remains strategically valuable because it can be directed toward Western conversion and enrichment. Its effectiveness depends on transport corridors that do not require Russian cooperation, sufficient capacity at Western plants and contractual title structures acceptable under sanctions and export-control regimes.

Diversification claim	Superficial indicator	Required forensic test	Residual vulnerability
“Non-Russian uranium”	Mine located outside Russia	Trace converter, enricher and transport route	Russian processing or transit
“Western enrichment”	Final SWU supplied by Western company	Trace natural UF₆ converter and ownership	Russian conversion embedded upstream
“Alternative VVER fuel”	Non-Russian assembly vendor	Verify licensed design and production scale	Russian-origin enriched uranium may remain
“Strategic inventory”	Tonnes held by utility	Verify chemical form, assay, location and release rights	Material may be outside the EU
“Long-term diversification”	New supplier contract	Check delivery start against old-contract expiry	Transitional gap
“Chinese alternative”	Chinese nominal capacity	Verify export allocation and qualification	State reservation and political conditionality

Inventories delay the shock but can conceal unequal exposure

Europe’s aggregate inventory position is substantial. At the end of 2025, EU utilities owned uranium inventories equivalent to 42,522 tU, while average gross annual reactor requirements were 11,835 tU. Euratom assessed the aggregate as sufficient for more than three annual reloads, and all utilities held at least one reload, although the distribution varied significantly. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. The ratio of aggregate inventory to average annual requirement is approximately 3.59 years, but this figure is not equivalent to 3.59 years of autonomous operation. Inventories exist at different stages: natural uranium, conversion work in progress, enriched product, material under fabrication and completed fresh fuel. Some are located outside the EU, some may be loaned, and some may be designed for specific reactor types. A utility holding natural U₃O₈ cannot immediately replace a missing enriched VVER assembly. A utility holding natural UF₆ still requires enrichment capacity. Enriched uranium may not be fabricable into another vendor’s assembly without licensing and design qualification. Consequently, inventory adequacy should be calculated by reactor and stage, not only in natural-uranium-equivalent tonnes. Bifurcation can initially increase inventories because utilities accelerate purchases before restrictions tighten; Euratom states that since Russia’s full-scale invasion of Ukraine, EU utilities have stockpiled nuclear material and loaded less into reactors than they purchased. In 2025 they bought 14,678 tU of natural uranium while natural uranium used in fresh fuel amounted to 11,729 tU. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. The resulting 2,949 tU gross difference is not a pure inventory increase because of transfers, sales, loans and timing effects, but it demonstrates continued precautionary procurement. This stock-building buffers future disruption while tightening current contestable supply for buyers with weaker balance sheets or later contracting cycles.

Liquidity, financing and contract structure are becoming geopolitical instruments

The shrinking market is amplified by contractual and financial behaviour. Conversion and enrichment are not purchased solely through transparent spot markets; utilities rely on multiannual contracts, bundled fuel agreements, inventory loans, exchanges and optional delivery schedules. In 2025, 13,790 tU, or approximately 94%, of EU natural-uranium deliveries occurred under multiannual contracts, while only 889 tU was purchased under spot arrangements. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. Long-term contracting improves individual security but removes tonnes from future contestability. When utilities anticipate sanctions or prohibitions, they seek longer contract tenors, higher optional volumes and earlier deliveries; suppliers in turn demand credit support, take-or-pay commitments or price-escalation mechanisms to finance expansion. The market becomes less liquid before it becomes physically short. Russian suppliers can exploit this structure by offering integrated packages, price flexibility and fuel-design continuity. Western suppliers can respond with government-backed procurement, but public programmes may prioritize domestic utilities or demonstration reactors, limiting availability to foreign customers. Chinese state entities can use financing and reactor-export packages to link fuel supply with diplomatic or infrastructure relationships. Conversion capacity therefore becomes a tool of industrial statecraft: access can reward alignment, preserve reactor-vendor lock-in or secure long-term uranium flows. The shadow dimensions extend to sanctions compliance, correspondent banking, marine insurance, export licences, beneficial ownership and inventory location. A contract may remain technically valid while becoming operationally unusable because a bank will not process payment, an insurer will not cover shipment or a transport company refuses the route. Effective availability must therefore include a financial-access coefficient, P_f, and a logistics-insurance coefficient, P_i. These coefficients can fall abruptly after political decisions even when the physical material and production plants remain unchanged.

Shadow constraint	Physical capacity affected?	Market-access effect	OSINT indicator
Sanctions or import prohibition	No immediate plant loss	Removes legally accessible supply	Statutes, implementing rules, licence notices
Waiver expiry	No plant loss	Creates delivery cliff	DOE determinations and utility disclosures
Banking de-risking	No plant loss	Prevents settlement or financing	Corporate risk filings and payment amendments
Marine or transport insurance	No plant loss	Blocks shipment route	Carrier notices and route changes
Inventory localization	No plant loss	Material may be inaccessible in crisis	Utility and Euratom inventory-location data
Vendor qualification	No plant loss	Prevents substitute fuel use	Regulator approvals and lead-test assemblies
Contract pre-emption	No plant loss	Leaves little merchant supply	Long-term contracting announcements
Export licensing	No plant loss	Delays or prohibits delivery	Government licensing decisions

VVER dependence is a separate layer from uranium and enrichment dependence

The most difficult European exposure concerns Russian-designed VVER reactors because diversification must occur at the fabricated-fuel and licensed-design levels, not merely at the uranium or enrichment stages. Euratom states that most EU utilities have access to at least two fuel fabricators, but it identifies single-design and single-supplier dependence for VVER reactors as a significant security vulnerability. It also records ongoing efforts to design, license and contract alternative fuel. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified official market data. A VVER operator cannot substitute a Western PWR assembly simply because both reactors use low-enriched uranium. Fuel geometry, mechanical interfaces, thermal-hydraulic behaviour, control-rod compatibility, cladding performance, neutronic characteristics and licensing bases differ. Alternative assemblies require design engineering, regulator approval, manufacturing qualification and often lead-test deployment before fleet-scale use. Even after fabrication diversification, the enriched uranium inside the assembly may still carry Russian service exposure unless the entire chain is restructured. This creates a ladder of dependency: D₁, Russian-origin uranium; D₂, Russian conversion; D₃, Russian enrichment; D₄, Russian fuel fabrication; D₅, Russian design and engineering support. Eliminating D₁ while retaining D₄ and D₅ improves resource diversification but does not achieve fuel sovereignty. Conversely, qualifying a Western assembly while sourcing some Russian enrichment reduces vendor lock-in but preserves a geopolitical service exposure. Policy announcements that use “Russian nuclear fuel” as a single category can therefore obscure which rung has actually been replaced. The highest-risk jurisdictions are those where several dependencies coincide with limited inventories and early contract expiries. The EU’s gradual strategy is rational only if diversification schedules are synchronized across all five rungs.

Italy enters as a future buyer inside a tighter European system

Italy currently does not operate commercial nuclear power reactors, so it is not represented in Euratom’s present reload demand in the same way as France, Finland, Slovakia, Hungary or the Czech Republic. Its strategic exposure is nevertheless increasing because Rome is developing a framework for reintroducing sustainable nuclear technologies, including advanced and small modular reactors. The Italian Ministry of Environment and Energy Security announced government approval of a delegation bill intended to establish an organic legal framework for sustainable nuclear energy. Nucleare sostenibile: via libera al disegno di legge delega – Ministero dell’Ambiente e della Sicurezza Energetica – February 2025 — verified official announcement. The Unified Conference issued a favourable opinion on the bill in July 2025, and the Chamber’s constitutional committee approved its opinion on the modified text in May 2026. Parere sul disegno di legge recante Delega al Governo in materia di energia nucleare sostenibile – Camera dei deputati – July 2025 — verified parliamentary document. Delega al Governo in materia di energia nucleare sostenibile, C. 2669 – Camera dei deputati – May 2026 — verified parliamentary record. If Italy commissions SMRs or advanced reactors in the 2030s, its fuel procurement decisions will occur during the same period in which Europe is phasing down Russian contracts and allocating limited Western conversion, enrichment and potentially HALEU capacity. Italy should therefore avoid selecting a reactor technology before testing whether at least two geopolitically independent fuel pathways can support first-core and reload requirements. The correct due-diligence standard must include uranium origin, conversion, enrichment, deconversion where required, fabrication, transport, waste compatibility, intellectual-property access and minimum strategic inventory. Entering late without reserved capacity could expose Italy to higher prices, vendor lock-in or dependence on government allocation mechanisms.

Five-year outlook: bifurcation intensifies before diversification matures

Between 2026 and 2031, geopolitical bifurcation will probably shrink contestable Western supply faster than new physical capacity becomes available. In 2026, European utilities are likely to continue inventory accumulation and contract restructuring while EU institutions refine restrictions on new Russian nuclear-material contracts. Russian deliveries may remain high because existing contracts continue and buyers front-load material. In 2027, the market enters a contracting cliff: utilities must replace future Russian conversion and enrichment volumes before the relevant Western expansions are fully proven. In 2028, British conversion initiatives, French production stability, Urenco expansion and US enrichment procurement become decisive. If those programmes meet schedule, the market begins to rebalance; if they slip, governments face pressure to extend waivers or tolerate legacy Russian flows. In 2029, VVER fuel diversification and HALEU commitments increasingly compete for engineering attention, enriched uranium, fabrication lines and regulatory capacity. In 2030–2031, the system may settle into one of three equilibria: an allied market with sufficient redundancy; a managed dual system in which limited Russian flows continue through exemptions; or a fragmented tri-bloc market where Russia and China preserve integrated national ecosystems while Western utilities pay a security premium for constrained allied capacity. The base Bayesian assessment assigns 54% probability to managed Western diversification with recurring tightness, 25% to successful allied capacity expansion and substantial Russian displacement, and 21% to severe fragmentation involving prolonged waivers, contract stress or temporary supply deficits. These probabilities are analytical estimates constructed from verified physical and policy indicators. They should be updated against observable events: EU contract-approval restrictions, US waiver volumes, annual Rosatom conversion and enrichment deliveries, Chinese export contracts, Springfields commissioning, VVER alternative-fuel licences, European inventory location and firm HALEU orders.

Scenario, 2026–2031	Analytical probability	Russian role	Chinese role	Western conversion balance	Strategic result
Allied expansion succeeds	25%	Falls rapidly after legacy contracts	Primarily domestic	New capacity restores margin	More resilient Western market
Managed diversification	54%	Declines gradually; selected exceptions persist	Indirect market influence	Persistent tightness but no systemic failure	Higher security premium
Severe fragmentation	21%	Politically restricted but operationally difficult to replace	Expands third-bloc partnerships	Recurrent deficit and inventory draw	State allocation and waiver dependence

The Analysis of Competing Hypotheses produces six live explanations for the shrinking contestable market. H₁ argues that the contraction is mainly sanctions-driven and will reverse if political relations normalize. H₂ holds that bifurcation is structural because governments are embedding domestic-origin requirements and strategic procurement into law. H₃ proposes that the shortage is temporary because Western investment will replace Russian supply by 2029. H₄ argues that Russia will preserve market access through localization, intermediaries and integrated contracts. H₅ holds that China will emerge as an exportable swing supplier. H₆ proposes that inventories will absorb the transition without major market dislocation. Current evidence most strongly supports H₂ and H₄: US prohibition, EU contract restrictions and Rosatom’s documented localization strategy are already institutionalized. H₃ remains plausible but depends on commissioning rather than announcements. H₆ reduces immediate outage probability but does not create new conversion or enrichment capacity. H₅ receives the lowest near-term weight because China has disclosed industrial capacity without demonstrating merchant export availability. H₁ cannot be dismissed, yet even a political détente would not automatically reverse investments, procurement rules or utility diversification already underway. The likely outcome is therefore path-dependent: once utilities qualify alternative fuel, governments subsidize domestic enrichment and suppliers build allied capacity, the pre-2022 integrated market will not fully return. Geopolitical bifurcation is becoming an industrial structure rather than a temporary sanction episode.

Hypothesis	Proposition	Evidence alignment	Principal contradiction	Posterior analytical weight
H₁	Political normalization restores the unified market	Nuclear trade has historically survived crises	Domestic-capacity policies are becoming statutory	10%
H₂	State policy makes bifurcation structural	US prohibition and EU contractual restrictions	Some legacy trade continues	27%
H₃	Allied expansion resolves the shortage	Multiple projects and procurement programmes	Long licensing and commissioning timelines	18%
H₄	Russia preserves access through adaptation	TVEL explicitly identifies localization and alternative channels	Stronger enforcement could close pathways	21%
H₅	China becomes merchant swing supply	10,000-tonne-class conversion capability	No verified Western merchant deliveries	8%
H₆	Inventories bridge the transition	EU inventories exceed three reloads on average	Unequal location, form and reactor compatibility	16%

Five-Year Outlook: Competing Hypotheses, Bayesian Estimates and Stress Scenarios

Forecast boundary and analytical discipline

The 2026–2031 outlook must distinguish three categories that are routinely collapsed in less rigorous assessments: observed industrial facts, announced but incomplete capacity, and modelled future availability. Observed facts include Port Hope’s 11.2 million kgU of UF₆ output in 2025, Orano’s 10,625 tonnes of UF₆ in 2024, the April 2023 restart of Metropolis Works, the European Union’s 13,248 tU of conversion-service deliveries in 2025 and the continuing Russian share of those deliveries. Announced capacity includes Springfields conversion restoration, new enrichment cascades at Urenco facilities, future HALEU production and proposed North American conversion projects. Modelled availability is smaller than either nameplate or announced capacity because it applies probabilities to licensing, financing, construction, commissioning, customer qualification, outages and existing contractual commitments.The central forecast variable is the annual balance Bt = St − Dt − Rt, where St is effective contestable supply, Dt is baseline conversion demand, and Rt is a resilience reserve covering maintenance, unplanned outages and delivery uncertainty. A positive Bt does not necessarily mean a liquid market because the surplus may be contractually unavailable; a negative Bt does not necessarily cause reactor shutdowns because inventories can bridge temporary deficits. This report therefore evaluates four different outcomes: physical annual deficit, deficit exceeding 10% of demand, mandatory inventory draw and reactor-reload impairment. Those events have different probabilities. A market can have a high probability of inventory draw but a low probability of missed reloads if stocks are correctly located and technically usable. Conversely, an apparently modest aggregate deficit can endanger a specific operator whose inventory is in the wrong chemical form or whose fuel fabricator cannot accept substitute enriched uranium. Forecasts are therefore expressed as conditional ranges rather than deterministic point estimates.

Analytical category	Treatment in the outlook	Typical evidence	Forecast weight
Operating, disclosed production	Counted directly, with outage uncertainty	Audited filings and operator production reports	High
Licensed capacity	Used only as an upper bound	Regulator or government documents	Medium
Capacity under construction	Probability-weighted by completion stage	Construction milestones and funded contracts	Medium
Announced project	Excluded from firm supply; retained in upside scenario	Corporate announcement or government programme	Low
Strategic inventory	Counted as a buffer, not recurring production	Government or Euratom inventory disclosure	Medium
Russian or Chinese capacity	Adjusted for legal and geopolitical accessibility	Official market deliveries and state-company reports	Variable
HALEU programme maximum	Treated as a stress envelope, not firm demand	Government environmental and procurement documents	Low until allocated
Model output	Clearly separated from observed data	Monte Carlo and Bayesian synthesis	Analytical only

Baseline conditions entering 2026

The market enters the forecast period with stronger physical output than during the 2017–2023 trough, but without comfortable redundancy. Cameco’s Port Hope plant produced a record 11.2 million kgU of UF₆ in 2025, close to the company’s previously stated objective of 12 million kgU annually. Management’s Discussion and Analysis for the Year Ended December 31, 2025 – Cameco – February 2026 — verified audited corporate filing. Orano reported 10,625 tonnes of UF₆ in 2024, up from 10,060 tonnes in 2023, as the Philippe Coste–Malvési system continued its ramp-up. Annual Activity Report 2024 – Orano – March 2025 — verified corporate report. Metropolis Works has a licensed ceiling of 15,000 tonnes annually, but the disclosed primary evidence reviewed does not establish sustained production at that level after the plant’s restart, so the model uses a broad distribution materially below the licence ceiling. Draft Environmental Impact Statement, Volume 2 – US Department of Energy – March 2024 — verified official technical assessment. On the demand side, EU utilities received 13,248 tU of conversion services during 2025: Orano supplied 3,292 tU, Rosatom 3,233 tU, ConverDyn 2,825 tU, Cameco 2,791 tU, and unspecified providers 1,107 tU. EU inventories totalled 42,522 tU of natural-uranium equivalent, compared with average annual gross reactor requirements of 11,835 tU. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified Euratom market data. The US Energy Information Administration reported that American reactor operators purchased 12.7 million SWU of enrichment services in 2025, of which only 2.93 million SWU were identified as US enrichment and 9.78 million SWU as foreign enrichment. Uranium Marketing Annual Report, Table 17 – US Energy Information Administration – July 2026 — verified official market table. This combination establishes the starting condition: production has recovered, inventories are material, but the Western system remains dependent on a few conversion sites and substantial foreign enrichment.

Five-year milestone calendar

The capacity calendar is not synchronized with the geopolitical calendar. Urenco USA is installing 700,000 SWU of additional capacity between 2025 and early 2027, representing approximately a 15% increase at its New Mexico site. The facility’s existing annual capacity is approximately 4.3 million SWU, around one-third of current US requirements. Urenco USA Continues Successful Installation of US Uranium Enrichment Capacity – Urenco – June 2026 — verified corporate project update. Urenco’s Almelo programme is expected to add approximately 750,000 SWU from 2027, with another 750,000 SWU from 2030; globally, the company has committed to 2.5 million SWU of new capacity under this group of expansions. Further Capacity Expansion at Urenco’s Site in the Netherlands – Urenco – November 2025 — verified corporate announcement. These additions reduce enrichment scarcity but can increase conversion demand if enrichers operate at higher tails assays or process more natural feed. A much larger US expansion of 2.1 million SWU, involving up to 24 cascades, will not begin production until 2032 and therefore falls outside the five-year relief window. Urenco USA Plans Significant Expansion of US Uranium Enrichment Capacity – Urenco – June 2026 — verified corporate announcement. The United Kingdom targets commercial HALEU enrichment at Capenhurst for 2031, supported by 196 million pounds of government funding and approximately 400 million pounds of combined investment. Statement on Civil Nuclear Fuel Use – UK Government – February 2026 — verified government statement. DOE’s revised allocation documentation expects US commercial HALEU enrichment capacity around 2030. HALEU Allocation Process – US Department of Energy – July 2026 — verified DOE programme document. The near-term implication is severe: conventional enrichment grows before commercial-scale HALEU systems and before major greenfield conversion projects can reliably contribute.

Year	Verified or announced milestone	Capacity effect	Conversion-market implication	Forecast treatment
2026	Continued Urenco USA cascade installation; DOE enrichment awards	Additional LEU and LEU+ capability	Incremental demand for natural UF₆	High probability
2027	Completion of current 700,000 SWU US expansion; Almelo first-stage entry	Material Western enrichment increase	Potential overfeeding pressure on conversion	High probability
2028	Earliest meaningful Springfields conversion-restoration window	Possible new European conversion pathway	Could reduce three-site concentration	Medium-low probability
2029	Urenco begins construction of larger New Mexico enrichment plant	No immediate operating output	Strengthens post-2031 expectations, not near-term balance	High construction probability
2030	DOE expects commercial US HALEU capacity around this period; Almelo second-stage entry begins	HALEU and LEU expansion	Higher natural-feed and deconversion requirements	Medium probability
2031	UK targets commercial HALEU operation at Capenhurst	New advanced-fuel supply node	Strong demand for qualified conversion and deconversion	Medium probability
2032+	Initial output from new 2.1 million SWU US project	Large enrichment addition	Outside current five-year relief window	Excluded from 2026–2031 supply

H₁: recovery and normalization

H₁ proposes that present tightness is primarily the residual effect of idling, restarts and incomplete ramp-ups rather than a structural shortage. Under this hypothesis, Port Hope sustains output near 11–12 million kgU, Orano progresses toward higher utilization, Metropolis improves throughput and downtime, Springfields enters from 2028, and enrichment additions are accompanied by efficient tails management. Evidence supporting H₁ includes record Canadian production, rising French output, the return of Metropolis and tangible construction progress at Urenco. Orano’s 2025 audited accounts also reversed 252 million euros of impairment on conversion assets after improvements in medium- and long-term conversion economics, suggesting that industrial investment incentives have strengthened. Consolidated Financial Statements as of December 31, 2025 – Orano – February 2026 — verified audited statements. The hypothesis is weakened by concentration and timing. Production recovery at existing plants increases utilization but does not create geographic redundancy. A plant operating near target is more valuable but not necessarily more resilient to a long-duration outage. Springfields has not yet demonstrated newly restored natural-uranium conversion at commercial scale, and major greenfield projects remain contingent. Moreover, new enrichment capacity can consume additional converted feed rather than relieve conversion pressure. H₁ therefore explains why an acute global shortage may be avoided, but it does not explain away the security premium or the persistence of outage sensitivity. The Bayesian posterior assigned to full normalization by 2030 is 18%, up from a prior of 15% because observable production and construction milestones have improved, but still below the combined weight assigned to continued tightness scenarios.

H₂: structural Western undercapacity

H₂ argues that Western conversion remains structurally underbuilt relative to politically accessible demand. The evidence is the limited number of operating plants, the single-site US configuration, heavy reliance on Port Hope and Orano, the long development cycle for new chemical facilities, and the fact that Russian conversion still supplied 24.40% of EU deliveries in 2025. Under this hypothesis, phasing down Russian supply transfers demand into a Western system that has insufficient uncommitted margin. Existing plants can meet average requirements only if they operate reliably and if inventories absorb temporary imbalances. Euratom’s earlier analysis warned that enrichment overfeeding could produce a Western conversion shortfall reaching 10,000 tU annually after 2025 and that new conversion capacity would be needed as global demand increased. Annual Report 2022 – Euratom Supply Agency – October 2023 — verified official report. H₂ is reinforced by the capacity calendar: enrichment expansion occurs during 2026–2030, while the largest confirmed US enrichment addition starts only in 2032 and commercial HALEU arrives near 2030–2031. Conversion projects must also compete for specialized engineering, fluorination equipment, environmental permits and long-term customer commitments. The disconfirming evidence is that aggregate inventories are large, French and Canadian output is improving, and some Russian deliveries may continue during transition. H₂ does not imply inevitable reactor shutdowns; it predicts a market with chronically high utilization, long contracting horizons, low merchant liquidity and recurring inventory intervention. Its posterior probability is 29%, the largest single hypothesis weight.

H₃: geopolitical exclusion shock

H₃ treats policy rather than engineering as the dominant driver. Under this hypothesis, Russian conversion and enrichment become inaccessible faster than Western replacements enter. The United States has already prohibited Russian LEU imports except under waiver, while the EU roadmap seeks gradual restrictions on Russian uranium, enriched uranium and other nuclear materials and intends to constrain new Euratom co-signed contracts with Russian suppliers. Environmental Impact Statement for Department of Energy Activities in Support of Commercial Production of HALEU – US Department of Energy – October 2024 — verified DOE assessment. Roadmap towards Ending Russian Energy Imports, Action 5 – Council of the European Union – May 2025 — verified Council document. The shock channel operates through contract expiry, waiver termination, financial de-risking, insurance and export controls rather than destruction of Russian plant capacity. H₃ gains support from continued reliance on Russian services: EU utilities received 3,233 tU of Rosatom conversion and 2,735 tSW of Russian enrichment in 2025. Replacing those flows requires both physical capacity and correctly timed contracts. H₃ is weakened by the gradual design of EU policy, existing inventories and the likelihood that waivers or transitional arrangements would be used to prevent reload failures. Its posterior probability is 17% for a material but manageable exclusion shock and approximately 6% for a shock severe enough to threaten multiple reactor reloads within one year. The two probabilities should not be added mechanically because the severe outcome is nested inside the broader shock hypothesis.

H₄: enrichment-driven overfeeding

H₄ argues that conversion tightness will be driven less by new reactors than by enrichment economics. An enricher can substitute separative work for natural feed by lowering the tails assay, or substitute natural feed and conversion for scarce separative work by raising it. The 2025 EU fleet used an average tails assay of 0.21%, while enrichment-service deliveries rose 17% over 2024. EU-origin enrichment increased to 8,844 tSW, or 72.91% of deliveries, while Russian enrichment supplied 2,735 tSW. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified Euratom data. At a fixed product assay, an increase in tails from 0.20% to 0.25% raises natural-feed demand materially, even though final enriched-product mass remains unchanged. Urenco USA’s existing 4.3 million SWU capacity, its 700,000 SWU addition by early 2027 and Almelo’s staged expansion create more Western separative work, which should eventually permit lower tails and reduce natural-feed intensity. The transitional period is uncertain because existing contracts, centrifuge availability and product-assay requirements determine operating choices. H₄ also interacts with Russian disengagement: if Russian SWU disappears before Western cascades replace it, Western enrichers may temporarily overfeed to maximize output. The hypothesis is supported by Euratom’s explicit warning that overfeeding could produce a conversion gap; it is weakened by the rapid progress of current Urenco expansions. Its posterior probability is 16% as the dominant cause of tightness, but overfeeding appears as a contributing factor in most other scenarios.

H₅: HALEU demand acceleration

H₅ proposes that advanced reactors create the principal conversion shock before 2031. DOE has evaluated acquisition of up to 290 metric tonnes of HALEU and estimated future US commercial requirements of 50 tonnes annually by 2035, rising toward 500 tonnes annually by 2050. It also stated that approximately 90% of the separative work needed to enrich natural uranium from 0.711% to 19.75% is consumed before the material reaches 10% assay. Environmental Impact Statement for Department of Energy Activities in Support of Commercial Production of HALEU – US Department of Energy – October 2024 — verified DOE assessment. DOE’s technical assessment estimated that producing 50 tonnes of HALEU annually could require roughly 2,600 tonnes of yellowcake per year, depending on process assumptions. Technical Report in Support of the HALEU Environmental Impact Statement – US Department of Energy – November 2023 — verified technical report. H₅ is nevertheless constrained by commercial reality: reactor construction delays, licensing, deconversion and fuel-fabrication readiness may postpone demand. DOE’s July 2026 allocation process expects commercial US HALEU enrichment only around 2030, and the UK targets 2031. Consequently, HALEU is unlikely to dominate conversion demand in 2026–2027. It becomes a meaningful upper-tail risk from 2029 onward, particularly if several first cores require delivery within a narrow period. H₅’s posterior probability as the dominant five-year bottleneck is 9%, but its contribution to the 2030–2031 stress tail is much larger than that single weight suggests.

H₆: inventory-mediated stability

H₆ holds that inventories, material exchanges and staggered contracting will prevent conversion tightness from becoming a fuel-delivery crisis. EU utilities ended 2025 with 42,522 tU of natural-uranium-equivalent inventory, sufficient in aggregate for more than three average annual reloads, and all utilities reportedly held at least one reload. In 2025 they purchased 14,678 tU of natural uranium while fresh fuel used 11,729 tU of natural uranium feed, reflecting continued stock-building after Russia’s invasion of Ukraine. European Union Market in 2025 – Euratom Supply Agency – August 2026 — verified Euratom data. In the United States, reactor operators had maximum contracted uranium deliveries of 174 million pounds U₃O₈ equivalent for 2026–2035 but also reported 186 million pounds of unfilled requirements through 2035. Uranium Marketing Annual Report – US Energy Information Administration – July 2026 — verified official market report. Inventories provide a genuine resilience layer, but their aggregate mass exaggerates flexibility. Material may be natural uranium rather than UF₆, may be stored outside the consuming jurisdiction, may already be committed, or may be incompatible with the relevant fabrication schedule. H₆ therefore predicts continuity with high working-capital cost and unequal exposure rather than effortless stability. Its posterior probability as the dominant system behaviour is 11%, while inventory buffering appears as an essential mitigating mechanism in every base and downside scenario.

Bayesian update matrix

The Bayesian assessment began with priors derived from market structure rather than price behaviour: recovery and normalization 15%; structural undercapacity 27%; geopolitical exclusion shock 16%; enrichment-driven overfeeding 15%; HALEU acceleration 12%; and inventory-mediated stability 15%. Evidence available through August 2026 alters those weights. Record Port Hope output and Urenco construction progress raise H₁. Continued Russian conversion deliveries, limited new conversion construction and delayed major capacity raise H₂. Gradual EU restrictions and US waivers prevent H₃ from dominating. The EU tails assay near 0.21% and the arrival of new SWU slightly constrain H₄. DOE and UK HALEU milestones confirm future demand but push large-scale operation toward 2030–2031, reducing H₅ inside the early forecast years. Large European inventories support H₆, but heterogeneous forms and locations limit its explanatory power. The normalized posteriors are therefore H₁ 18%, H₂ 29%, H₃ 17%, H₄ 16%, H₅ 9% and H₆ 11%. These values describe which mechanism is most likely to dominate, not mutually exclusive physical worlds. In the actual market, H₂ may define structure, H₃ trigger the shock, H₄ amplify feed demand and H₆ prevent missed reloads. Bayesian updating should therefore be repeated against milestone evidence, Ei,t, using posterior odds Oi,t = Oi,t−1 × L(Ei,t|Hi) ÷ L(Ei,t|¬Hi). The formula is shown only to make the updating logic transparent; likelihood ratios must be supported by observed milestones rather than subjective narrative.

Hypothesis	Prior	Evidence increasing weight	Evidence decreasing weight	Posterior
H₁ Recovery and normalization	15%	Port Hope record; Orano ramp; Urenco construction	Lack of new conversion redundancy	18%
H₂ Structural Western undercapacity	27%	Few plants; Russian displacement; long project timelines	Inventory and incumbent-output improvement	29%
H₃ Geopolitical exclusion shock	16%	US prohibition; EU contract restrictions	Waivers and gradual transition	17%
H₄ Enrichment overfeeding	15%	Russian SWU exposure; conversion sensitivity to tails	Western SWU additions	16%
H₅ HALEU acceleration	12%	DOE 290-tonne envelope; 2030–2031 facilities	Reactor and fuel-chain schedule risk	9%
H₆ Inventory-mediated stability	15%	EU stocks above three reloads	Location, form and allocation constraints	11%

Monte Carlo architecture

The Monte Carlo model uses 200,000 annual trials for each year from 2026 through 2031. It does not claim proprietary knowledge of operator books. Port Hope is represented by a triangular effective-output distribution bounded below the stated 12-million-kgU target and centred on disclosed 2025 production. Orano is centred above disclosed 2024 production with gradual ramp potential but capped at a higher industrial bound. Metropolis uses the broadest distribution because sustained post-restart output is not publicly disclosed; the licence ceiling of 15,000 tonnes is treated as an upper bound rather than an expected value. A combined “other allied capacity” term covers smaller, secondary or indirectly available sources without assigning unsupported plant-level precision. Springfields receives zero firm output before 2028 and probability-weighted entry thereafter. Each major plant has an annual 4.5% probability of a serious disruption causing a 15–50% output loss, plus ordinary operating variability. Baseline Western demand begins around 35,200 tU, grows approximately 1% annually, and includes uncertainty reflecting reload timing. Additional demand from declining Russian accessibility grows through the period. HALEU demand is modelled as a zero-inflated distribution: most early trials contain no material increment, while the probability and size of demand rise toward 2031. This structure avoids treating the DOE maximum as guaranteed consumption. The primary outputs are the probability that annual demand exceeds effective contestable supply, the probability that the deficit exceeds 10% of demand, the median balance and the 90th-percentile adverse gap. Inventory is excluded from productive supply and analysed separately as a mitigation layer, preventing stock draw from being mislabelled as recurring capacity.

Monte Carlo variable	2026 treatment	Evolution through 2031	Distribution logic
Port Hope effective output	Centred near 11.2 million kgU	Broadly stable	Triangular plus outage loss
Orano effective output	Centred moderately above 2024 production	Gradual ramp	Triangular plus outage loss
Metropolis output	Wide range below 15,000 t licence ceiling	Slow improvement	Broad triangular plus outage loss
Springfields	Zero firm baseline	Increasing entry probability from 2028	Bernoulli entry plus triangular output
Other allied capacity	3.5–8.0 ktU range	Gradual increase	Triangular
Baseline demand	Around 35.2 ktU	Approximately 1% annual growth	Normal
Russian replacement demand	Limited in 2026	Increases with restrictions	Triangular
HALEU increment	Usually zero in early years	Rising probability and magnitude	Zero-inflated triangular
Serious plant outage	4.5% per major plant-year	Constant probability	Bernoulli severity shock

Monte Carlo results

The model produces a 31.5% probability of an annual contestable-capacity deficit in 2026, rising to 36.7% in 2027 as Russian replacement demand grows before all allied projects mature. The probability remains 36.2% in 2028 because tentative Springfields entry offsets only part of the demand increase, then falls to 34.1% in 2029 and 32.8% in 2030 as probability-weighted new capacity improves the balance. It rises slightly to 34.2% in 2031 because commercial HALEU demand and Russian displacement begin to absorb the added capacity. A deficit exceeding 10% of annual contestable demand remains much less likely: 4.6% in 2026, 6.1% in 2027, 6.5% in 2028, 6.7% in 2029, 6.7% in 2030 and 7.1% in 2031. Median supply remains above median demand in every year, but the margin is narrow: approximately 1.4 ktU in 2026, 1.0 ktU in 2027, 1.2 ktU in 2028, 1.5 ktU in 2029, 1.7 ktU in 2030 and 1.6 ktU in 2031. The 90th-percentile adverse gap rises from approximately 2.5 ktU in 2026 to 3.5 ktU by 2031. These results support a precise conclusion: the base system is more likely than not to meet annual demand, but roughly one year in three produces a deficit requiring inventory draw, delivery rescheduling, lower tails assays or additional non-Western supply. The risk of a deficit large enough to challenge several reload programmes simultaneously is much lower but rises gradually because demand diversification and HALEU expansion counterbalance capacity additions.

Model output	2026	2027	2028	2029	2030	2031
Probability demand exceeds contestable supply	31.5%	36.7%	36.2%	34.1%	32.8%	34.2%
Probability deficit exceeds 10% of demand	4.6%	6.1%	6.5%	6.7%	6.7%	7.1%
Median supply	38.0 ktU	38.5 ktU	39.6 ktU	41.0 ktU	42.2 ktU	43.2 ktU
Median demand	36.5 ktU	37.5 ktU	38.4 ktU	39.4 ktU	40.5 ktU	41.5 ktU
Median balance	+1.4 ktU	+1.0 ktU	+1.2 ktU	+1.5 ktU	+1.7 ktU	+1.6 ktU
90th-percentile adverse gap	2.5 ktU	3.0 ktU	3.2 ktU	3.2 ktU	3.3 ktU	3.5 ktU

Values are outputs of the disclosed analytical model, not forecasts published by governments, Euratom or fuel-cycle companies. Totals are rounded, so displayed supply minus demand may differ slightly from the displayed median balance.

Scenario A: managed tightness

The base scenario receives a 54% probability and describes a system that avoids generalized failure through high incumbent output, inventory use, Russian transition arrangements and incremental Western additions. Port Hope remains near 10.5–11.5 million kgU; Orano sustains or modestly improves output; Metropolis operates reliably enough to support US demand; Springfields contributes limited volumes late in the decade; and Urenco’s additions reduce enrichment scarcity. Russian conversion and enrichment decline gradually rather than disappearing suddenly. HALEU demand remains programme-led until 2029, then expands without reaching the full DOE environmental envelope. Conversion prices and contract tenors remain elevated because utilities pay for security and suppliers preserve capacity for long-term customers. Inventory draws occur in one or two years but are later rebuilt. The scenario’s defining characteristic is the absence of a clean return to liquidity: physical supply broadly covers demand, yet new buyers struggle to obtain flexible deliveries. Government programmes increasingly influence allocation, especially in the United States and United Kingdom. The risk is not average insufficiency but sequencing. A maintenance outage coinciding with a Russian waiver expiry or an advanced-reactor first-core order produces temporary deficits. Utilities with three or more reloads remain protected; utilities with less diversified material suffer higher replacement costs. By 2031, the Western system is larger but still concentrated, and major US enrichment output scheduled from 2032 arrives just outside the forecast period. Managed tightness is therefore stable only if inventories and policy flexibility remain available.

Scenario B: allied expansion succeeds

The upside scenario receives a 25% probability. It requires Port Hope and Orano to sustain high output, Metropolis to demonstrate reliable throughput, Springfields to enter on or near the earliest credible schedule and Western enrichment projects to remain on budget. Urenco USA completes the 700,000 SWU addition by early 2027; Almelo begins delivering its first 750,000 SWU from 2027 and progresses toward the second stage from 2030; LEU+ becomes commercially transportable to fabricators; and DOE procurement converts framework awards into bankable task orders. DOE announced 2.7 billion dollars in enrichment support over ten years in January 2026, with three companies receiving contracts valued at up to 900 million dollars each for LEU and HALEU services. US Department of Energy Awards 2.7 Billion Dollars to Restore American Uranium Enrichment – US Department of Energy – January 2026 — verified DOE announcement. Under the upside scenario, these contracts produce private co-investment rather than merely reserving future options. HALEU reactor schedules remain staggered, preventing first-core demand from concentrating in one year. Russia’s EU share declines in step with Western entry rather than ahead of it. By 2030, contestable supply margin expands, contract premia stabilize and inventories stop increasing. The scenario still does not create abundant spot capacity because nuclear-fuel investments require long-term commitments. Its principal failure point is conversion: enrichment additions can outpace natural-UF₆ supply unless Springfields or another converter becomes operational.

Scenario C: synchronized outage and geopolitical cutoff

The severe disruption scenario receives a 13% probability, while the probability of its most extreme form—multiple reload impairments—is assessed near 5–7% depending on year. It combines four shocks: a major outage lasting three to six months at Port Hope, Orano or Metropolis; accelerated reduction of Russian supply; delay of Springfields or other projects; and elevated enrichment overfeeding. A six-month outage at a 12,000-tU converter removes roughly 6,000 tU of annualized output before restart inefficiency. If this coincides with replacement of Rosatom’s 3,233 tU of EU conversion deliveries, the gross stress can approach 9,000 tU, close to Euratom’s earlier high-stress estimate. Inventories would prevent immediate system failure, but government and utility responses would remove material from circulation. Likely measures include strategic UF₆ releases, enrichment rescheduling, lower tails assays, bilateral material loans, emergency waiver extensions and prioritization of conventional reactor reloads over demonstration HALEU. The market would cease functioning through ordinary price discovery and move toward administrative allocation. Cyber-physical events are included as outage initiators but not assigned a separate numerical probability because public incident data are insufficient. Conversion plants contain distributed control systems, fluorine production, ventilation, criticality monitoring and hazardous-chemical controls; a cyber event need not manipulate nuclear material directly to cause a precautionary shutdown. Recovery duration, not initial intrusion severity, would determine market impact.

Scenario D: HALEU schedule compression

The HALEU acceleration scenario receives an 8% probability as a distinct dominant pathway, overlapping partly with H₅. It arises if several advanced-reactor projects move from policy aspiration to simultaneous first-core procurement between 2029 and 2031. DOE’s contemplated maximum of 290 tonnes of HALEU is not assumed to materialize, but even a fraction creates disproportionately large upstream requirements. At a programme-level estimate of approximately 52 tonnes of yellowcake per tonne of HALEU product, 10 tonnes of additional HALEU can imply roughly 520 tonnes of upstream natural-uranium feed before precise plant-specific adjustments. The pressure is not limited to natural conversion. Material above 10% enrichment requires Category II security and licensing in the United States, and deconversion, storage and fabrication must meet the same category. DOE stated that domestic commercial capability was insufficient and that the initial American Centrifuge Plant capacity was 900 kg per year at the time of the 2024 assessment. Environmental Impact Statement for Department of Energy Activities in Support of Commercial Production of HALEU – US Department of Energy – October 2024 — verified DOE assessment. In a compressed schedule, the government must choose whether to allocate scarce enriched material to demonstrations, commercial first cores or strategic inventories. Conventional LEU reactors would probably retain priority because they supply existing electricity, pushing advanced projects into delay rather than causing widespread reactor outages. The economic consequence would nevertheless be material: reactor developers would face carrying costs, financing resets and possible redesign toward LEU+.

Annual forecast and trigger thresholds

For 2026, the central risk is incumbent-plant reliability; the system has limited time to compensate for an extended outage. For 2027, the risk shifts toward enrichment-conversion coupling as new SWU comes online and Russian replacement demand rises. For 2028, project execution becomes decisive: the absence of a credible Springfields commissioning path by year-end would lower the probability of meaningful conversion relief before 2030. For 2029, HALEU demand must be classified into firm, probable and aspirational quantities. For 2030, commercial US HALEU expectations meet the reality of Category II licensing, deconversion and fabrication. For 2031, UK HALEU operations and advanced-reactor first-core schedules determine whether the market enters the next decade with adequate infrastructure or a new bottleneck. A red trigger should be activated if any major Western converter loses more than 90 production days, if Rosatom’s EU conversion share falls by more than 10 percentage points in a single year without equivalent replacement, if effective inventories fall below two reloads, or if firm HALEU commitments exceed 20 tonnes annually before new deconversion capacity operates. An amber trigger applies when converter utilization exceeds 90%, contract lead times exceed five years, tails assays rise above 0.25%, or more than 25% of a jurisdiction’s future conversion needs remain uncontracted inside a three-year window.

Year	Base assessment	Principal upside indicator	Principal downside indicator	Bayesian revision trigger
2026	Tight but manageable	Stable output at all three major Western plants	More than 90 outage days at one plant	Raise disruption posterior by 8–12 points
2027	Peak near-term contracting pressure	Urenco additions reduce SWU scarcity	Overfeeding increases UF₆ demand	Raise H₄ if average tails exceed 0.25%
2028	Conversion investment decision point	Springfields first qualified production	Licensing or construction slippage	Raise H₂ by 5–8 points
2029	Advanced-fuel commitments become material	Staggered reactor schedules	Multiple first cores converge	Raise H₅ if firm HALEU exceeds 10 t annually
2030	Potential relief, still fragile	US commercial HALEU and Almelo expansion	Deconversion or fabrication not ready	Shift probability from upside to managed tightness
2031	Transition into a larger but more complex system	UK HALEU operates; inventories stable	Demand absorbs all new margin	Raise structural-deficit probability

Strategic judgement

The five-year forecast does not support the claim that uranium conversion will inevitably halt large numbers of reactors. It supports a narrower and more defensible judgement: between 2026 and 2031, the Western fuel cycle will operate with sufficient average capacity in most years but insufficient redundancy to absorb every plausible combination of plant outage, Russian disengagement, enrichment overfeeding and HALEU acceleration without drawing inventories or using policy intervention. The most likely path is managed tightness, not catastrophic shortage and not full normalization. The Bayesian distribution places 54% on managed tightness, 25% on successful allied expansion, 13% on synchronized geopolitical and outage disruption, and 8% on HALEU-driven schedule compression as the dominant stress. Monte Carlo results show an annual contestable-deficit probability around one-third, but severe deficits remain below 8% in each modelled year. This gap between ordinary deficit probability and severe-deficit probability is the central intelligence finding: inventories, contract flexibility and lower tails assays are likely to prevent most annual imbalances from becoming fuel failures. Those defences are costly and finite. Every inventory draw reduces future resilience; every waiver delays investment certainty; every enrichment adjustment consumes more SWU; and every government allocation displaces another buyer. The window for reducing systemic risk is 2026–2028, because capacity commissioned after 2030 cannot protect earlier contract expiries or outages. Investment should therefore prioritize conversion redundancy, qualified cylinder and transport capacity, HALEU deconversion, cross-compatible fabrication and inventories measured by usable reactor reload rather than natural-uranium equivalent alone.

---

Copyright of debuglies.com – Even partial reproduction of the contents is not permitted without prior authorization – Reproduction reserved