Uranium Supply Chain

How enrichment concentration, the five-stage mine-to-reactor pathway, and the legacy of warhead downblending create a commodity system where geopolitics determines supply more than geology does.

Introduction

A uranium supply chain describes how physical products — yellowcake concentrate, uranium hexafluoride gas, enriched uranium oxide pellets, and finished fuel assemblies — move from mines through a sequence of chemical and isotopic transformations into the cores of nuclear reactors. Each transformation crosses organizational, national, and regulatory boundaries. The chain is unusual among commodity systems because every stage is constrained not only by economics and engineering but by international treaty, weapons-proliferation controls, and sovereign policy decisions that operate on logic entirely separate from market signals.

What makes the uranium supply chain structurally distinct from other mineral supply chains is the interaction of three root constraints. Enrichment — the process of increasing the concentration of fissile uranium-235 from its natural 0.7 percent to the three to five percent required for power reactors — is performed by a small number of state-affiliated facilities, and the centrifuge technology that performs it can also produce weapons-grade material. The mine-to-reactor pathway passes through five discrete stages, each with its own qualification barriers, regulatory requirements, and a remarkably small number of participants. And for twenty years, from 1993 to 2013, roughly half of all fuel loaded into American reactors came not from mines but from dismantled Russian nuclear warheads — a secondary supply source that suppressed uranium prices, discouraged mining investment, and created a structural dependency that only became visible when it ended.

These three forces produce a supply chain where the binding constraints are political and institutional rather than geological. The earth contains abundant uranium. The problem is not finding it or extracting it. The problem is transforming it through a sequence of bottlenecks where the number of qualified participants at each stage can be counted on one hand, where the technology is controlled by treaty, and where decades of secondary supply created habits and assumptions that primary mining was sufficient when it was not.

Root Constraints

Why is uranium enrichment the most tightly controlled industrial process?

Uranium enrichment is the process of increasing the proportion of uranium-235 — the fissile isotope that sustains a chain reaction — from its natural abundance of 0.711 percent to the three to five percent concentration used in light water reactor fuel. The physics of isotope separation requires processing enormous quantities of uranium hexafluoride gas through thousands of centrifuges arranged in cascades. The same centrifuge technology that produces low-enriched uranium for power reactors can, with continued processing, produce highly enriched uranium suitable for weapons. This dual-use reality places enrichment under the tightest international controls of any industrial process.

Global enrichment capacity is concentrated in four primary operators. Russia’s Rosatom, through its subsidiary TENEX and the enrichment arm Techsnabexport, controls approximately 40 percent of global enrichment capacity. The European consortium Urenco — jointly owned by the United Kingdom, the Netherlands, and Germany — operates centrifuge plants in those three countries plus the United States and accounts for roughly 30 percent. France’s Orano operates the Georges Besse II plant. China’s CNNC has built domestic enrichment capacity primarily for its own growing reactor fleet. Together, these four entities account for essentially all commercial enrichment on earth.

This concentration is not a market outcome. It is a consequence of the Nuclear Non-Proliferation Treaty and associated export control regimes, which restrict the transfer of enrichment technology. Building new centrifuge capacity requires not just capital and engineering capability but sovereign permission, international safeguards agreements, IAEA inspection protocols, and political alignment with existing nuclear powers. Countries that did not develop enrichment technology during the Cold War face structural barriers to acquiring it that cannot be overcome by commercial investment alone. The barrier is legal and geopolitical, not technical or financial.

The consequence is a structural dependency that became starkly visible after 2022. Western reactor operators — particularly in the United States, which operates the world’s largest fleet of 93 reactors — had relied on Russian enrichment services for a significant share of their fuel supply. When geopolitical relations deteriorated, operators discovered that alternative enrichment capacity could not be scaled on any commercially relevant timeline. Urenco and Orano were operating near capacity. New centrifuge facilities require five to ten years to build, certify, and bring to full production. The system had optimized for cost over decades and could not reorganize when the political landscape shifted.

Why must uranium pass through five sequential stages before it can fuel a reactor?

Uranium must pass through five discrete transformation stages before it can produce heat in a reactor core. Each stage has its own chemistry, its own regulatory framework, its own qualification requirements, and a remarkably small number of global participants. The stages are sequential and cannot be skipped or reordered. A disruption at any single stage halts the entire downstream chain.

Stage one is mining. Uranium ore is extracted through conventional open-pit or underground mining, or increasingly through in-situ leaching — a process where a solution is pumped underground to dissolve uranium from the ore body and bring it to the surface. Kazakhstan dominates global uranium mining, producing roughly 45 percent of world output, primarily through in-situ leaching operations run by the state-owned enterprise Kazatomprom. Canada, Namibia, Australia, and Uzbekistan account for most of the remainder. Mining is the most commercially accessible stage of the fuel cycle, but new mines still require 10 to 15 years from discovery to first production, owing to permitting, environmental review, and the capital-intensive nature of development.

Stage two is milling. Mined ore or leach solutions are processed into uranium oxide concentrate — the yellow powder known as yellowcake, with the chemical designation U3O8. Milling is typically co-located with mining operations and is the least constrained stage of the fuel cycle. Yellowcake is a stable, transportable commodity that can be stored and shipped under standard hazardous materials protocols.

Stage three is conversion. Yellowcake must be chemically transformed into uranium hexafluoride — UF6 — the only uranium compound that exists as a gas at practical temperatures and pressures, which is the required feedstock for centrifuge enrichment. This step is an overlooked bottleneck. Only four to five conversion facilities operate globally: Orano’s Malvesi and Pierrelatte plants in France, Cameco’s facility in Port Hope, Canada, ConverDyn’s Metropolis plant in Illinois, and Rosatom’s facilities in Russia. When any of these facilities experiences an outage — as occurred at the Metropolis plant during extended shutdowns — the constraint propagates immediately through the downstream chain because there is no substitute process and minimal surplus capacity.

Stage four is enrichment, described above. Stage five is fuel fabrication. Enriched UF6 is converted to uranium dioxide powder, pressed into ceramic pellets, sintered at high temperature, ground to precise dimensions, loaded into zirconium alloy cladding tubes, and assembled into fuel bundles engineered to the exact specifications of a particular reactor design. Fuel assemblies are not interchangeable. A fuel assembly designed for a Westinghouse AP1000 cannot be loaded into a French EPR or a Russian VVER. Fabrication plants must be qualified for each reactor type they serve, and qualification requires years of testing and regulatory approval.

The consequence of this five-stage structure is that the uranium supply chain has serial dependencies that multiply rather than average vulnerability. If any single stage is disrupted — a mine closure, a conversion plant outage, an enrichment supply restriction, a fabrication qualification failure — the entire downstream path is blocked. Redundancy at one stage does not compensate for concentration at another. The chain is as constrained as its most constrained link, and several links are severely constrained simultaneously.

How did dismantled nuclear warheads come to supply half of America's reactor fuel?

From 1993 to 2013, the Megatons to Megawatts program — formally the HEU Purchase Agreement between the United States and Russia — converted 500 metric tons of Russian weapons-grade highly enriched uranium into low-enriched uranium suitable for reactor fuel. This material, extracted from approximately 20,000 dismantled nuclear warheads, supplied roughly half of all fuel loaded into American nuclear reactors during that period. It was the largest disarmament program in history and simultaneously the largest single source of reactor fuel on earth.

The program’s effect on the uranium market was profound and structural. By providing a massive, reliable, below-market-cost source of enriched uranium, it suppressed uranium prices for two decades. The spot price of uranium remained below $20 per pound for most of the 1990s and early 2000s — well below the cost of production at most mines. The predictable consequence was systematic underinvestment in primary mining. Exploration budgets collapsed. Mines were mothballed. Skilled workforces dispersed. The uranium mining industry contracted to a fraction of the capacity that would have been needed if reactors had relied entirely on freshly mined material.

When the program ended in 2013, the secondary supply it had provided did not disappear overnight — commercial inventories built during the program provided a buffer — but the structural gap between reactor demand and primary mine supply became progressively more visible. The Fukushima accident in 2011 temporarily masked this gap by reducing demand as Japan shut down its reactor fleet. But as reactors restarted and new construction proceeded in China, the gap between primary mine supply and reactor requirements widened.

The warhead downblending program created what might be called a structural illusion: the appearance that the uranium market was adequately supplied, when in reality the supply was coming from a non-renewable, non-repeatable source. Once the warheads were consumed, the mining industry needed to replace not just the incremental growth in demand but the entire volume that secondary supply had been providing. This replacement has not yet fully occurred. The mine-to-reactor timeline of 10 to 15 years means that investment decisions made after 2013 are only now beginning to produce new supply.

How the Constraints Shape the System

Why is uranium mined in different countries from where it is processed into fuel?

The uranium supply chain exhibits a striking geographic separation between where the raw material is extracted and where it is transformed into usable fuel. Kazakhstan produces roughly 45 percent of the world’s mined uranium, but it has no commercial enrichment capacity. Australia holds the world’s largest uranium reserves but does not mine at scale commensurate with its geological endowment and performs no conversion, enrichment, or fabrication domestically. Canada is a major miner and hosts a conversion facility but relies on foreign enrichment.

Enrichment capacity, by contrast, is concentrated in Russia, Western Europe, and increasingly China — none of which are major uranium mining countries. This means that mined uranium must physically cross multiple national borders and pass through multiple corporate entities before it becomes reactor fuel. The material flows from Central Asian and African mines to conversion plants in France, Canada, or the United States, then to enrichment facilities in Europe or Russia, then to fabrication plants that may be in yet another country, before finally arriving at a reactor.

Each border crossing carries regulatory requirements. Each transfer between corporate entities involves contractual negotiations. The geographic mismatch between mining and processing is not an inefficiency that optimization can resolve — it is a structural feature of how enrichment technology was historically developed and how treaty controls prevent its diffusion. The supply chain’s physical geography reflects its political geography.

Why does uranium trade through long-term contracts rather than a spot market?

Unlike oil, copper, or gold, uranium trades primarily through long-term bilateral contracts rather than on liquid spot markets. The spot market — where uranium can be purchased for near-term delivery — typically accounts for only 10 to 20 percent of total transaction volume. The remaining 80 to 90 percent moves under multi-year contracts negotiated directly between mining companies and utilities, with terms that can extend five to fifteen years.

This contract structure reflects the physical realities of the fuel cycle. A utility planning a reactor refueling outage two years hence needs assurance that enriched fuel assemblies will be delivered on schedule. The fabrication process alone takes 12 to 18 months. Enrichment must be arranged before that. Conversion before that. A spot purchase of yellowcake cannot be transformed into a loaded fuel assembly on any timeline shorter than roughly two years. Utilities therefore secure supply years in advance through contracts that specify not just uranium quantities but conversion, enrichment, and fabrication services.

The thin spot market has consequences for price discovery and price volatility. Small changes in spot demand or supply — a single utility entering the market unexpectedly, a single mine experiencing a production shortfall — can move prices dramatically because the spot market lacks the depth to absorb even modest shocks. The spot price of uranium has historically experienced swings of 100 percent or more within single years, driven by movements that would be trivially absorbed in more liquid commodity markets. The spot price is visible and widely quoted, but it represents a fraction of actual transactions and can diverge significantly from the contract prices at which most uranium actually changes hands.

Why can't a reactor's fuel assemblies be swapped between different reactor designs?

Fuel assemblies are engineered to precise specifications that vary by reactor design. A pressurized water reactor built to Westinghouse specifications uses fuel assemblies with different geometry, enrichment profiles, and cladding materials than a French EPR, a Russian VVER-1000, or a Canadian CANDU. The ceramic pellet dimensions, the fuel rod spacing, the enrichment gradients within an assembly, and the structural components that hold the assembly together are all design-specific. A fuel fabrication plant qualified for one reactor type cannot produce assemblies for another without a separate qualification program that may take years.

This specificity creates a structural lock-in. An operator of VVER reactors — the Russian-designed pressurized water reactors that operate across Eastern Europe, China, India, and elsewhere — has historically relied on Russian fuel fabrication through TVEL, a Rosatom subsidiary. Switching to a Western fuel supplier requires not just a commercial decision but a multi-year engineering qualification program to demonstrate that alternative fuel assemblies perform safely in a reactor designed for different fuel. Westinghouse has developed VVER-compatible fuel assemblies, and several Eastern European operators have undertaken the qualification process, but it is measured in years and millions of dollars, not in purchase orders.

The consequence is that reactor design choices made decades ago continue to determine fuel supply dependencies today. A country that built VVER reactors in the 1980s remains structurally connected to Russian fuel supply chains in the 2020s, regardless of how geopolitical relationships have changed. The switching cost is not a price premium. It is a multi-year engineering and regulatory process that cannot be accelerated by spending more money.

Why do utilities hold years of uranium inventory as a buffer?

Utilities and fuel cycle intermediaries maintain uranium inventories — typically two to three years of forward fuel requirements — as a buffer against supply disruptions. These inventories exist at various stages of the fuel cycle: as yellowcake in storage, as UF6 awaiting enrichment, as enriched UF6 awaiting fabrication, and as fabricated fuel assemblies in on-site storage at reactors. The total commercial inventory globally is estimated at several years of reactor requirements.

This buffer provides resilience against short-term disruptions but creates its own structural effects. When inventories are high, utilities feel less urgency to contract for new supply, which suppresses prices and discourages mining investment. When inventories draw down — as they have been doing since the mid-2010s as secondary supply ended and utilities consumed strategic reserves — the market discovers that rebuilding inventory requires purchasing from a mining industry that contracted during the years of oversupply. The inventory cycle amplifies rather than dampens the underlying supply-demand imbalance, because the decision to draw down inventory and the decision to invest in new mines operate on entirely different timescales.

What Disruptions Have Revealed

The post-2022 geopolitical disruption involving Russia exposed the depth of Western dependence on Russian nuclear fuel services. The United States imported roughly 25 percent of its enriched uranium from Russia. European utilities sourced significant volumes of both enrichment services and fabricated fuel from Rosatom subsidiaries. When political pressure mounted to reduce this dependence, operators discovered that the structural constraints described above — enrichment concentration, long qualification timelines, reactor-specific fuel — meant that reducing Russian dependence was a multi-year process, not a procurement decision.

The U.S. Congress eventually passed legislation restricting Russian uranium imports, with a phased implementation recognizing that immediate cessation was physically impossible given the supply chain’s structure. Utilities needed time to secure alternative enrichment contracts, qualify new fuel fabrication sources, and in some cases modify reactor operating parameters to accommodate fuel from different suppliers. The disruption did not reveal a market failure. It revealed the structural reality of a system where substitution timelines are measured in years and where the number of alternative suppliers at each stage is very small.

Kazakhstan’s uranium production, while dominant in volume, has demonstrated its own fragility. Kazatomprom — the state-owned enterprise that controls Kazakh uranium mining — has repeatedly revised production guidance downward in recent years, citing sulfuric acid shortages needed for in-situ leaching, construction delays at new wellfields, and logistical constraints in a landlocked country that must transport yellowcake through Russia or across the Caspian Sea to reach Western markets. The world’s largest uranium producer depends on inputs and transit routes that are themselves subject to disruption, creating nested vulnerabilities that compound through the supply chain.

Japan’s post-Fukushima reactor shutdowns in 2011 produced a different kind of structural revelation. When Japan took its entire reactor fleet offline, it released both uranium inventory and contracted supply into a market that was simultaneously losing its largest demand source. Prices collapsed and remained depressed for nearly a decade, further discouraging mining investment. The disruption demonstrated how a single country’s policy decision could reshape the global uranium market’s price structure for years, because the market is small enough — roughly 180 million pounds per year of reactor demand — that the addition or subtraction of a single major consumer visibly changes the supply-demand balance.

The Emerging Demand Layer

Several structural developments are adding new demand to a supply chain already struggling to replace secondary supply and recover from two decades of underinvestment. China has the world’s most ambitious reactor construction program, with over 20 reactors under construction and plans for a fleet exceeding 150 units by 2035. Each gigawatt of new nuclear capacity requires roughly 200 tons of natural uranium per year in fuel, plus the associated conversion, enrichment, and fabrication services. China is simultaneously building domestic enrichment capacity, but the pace of reactor construction outstrips the pace of fuel cycle expansion.

Small modular reactors — a category of reactor designs with electrical output below 300 megawatts — are moving from concept to deployment in several countries. Some SMR designs require high-assay low-enriched uranium (HALEU) — uranium enriched to between 5 and 20 percent U-235, above the level produced by existing commercial enrichment facilities. HALEU production requires either new enrichment cascades or modification of existing ones, and currently only Russia produces HALEU at commercial scale. A new reactor technology intended to reduce nuclear energy’s dependence on large-scale infrastructure would, paradoxically, increase dependence on the most constrained segment of the existing fuel cycle.

Reactor life extensions add a subtler form of demand. Existing reactors in the United States and Europe are receiving license extensions from 40 to 60 years, and some operators are pursuing extensions to 80 years. Each decade of extended operation requires continued fuel supply from a chain that was sized for a fleet expected to retire on its original schedule. Life extensions do not create headlines, but they lock in fuel demand for decades beyond original planning horizons.

Epistemic Boundaries

This analysis describes the structural constraints and dependencies visible in the uranium supply chain. It cannot predict how geopolitical relationships will evolve, whether enrichment sanctions will be tightened or loosened, or how quickly new mining capacity will come online. The uranium market’s opacity — dominated by private bilateral contracts with limited public price discovery — means that the actual supply-demand balance is estimated rather than precisely measured. Inventory levels held by utilities are commercially sensitive and not fully reported. The analysis describes the system’s architecture and where constraints bind, not the specific outcomes those constraints will produce.

What This Reveals

• Enrichment is the structural chokepoint — In a supply chain where many stages are constrained, enrichment stands apart because it is simultaneously restricted by physics, treaty, and geopolitics. No other commodity’s critical transformation is controlled by international nonproliferation agreements. The concentration of 40 percent of global capacity in a single state entity creates a dependency that cannot be resolved by commercial procurement.
• Secondary supply created structural debt — Twenty years of warhead downblending suppressed prices, discouraged mining investment, and allowed the industry to undersize relative to actual reactor demand. The end of that program in 2013 created a supply gap that the mining industry has not yet closed, and the 10-to-15-year mine development timeline means the gap will persist regardless of price signals.
• Conversion is the overlooked bottleneck — With only four to five facilities globally and no significant new capacity, conversion may be the nearest-term constraint in the fuel cycle. It receives less attention than enrichment because it lacks the geopolitical drama, but the physics is the same: without UF6, centrifuges have nothing to enrich.
• Reactor design choices create decades-long dependencies — Fuel assemblies are reactor-specific and cannot be swapped between designs. A decision to build a particular reactor type in 1980 determines fuel supply chain dependencies in 2025. Switching suppliers requires years of engineering qualification, not a purchase order.
• Geographic separation between mining and processing reflects political geography — Kazakhstan mines uranium but cannot enrich it. The United States operates reactors but has limited enrichment capacity. Russia enriches uranium but mines relatively little. The physical supply chain crosses borders that carry geopolitical meaning at every transfer point.

Inside CompanyGraph

View this perspective inside CompanyGraph.