Why Fuel Assembly Manufacturing Is Nuclear’s Hidden Bottleneck

Most talk about the nuclear fuel supply chain stops at enrichment, as if a stockpile of enriched uranium could power a reactor on its own. It cannot. That material first has to become precisely engineered fuel assemblies, approved by regulators and built for one specific reactor design, and nuclear fuel assembly manufacturing is where much of the supply chain’s lock-in actually sits.

The timing matters. European utilities running Russian-designed reactors are working to replace Russian fuel. At the same time, advanced fuels are exposing gaps in a manufacturing base that is otherwise adequate for today’s fleet. Mining, conversion and enrichment come before all of this, but the bottleneck worth watching now sits further down the line.

Here you will see how a fuel assembly is built, who builds them, why changing suppliers takes years, and which listed companies offer indirect exposure. This is educational material only, not investment advice.

How uranium becomes a fuel assembly: the fabrication step explained

Start with the finished object. A fuel assembly is a tall, rigid bundle of metal rods held in a precise grid. A large reactor core contains well over a hundred of them, and together they hold tens of thousands of rods.

Every one of those rods began as a ceramic pellet about the size of your fingertip.

Fabrication sits in the middle of the complete uranium fuel cycle, after mining, conversion and enrichment and before reactor loading, so each upstream stage shapes how much fuel a fabricator can actually turn into assemblies for you.

From powder to rod

Fabrication follows a fixed sequence:

1. Powder: Enriched uranium hexafluoride (UF6, the gas form used during enrichment) is converted into uranium dioxide (UO2) powder.
2. Pellet: The powder is pressed and sintered (baked at high temperature until it fuses into a dense solid) into ceramic pellets roughly 10 mm wide and 10 mm tall.
3. Rod: Pellets are stacked inside zirconium alloy cladding tubes, which are welded shut to form sealed fuel rods.
4. Assembly: Rods are bundled into an assembly with spacer grids and guide tubes.
5. Reactor loading: Finished assemblies are shipped and loaded into the core during a refuelling outage.

Cladding is the metal tube that holds the pellets and keeps radioactive material sealed in. The zirconium alloy used for it is a specialised input, and tubing, welding and inspection capacity can limit how many assemblies a plant produces even when pellet output is available.

From rod to reactor core

Fabrication is where generic fuel becomes reactor-specific. Rod dimensions, grid design, guide tube positions and enrichment zoning (where different enrichment levels sit within the assembly) all have to match the core they are going into.

According to the World Nuclear Association (WNA) and the International Atomic Energy Agency (IAEA), a typical 1,000 MWe pressurised water reactor (PWR) core holds about 157 assemblies. A larger 1,100 MWe unit holds about 193.

One large core, by the numbers A 1,100 MWe PWR holds more than 50,000 fuel rods and roughly 18 million pellets, according to WNA data.

None of this is a one-off purchase. At each outage, operators replace roughly a quarter to a third of the core, about 60 assemblies in a 157-assembly reactor. IAEA and WNA material from October 2026 cites typical refuelling intervals of 12 to 18 months, while other industry sources place many large PWRs on 18 to 24 month cycles, so the realistic range spans about 12 to 24 months depending on design.

For you, the implication is simple: fabrication demand tracks the operating fleet. Every running reactor needs a steady stream of design-matched assemblies, which makes this a recurring link in the chain rather than a commodity conversion step.

Who makes the world’s reactor fuel assemblies?

That recurring demand flows to a small group of suppliers. Four names dominate the conversation, but the real story is concentration by reactor design family.

The four headline fabricators

Westinghouse has been owned 51% by Brookfield Renewable and institutional partners and 49% by Cameco since the deal closed on 7 November 2023, according to Cameco’s filings updated in October 2026. Cameco states that it does not control the company. Westinghouse is one of the largest fabricators, with the biggest US capacity among major Western suppliers in WNA’s profile, though no reviewed source ranks it the world’s largest.

Framatome is a subsidiary of EDF, which is wholly owned by the French state. Framatome’s governance disclosure of 29 August 2025 listed EDF at 75.5%, Mitsubishi Heavy Industries at 19.5% and Assystem at 5%; EDF has since described acquiring Assystem’s stake, so you should confirm the current split against EDF reporting.

Global Nuclear Fuel (GNF) specialises in boiling water reactor (BWR) fuel. It is described publicly as a GE Vernova-led alliance with Hitachi, and current ownership percentages should be checked against parent filings. TVEL, the fuel arm of Russia’s state nuclear company Rosatom, has historically been the main supplier of fuel for Russian-designed VVER reactors in Eastern Europe and Asia.

Treat the capacity column as a rough guide to scale only. Sources report in different units (tonnes of uranium versus tonnes of heavy metal) and cover different scopes, so precise comparisons between rows would mislead you.

Springfields, near Preston, is the UK’s sole commercial fuel fabrication site and a Westinghouse subsidiary, according to the UK Office for Nuclear Regulation.

Beyond the big four

Four is not the full count. Other commercial fabricators serve national or regional fleets:

• KEPCO Nuclear Fuel (South Korea)
• Mitsubishi Nuclear Fuel (Japan, including APWR designs)
• ENUSA (Spain)
• CNNC (China, serving the domestic fleet)

What you should take from this is that each reactor family has only a handful of qualified suppliers, sometimes just one. That concentration becomes very visible when a supplier suddenly becomes politically unacceptable.

Replacing Russian fuel: where VVER substitution stands

Ukraine shows substitution can work. Westinghouse has supplied VVER fuel there for years and now supplies all nine operating units under Ukrainian control, building deep experience with the larger VVER-1000 design.

The harder part is the smaller reactor.

Two designs, two qualifications VVER-440 and VVER-1000 assemblies use different hexagonal geometries. Each needs its own engineering, testing and regulatory qualification, so success with one does not transfer to the other.

Westinghouse’s VVER-440 fuel development was targeted for completion around 2028, according to Carnegie Endowment analysis from November 2025. Elsewhere, progress is reactor by reactor.

Slovakia is the clearest warning. Commentary reviewed by Bellona and Bruegel indicates that Framatome’s VVER-440 supply relies on licensed Russian design know-how during the transition, which means independence from Russia is partial for now.

For you, the read is that a geopolitical shock turned fabrication capability into a security asset. Any claim that Europe has simply switched suppliers deserves scepticism.

Sanctions: what is and is not restricted

Restrictions are specific, not blanket:

• United States: Russian uranium imports have been restricted by legislation since 2024, with waivers available. Reviewed sources did not establish a blanket ban on every TVEL service.
• European Union: As of 2025 reporting, there was no comprehensive ban on Russian nuclear fuel imports, and deliveries to some VVER operators continued.
• United Kingdom: The current UK position on TVEL fabrication sanctions could not be confirmed from reviewed sources.

Why switching costs are high, and what that means for investors

The slow pace of VVER substitution is not an accident. It reflects how fuel is regulated everywhere.

Why switching suppliers takes years

Regulators in every jurisdiction must separately sign off on each assembly design, and bringing a new supplier up to standard for a given reactor type typically means several years of work. Five barriers explain why:

1. Fuel is reactor-specific. Dimensions, grids, guide tubes, hydraulics and enrichment zoning must match the core.
2. Fuel sits inside the licensed safety case. A new supplier must provide mechanical, thermal, neutronic and accident-performance evidence.
3. Lead-use assemblies come first. A small number are irradiated and inspected over operating cycles before a full reload is approved.
4. Manufacturing must be qualified separately. The vendor has to prove it can produce pellets, rods and assemblies repeatably.
5. VVER fuel adds another layer. Two distinct geometries mean two qualification programmes.

Conventional versus advanced fuel In a report dated 21 September 2025, the IAEA found that industrial capacity for conventional fuels exists worldwide, but capacity for LEU+ and HALEU fuels does not yet meet expected demand.

LEU+ is uranium enriched slightly above the usual 5% limit for commercial fuel, and HALEU (high-assay low-enriched uranium) runs higher still, mainly for advanced reactor designs. The scarce asset is qualification-adjusted capacity, not factory floor space, so judge any fabricator on its licensed designs and contracted demand rather than headline plant size.

For readers wanting to go deeper on lock-in, our dedicated guide to nuclear fuel assembly manufacturing barriers breaks down each of the five barriers protecting incumbent fabricators.

Routes to exposure

Direct access is limited, and these are educational observations only:

• Westinghouse is private; indirect exposure runs through Cameco (TSX/NYSE: CCJ, 49%) and Brookfield Renewable (confirm listing and stake in current filings).
• Framatome sits inside state-owned EDF and is generally not accessible to international investors.
• TVEL belongs to the Russian state and faces sanctions across many Western jurisdictions.
• GNF exposure runs through GE Vernova (NYSE: GEV) and Hitachi (TSE: 6501), subject to verification of current participation.

Fabrication is a small slice of each parent’s business, so your exposure is diluted. Contract terms vary by customer and are typically multiyear.

Risks to weigh

• Customer concentration on a few utilities or one reactor design
• State ownership and political influence on contracts
• Shifting sanctions rules
• Dependence on enriched uranium supply
• Zirconium alloy cladding constraints
• Qualification delays pushing revenue years past investment
• Overcapacity risk in conventional fuel if reactor builds slip
• Inconsistent public capacity data

What to remember about fuel fabrication before drawing conclusions

Conventional capacity is broadly adequate for today’s reactors. What is scarce is qualified, regulator-approved manufacturing for specific reactor designs and advanced fuels, and that scarcity explains both the slow VVER transition and the stickiness of supplier relationships.

Advances in enrichment technology are widening the supply of higher-assay material, which shifts the pressure point to the qualified fabrication capacity needed to turn that material into licensed assemblies.

If you follow this corner of the fuel cycle, four signals deserve your attention:

• VVER-440 licensing milestones in Central Europe
• LEU+ and HALEU fuel qualification progress
• Changes to US, EU and UK sanctions
• Ownership disclosures from Cameco, Brookfield, EDF, GE Vernova and Hitachi

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research, check current company filings and consult with financial professionals before making investment decisions.