UK Fusion Energy (UKFE) has opened a five-lot, up-to-£115 million procurement for "System Partners" to design, integrate, test and control the key systems of the STEP spherical tokamak at West Burton, Nottinghamshire. The headline number is modest next to the £200 million ILIOS construction tranche awarded in March 2026, and it buys almost no hardware. That is precisely why it is worth reading closely: the lots map the programme's real technical and organisational bottlenecks, and they tell suppliers where STEP intends to place engineering risk over the next two to four years.
The lots, and what each one is actually buying
| Lot | Partner | Ceiling (ex-VAT) | What UKFE is outsourcing |
|---|---|---|---|
| 1 | Breeder Blanket System Partner | £20m | Design, integration and realisation through manufacture; management of engineering data; detailed design and supply-chain contracting |
| 2 | Fuel Cycle System Partner | £20m | Support then lead fuelling and fuel-cycle sub-systems, from design to manufacture and operation |
| 3 | Test Operations Partner | £35m | Operation of UKFE test facilities; data feeding SPP design and safety justification |
| 4 | Control System Partner | £15m | Plant-wide control system delivery and integration |
| 5 | Systems Engineering & Whole Plant Integration Partner | £25m | Capability inside the STEP Design Authority during Tranche 2A |
Reported facts unless stated. Values are lot ceilings from the Find a Tender notice (reference 107283-2026, first published 11 November 2026); the £115m is the arithmetic sum of the five maxima, not a forecast spend.
Three features deserve attention.
The Test Operations Partner is the largest single lot at £35m — larger than the breeder blanket or fuel cycle. That is an unusual allocation of money and a signal about programme logic. Fusion designs cannot be qualified by analysis alone; STEP's safety case and design freeze will rest on experimental data from test rigs. UKFE is explicitly retaining ownership of test requirements and engineering intent, with rig suppliers responsible for rig design. In effect it is buying an operator for its own test estate, and the deliverable that matters is data and safety justification, not equipment.
The Systems Engineering and Whole Plant Integration lot is an admission of bandwidth constraint. £25m to place people inside the Design Authority during Tranche 2A is a resourcing decision dressed as a contract. For a first-of-a-kind plant integrating a tokamak, tritium plant, blanket, divertor and balance of plant into a single licensable facility, the scarce resource is not steel or concrete but experienced systems engineers capable of owning interfaces. Firms with nuclear or complex-systems integration pedigree — rather than fusion specialists alone — should read this lot as the entry point to the programme.
The blanket and fuel cycle lots concentrate the hardest physics. Tritium self-sufficiency is a formal STEP objective, and it depends on both the breeder blanket and the fuel cycle working together. Published STEP blanket development work indicates a move from an early helium-cooled liquid-lithium concept with V-4Cr-4Ti structure to a CO₂-cooled solid Li₂O breeder with a titanium beryllide multiplier and 15-15-Ti austenitic steel, with reduced-activation ferritic/martensitic steels also under consideration. A partner taking Lot 1 is therefore inheriting unresolved materials and manufacturability questions — which is why prior UKFE notices have sought scaled demonstrators to prove manufacturability of ferritic-martensitic steels for the blanket. Lot 2 carries comparable uncertainty: the fuel cycle must breed, recover, purify and re-inject tritium at rates no commercial plant has yet demonstrated.
Commercial terms that change the bidding calculus
The framework is a two-year term with options to extend by one year twice, capped at four years in total. Contracts are expected to start in late April 2027, with tenders due 29 January 2027 and awards anticipated 6 April. (Sources differ by a day on the start date — 22 versus 23 April — which is immaterial but worth treating as "expected" rather than fixed.)
Two further points matter more than the headline value:
- A single supplier may win up to three of the five lots. Cross-lot capability is therefore a genuine competitive advantage. A bidder able to combine control systems with whole-plant integration, or blanket design with test operations, can capture a materially larger share of the £115m ceiling and, more importantly, shape interfaces.
- The £115m is a ceiling across all lots and extension periods combined. If fully spent over the two-year initial term, implied run-rate is roughly £57.5m a year; spread over four years, about £29m a year. Awards may fall well below the ceilings, and the competitive flexible procedure under the Procurement Act 2023 allows UKFE considerable freedom to shape and negotiate. Suppliers should plan on scope, not on the advertised number.
Where this sits in the STEP money trail
The sequencing is now visible. ILIOS — the Kier–Nuvia joint venture with AECOM, AL_A and Turner & Townsend — took the design-and-build tranche in March 2026, covering buildings, infrastructure, site management and constructability assurance, with up to 8,000 peak construction jobs cited by government. The System Partners now being tendered sit on the plant side of the interface: they define and integrate the machine that ILIOS houses. This split creates a classic interface risk. ILIOS owns the site and buildings; Lot 5's partner must make the plant's systems cohere inside them; Lots 1, 2 and 4 own specific technical envelopes. Who arbitrates when blanket maintenance requirements clash with building layout? The Design Authority — which is why Lot 5's placement inside it is the most strategically significant element of the tender, despite being neither the largest nor the most technically glamorous lot.
Note also a genuine source disagreement on the earlier tranche: NUVIA describes phase one as a €230 million, 4.5-year contract, while the UK government and Kier describe £200 million over three years. The UK and Kier framing should be treated as authoritative, but the discrepancy is a reminder that STEP's commercial disclosures are still being normalised.
What this means for the metals and manufacturing supply chain
The direct read-across to steel demand is close to nil. This procurement buys engineering services and intellectual capability, not tonnage, and the largest material volumes at West Burton will flow through ILIOS and later hardware packages.
However, the indirect signal is more useful than the headline figure suggests. Lot 1 explicitly includes "realisation of designs through to manufacture" and contracting through the supply chain, and Lot 2 extends to manufacture and delivery. The Breeder Blanket System Partner will therefore become a de facto route to market for specialist fabricators: vacuum-vessel-grade stainless and advanced austenitic steels, RAFM steels, beryllium-containing components, lithium ceramic breeder pebbles and cooling-plate assemblies with tolerances and cleanliness standards that sit far outside conventional power-plant practice. Qualification, not capacity, will be the barrier. Suppliers who begin now to build a documented route into the blanket demonstrator programmes and UKFE test campaigns will be better positioned than those waiting for a plant hardware tender.
Judgement
The £115m is small relative to STEP's eventual cost and smaller still against the up-to-£10 billion programme envelope that ILIOS's appointment referenced. Its importance is structural rather than financial. UKFE is populating the layer between a construction partner and a future hardware supply chain — the systems integrators, test operators and design-authority augmentees who will determine whether a 2040 operating target remains credible. Two things to watch: whether the awards concentrate in a small number of multi-lot winners (which would signal thin depth in the UK engineering base, or deliberate consolidation around proven integrators), and whether the Test Operations Partner lot produces a visible test-data pipeline feeding the safety case. If the latter stays opaque, the 2040 date should be treated with scepticism.
For bidders, the practical message is that the enquiry deadline of 8 January 2027 comes well before the 29 January tender deadline, and that the most valuable question to put to UKFE is not about price but about interface authority: who owns the boundaries between Lots 1, 2 and 4, and how much decision rights the Lot 5 partner actually holds inside the Design Authority.
Source & methodology +
MetalMate separates original-source evidence from summary and analysis. Verify the original source before procurement, legal or engineering decisions.
- Article status
- Published · Version 1
- Source date
- Not supplied
- MetalMate publication
- 12 November 2026
- Verification
- Automated evidence checks; not human reviewed
Bids invited for construction of UK prototype fusion plant — World Nuclear News

