A consortium led by SGE, backed by Polish billionaire Michal Solowow, says it wants to deploy 14 small modular reactors across the UK, with a combined capacity of 4.2 gigawatts and an investment envelope of roughly £35 billion. If the unit count holds, this is the largest single private-sector pitch for serial SMR deployment yet announced in the UK. The interesting question is not how many reactors appear on the slide, but whether factory-built nuclear can do what conventional nuclear has consistently failed to do: drive costs down by repeating the same design.
Small modular reactors are nuclear plants built in a factory and shipped to site in pieces, sized in the hundreds of megawatts rather than the gigawatt-plus scale of Hinkley Point C or Sizewell C. The mechanism behind SGE's pitch is serial manufacturing: build the same reactor many times on a production line, and per-unit cost should fall as the line keeps running. Conventional nuclear costs have risen because every plant is a one-off, built on a unique site with bespoke supply chains. The SMR thesis assumes the opposite, that repetition is what drives cost down.
To anchor the headline number, 4.2 gigawatts of SMR capacity would represent generation on the same order as EDF's Hinkley Point C, currently under construction at 3.2 GW. EDF has said Hinkley will supply around six million UK homes when complete. Four-point-two gigawatts of SMR capacity, if delivered, would represent broadly comparable generation, though the homes-powered equivalent depends on capacity factor assumptions and grid absorption.
The reactor SGE wants to deploy, the BWRX-300, is a generation III+ boiling-water design developed by GE Hitachi. It has already cleared a key UK regulatory gate. The Generic Design Assessment milestone is the process by which UK regulators evaluate whether a reactor design is safe enough to be built at all, and GE Vernova has confirmed the milestone independently. The company's UK SMR page markets the design as UK-ready. That step does not authorise construction of any specific plant, but it removes a class of design-level objections from the table.
What the announcement does not include is the harder half of the project. No sites are confirmed. No power purchase agreements have been disclosed. No financing has closed. The £35 billion figure is an envelope described in SGE's own press release and corroborated by World Nuclear News, Global Banking and Finance, and BeBeez. The Guardian attributes the funding commitment directly to Solowow. None of those are audited financing closes.
The proposal is positioned to enter the UK's Advanced Nuclear Pipeline process, a published framework setting out how project proposers should engage with government on advanced nuclear projects. The Haynes Boone legal alert on the new Advanced Nuclear Framework describes the broader regulatory backdrop. That pipeline is the next gate, not the last one.
SMR timelines have slipped before. The technology is real, but the gap between an announced unit count and steel in the ground has been the consistent failure mode of the sector globally. Sites require planning consent in a UK system with local objections and grid-connection queues. Power offtake is contested; data centres and heavy industry want clean power, but the long-term contracts that finance new nuclear are not in evidence here. The cost-down thesis depends on serial manufacturing actually starting, which is itself a capital project nobody has broken ground on yet.
The watch item is whether SGE produces a site list, an offtake partner, and a financing structure within the next twelve months. Without any one of those, the 14-reactor number stays in the press-release column. With all three, the proposal becomes a real test of whether the same reactor, built fourteen times in a factory, can deliver electricity more cheaply than a reactor built once on a construction site.