Why this conversation is happening now
Two events have converged to bring SMRs from speculative technology to active programme. The first was the European gas price crisis of 2022–2024, which demonstrated with brutal clarity what happens to UK electricity bills when global LNG markets tighten. The second is the Middle East conflict that began in early 2026, which has driven renewed urgency around energy sovereignty — with Energy Secretary Ed Miliband explicitly citing global instability as context for the Wylfa contract announcement.
Against this backdrop, the structural case for dispatchable, domestic, fuel-stockpileable baseload generation has become harder to dismiss. Wind power has expanded dramatically and will continue to do so — but wind cannot run at 90%+ capacity factor regardless of weather, cannot be stockpiled for multi-week periods of low pressure systems, and cannot currently be backed by economic battery storage at national scale. Gas fills that gap today. Something else needs to fill it in 2045.
Nuclear fuel requires far less volume than fossil fuels and can be stockpiled for extended periods — insulating operators from the kind of short-term supply disruption that caused UK gas prices to spike 500% in winter 2021–22. This is not a secondary consideration. It is a strategic asset that no renewable technology currently offers at equivalent scale.
The programme that just became real
For years the UK's SMR ambition existed as a competition, a shortlist, and a series of favoured bidder announcements. In April 2026, Rolls-Royce SMR was formally awarded the contract for technology design activities by Great British Energy – Nuclear, enabling work to begin immediately on the delivery of three 470 MW SMRs at Wylfa in North Wales.
Selected from a shortlist of four bidders. £2.6 billion was allocated in the 2025 Spending Review to enable the contract and wider programme delivery costs. The selection confirmed the UK's intent to pursue a domestic SMR programme with a British-designed reactor.
Prime Minister Keir Starmer announced that three SMRs will be sited at Wylfa, on the coast of Ynys Môn in North Wales. The site — home to decommissioned Magnox reactors that ran from 1971 to 2015 — already has grid connections, cooling water access, and a long history of nuclear operations. The site could potentially host up to eight SMRs.
Rolls-Royce SMR and Great British Energy–Nuclear signed the contract allowing work to begin on site-specific design, regulatory engagement, and ordering of long lead-time equipment from the supply chain. The National Wealth Fund committed up to £599 million to support development of the reactor design. GBE-N had already awarded more than £350 million in contracts to the supply chain.
Rolls-Royce SMR's design is currently at the final stage of the Generic Design Assessment process, which is expected to conclude by December 2026. Completion of GDA is a prerequisite for construction approval at Wylfa and any subsequent sites.
A spokesperson for Rolls-Royce SMR confirmed that the Wylfa SMRs will likely be operational by the mid-2030s. The 1.4 GWe three-reactor project will supply enough electricity to power the equivalent of three million homes for more than 60 years.
The case for SMRs over traditional nuclear
The UK's experience of large-scale nuclear construction at Hinkley Point C — currently projected to cost over £35 billion for 3.2 GW — illustrates the key problem with gigawatt-scale nuclear in a modern construction market: spiralling costs, extended timelines, and single-site concentration risk. SMRs are designed to sidestep these problems, not by being cheaper per unit of capacity, but by changing the risk profile entirely.
A Rolls-Royce SMR is designed for factory manufacturing — standardised components built in controlled industrial conditions rather than assembled piece by piece on a windy hillside. Each 470 MW module is smaller enough to be genuinely modular: if one site is delayed, the others continue. If cost overruns occur on unit one, they can be corrected before unit two. The learning curve — so damaging on one-off nuclear megaprojects — becomes an asset on a series-built fleet.
// The Rolls-Royce SMR design — key parameters
Output: 470 MW nameplate per unit · three units at Wylfa = 1,410 MW (1.4 GW)
Capacity factor: 90–95% operational availability
Operating life: 60+ years
Construction approach: Factory-built standardised modules — 90% of manufacturing and assembly in UK factories
Ownership: Rolls-Royce plc (majority) · BNF Resources (France) · Qatar Investment Authority · ČEZ (Czech Republic, 20% — also deploying up to 3 GW in Czechia)
Regulatory status: Generic Design Assessment final stage — completion expected December 2026
Programme estimate: £73 billion contribution to UK economy projected between 2025 and 2105
The arithmetic of replacement
Replacing the UK's gas fleet is not a question of whether SMRs work. The physics and the data from existing pressurised water reactor technology are well established. It is a question of scale, timeline, and industrial capacity. The numbers are large but not unprecedented.
How many reactors?
Rolls-Royce SMR nameplate output per unit: 470 MW
Operational availability: 90%
Effective continuous output per SMR: 423 MW
Minimum units to match gas capacity: 30,000 ÷ 470 = ~64 SMRs
With reserve margin, maintenance & demand growth: 70–75 SMRs
Working figure: 70 SMRs for full gas fleet replacement with margin
What does 70 SMRs cost?
Number of SMRs: 70
Total programme investment: 70 × £2.5bn = £175 billion
// Note: unit costs typically fall with series production. The 10th SMR should cost
// significantly less than the first. This estimate uses a conservative flat-rate figure.
What does it cost per MWh over 60 years?
Annual generation: 423 MW × 8,760 hrs = 3,705,480 MWh
60-year lifetime generation: 3,705,480 × 60 = 222,328,800 MWh
Capital cost ÷ lifetime generation:
£2,500,000,000 ÷ 222,328,800 MWh
Capital-only cost: ~£11.25 per MWh
// Even tripling this for operations, fuel, staffing, decommissioning and financing
// gives ~£34/MWh — compared with £150–300/MWh during the 2022 gas crisis.
The deployment timeline
Seventy reactors in thirty years is not a gentle programme. It requires a sustained industrial mobilisation — but it is comparable in scale to the expansion of North Sea oil infrastructure in the 1970s and 1980s, or the current rate of offshore wind deployment. The phasing matters: the first decade is the hardest, because supply chain, regulatory capacity, and construction skills all have to be built simultaneously.
| Period | Phase | SMRs delivered | Cumulative GW | Annual investment |
|---|---|---|---|---|
| 2026–2030 | Design, GDA, Wylfa construction | 0 (in build) | — | £2–3bn/yr |
| 2030–2035 | Pilot deployment | 5–10 | ~2.4–4.7 GW | £3–5bn/yr |
| 2035–2045 | Industrial scale-up | 25–35 | ~14–19 GW | £5–8bn/yr |
| 2045–2055 | National fleet completion | 30–35 | ~33 GW | £6–9bn/yr |
| 2026–2055 total | Full gas replacement | ~70 SMRs | ~33 GW | avg £5.8bn/yr |
Putting £5.8 billion a year in context
£175 billion sounds enormous. Spread over 30 years it is £5.8 billion per year — a figure that becomes more tractable when placed alongside other UK infrastructure commitments and the cost of the status quo.
The UK spent more supporting energy markets during 18 months of the gas crisis than the annualised cost of a full SMR fleet replacement programme represents in a decade. The question is not whether the country can afford SMRs. It is whether the country can afford to continue importing the volatility they would replace.
SMRs vs wind: not either/or, but both
The framing of nuclear versus renewables is a political narrative, not an engineering one. The UK's future grid almost certainly needs both — wind and solar for their now-proven cost trajectory and scalability, nuclear for the dispatchable baseload that wind structurally cannot provide. The question is not which to choose, but what role each plays.
Small Modular Reactors
Offshore wind (comparison)
The critical entry in this table is backup requirement. Every gigawatt of offshore wind currently installed in the UK requires a corresponding gigawatt of gas backup capacity standing ready for periods of low wind. Until battery storage or hydrogen can economically substitute for that backup at multi-day scale, removing gas from the system requires replacing it with something else that runs when the wind doesn't. SMRs are the only currently-engineered technology that does this at the required scale.
The industrial prize
Beyond energy security, the SMR programme represents one of the more credible industrial policy opportunities available to the UK in the coming decades. Rolls-Royce SMR estimates its programme could contribute up to $73 billion to the UK economy between 2025 and 2105, with 90% of manufacturing and assembly taking place in UK factories. The Wylfa project alone will create 3,000 jobs local to the site and an additional 5,000 nationally.
Scale that across 70 reactors over 30 years and the employment figures become transformative for specific regions. The Humber, Teesside, South Wales, and the Pembrokeshire coast — all areas with existing energy infrastructure, workforce skills, and economic need — are natural candidates for SMR deployment on or near retired gas plant sites. These are exactly the communities where the energy transition has so far been something that happens to them rather than for them.
There is also an export dimension. Rolls-Royce SMR is the only SMR company with multiple commitments in Europe — initial units at Wylfa and up to six units in Czechia. A domestically-proven British design, built at scale, has genuine export potential to European countries facing the same gas replacement challenge. The manufacturing investment required to build 70 reactors also creates the industrial base from which to sell the next 70 abroad.
The challenges that need honest answers
A programme of this ambition carries genuine risks. Acknowledging them is not a reason to reject the case for SMRs — it is a requirement for making it responsibly.
- Regulatory timelines. Nuclear licensing in the UK is thorough, rigorous, and slow. The Generic Design Assessment for Rolls-Royce SMR has been running since 2021 and concludes in late 2026 at the earliest. Planning consent for individual sites, beyond the GDA, adds further years. The mid-2030s target for first Wylfa power is achievable — but leaves limited margin for slippage if the programme is to build meaningful capacity before 2040.
- Cost overruns. The history of nuclear construction in the Western world is not encouraging — from Flamanville to Vogtle to Hinkley. The modular manufacturing argument for SMRs is theoretically compelling, but no commercial SMR fleet has yet been built at scale in the UK. The learning curve improvement that should make reactor 20 cheaper than reactor 1 has not yet been demonstrated in practice. Unit cost assumptions carry material uncertainty.
- Waste management. Each additional reactor adds to the UK's long-term nuclear waste obligation. The geological disposal facility programme remains in early stages. This is a manageable engineering challenge — France has operated far larger nuclear fleets for decades — but it requires political commitment, community consent, and sustained funding that spans multiple parliamentary terms.
- Public acceptance. Nuclear energy divides opinion. Some environmental groups — historically opposed — have shifted position as the climate emergency has intensified. Others have not. Community engagement at Wylfa and any subsequent sites will require genuine dialogue, not top-down imposition. The existing nuclear heritage at Wylfa is an asset; it means local familiarity with the industry and its risk profile.
- Financing structure. £2.5 billion per reactor is not a sum that private capital will finance at acceptable rates without government support, particularly in early deployment phases. The Regulated Asset Base model used at Hinkley provides one template; direct government investment via Great British Energy provides another. Getting the financing structure right — distributing risk appropriately between state and private capital — is as important as the engineering.