Nuclear power has quietly kept Britain's lights on for seventy years, but the fleet doing that work today is old and shrinking fast. Most of the reactors running in 2026 were built in the 1980s and were only ever meant to last a few decades. One by one they are being switched off. At the same time, the country is trying to build the biggest wave of new nuclear it has attempted since the 1980s — two enormous power stations, a fleet of factory-built "small" reactors, and now a surprise £35 billion bid from a Polish billionaire. This article walks through what exists, what's coming, and what it all adds up to.
1. The existing fleet: powerful, but on the way out
The UK currently runs nine reactors across four sites, with a combined capacity of roughly 5.9 gigawatts (GW) — enough, on a good day, to power millions of homes. Eight of these are a distinctively British design called the Advanced Gas-cooled Reactor (AGR), and one is a more conventional Pressurised Water Reactor (PWR). All are run by EDF.
The catch is age. The AGRs are held together by graphite cores that slowly crack as they age, so their closing dates are decided by regular inspections rather than a fixed calendar. EDF has repeatedly nudged the dates outward when inspections came back clean — but the direction of travel is only one way: off.
| Station | Type | Capacity | Planned closure | Status |
|---|---|---|---|---|
| Heysham 1 | AGR | ~1.2 GW | March 2028 | Running |
| Hartlepool | AGR | ~1.2 GW | March 2028 | Running |
| Heysham 2 | AGR | ~1.25 GW | March 2030 | Running |
| Torness | AGR | ~1.25 GW | March 2030 | Running |
| Sizewell B | PWR | ~1.2 GW | ~2055 | Running |
Heysham 1 and Hartlepool were last extended in September 2025 to March 2028. Sizewell B, the newest and only PWR, opened in 1995 and is targeted to run a full 60 years to around 2055.
The pattern is stark: by March 2028 two of the four AGR stations shut down, and by March 2030 the other two follow. After 2030, unless something new has come online, the UK's entire nuclear output would rest on a single station — Sizewell B — until the new build arrives.
How we got here: the historical picture
Britain was a nuclear pioneer. Through the 1970s and 80s it built out a large fleet of Magnox and then AGR reactors, and by the mid-1990s — with Sizewell B added in 1995 — installed nuclear capacity peaked at roughly 12.7 GW. Generation peaked in 1998 at 99.5 terawatt-hours, more than a quarter of all UK electricity. Since then it has been a long, steady decline as the old Magnox stations closed through the 2000s (the last, Wylfa, in 2015) and the AGRs began retiring in the early 2020s.
Figure 1 — UK nuclear capacity, 1980–2026 (approximate). Capacity roughly doubled from the early 1980s to a mid-1990s peak near 12.7 GW, then fell by more than half as Magnox and AGR reactors retired. Figures are illustrative and rounded to show the trend.
2. The big new build: Hinkley Point C and Sizewell C
The centrepiece of the UK's nuclear revival is two very large stations, each based on the French EPR design and each roughly equal to the entire current fleet in output.
Hinkley Point C (Somerset)
Two reactors of 1,600 megawatts each — 3.2 GW total. It is the first new British nuclear station in a generation, but it has become a byword for delay and cost overruns. As of February 2026, EDF confirmed the first reactor won't start until 2030 — five years later than the original 2025 target — and the price tag has climbed to around £48 billion in today's money (about £35bn in 2015 prices). The latest slip was blamed on slow progress installing the plant's vast web of pipes, cables and systems.
Sizewell C (Suffolk)
A near-copy of Hinkley, also two EPR units — this time 3.2 GW (1,630 MW each) — intended to be cheaper to build because it reuses Hinkley's design. After years of uncertainty, the government took a final investment decision in July 2025 and reached financial close in November 2025, with the state taking a 44.9% stake alongside Canada's La Caisse, Centrica, EDF and Amber Infrastructure. Total cost is put at £38 billion, and the station is expected to start generating in the mid-to-late 2030s, powering the equivalent of six million homes.
3. Small Modular Reactors: the factory-built future
The most talked-about shift in nuclear is away from vast one-off mega-projects and towards Small Modular Reactors (SMRs) — reactors small enough to be built in factories in standard modules and assembled on site, in theory faster and cheaper than a Hinkley. Two very different SMR efforts are now underway in Britain.
Rolls-Royce SMR — the government's pick
In June 2025, after a two-year competition, the government's Great British Energy – Nuclear body chose Rolls-Royce SMR as its preferred partner. The plan is for three reactors delivering at least 1.4 GW — enough for around three million homes — with Wylfa in Wales confirmed in November 2025 as the first site. A contract to begin site design and planning was signed in April 2026, £2.6bn was set aside in the 2025 Spending Review, and a final investment decision is expected around 2029, with grid connection in the mid-2030s.
The billionaire's £35bn bet
In July 2026, a consortium led by Polish billionaire Michał Sołowow — Poland's richest person, through his firm Synthos Green Energy — unveiled a plan to invest around £35 billion in 14 small modular reactors across three UK sites, with the first units targeted for commercial operation in 2034.
The reactors would be GE Vernova Hitachi BWRX-300s — 300-megawatt boiling-water reactors (so roughly 4.2 GW in total if all 14 are built) — at an estimated £2.2–2.5bn per unit. Crucially, the consortium says it would lead with private capital first, seeking a government price-support contract only once the reactors are generating, rather than up-front public funding. It has also floated a tie-up with Google Cloud, which may invest up to £4.5bn in data centres powered by the reactors — a sign of how the AI-driven surge in electricity demand is pulling tech giants toward nuclear.
4. Putting it together: the capacity outlook
So what does the arithmetic look like? In the near term, capacity keeps falling as the AGRs retire faster than new plants arrive — a genuine dip around the end of this decade. Then, if Hinkley, Sizewell C and the SMR programmes all deliver, capacity climbs back well above today's level through the 2040s. The government's stated ambition is up to 24 GW of nuclear by 2050, around a quarter of projected electricity demand.
Figure 2 — Illustrative capacity outlook to 2050. Existing reactors (grey) fall away by 2030; Hinkley Point C (blue), Sizewell C (teal) and SMRs (violet) build the fleet back up. The dashed amber line marks the government's 24 GW / 2050 ambition. This is a scenario based on announced projects and target dates — real delivery depends on construction, financing and regulatory approvals, all of which have slipped before.
The honest caveats
Every date on the chart above is a plan, not a promise. Hinkley Point C alone has slipped five years and billions over budget. SMRs have never been built at scale in Britain, so their timelines and costs are estimates. And a "final investment decision" — the point of no return for spending — still lies ahead for both the Rolls-Royce and Sołowow projects. The direction of policy is clearly pro-nuclear and cross-party, but the gap between ambition and steel in the ground remains the central question.
The bottom line
Britain is running two races at once: retiring a worn-out fleet, and building its replacement. For the next few years the retirements win, and nuclear's share of UK electricity keeps drifting down. The pay-off — a larger, more modern fleet built around two EPR giants and a wave of factory-built SMRs — is real but sits mostly in the 2030s and beyond. Whether the UK hits its 24 GW goal will depend less on ambition, which it has in abundance, than on delivery, which has been its weak spot for decades.