On the day before yesterday, 15 September 2025, the British government announced a new ‘golden age of nuclear’, featuring the biggest build-out of nuclear power in generations and a ‘landmark’ US-UK partnership. To those who have been in the industry a while, it feels like Groundhog Day.

Fifty years ago, the then Secretary of State for Energy, Tony Benn, championed the British AGR programme of new nuclear power stations. While he might not have used the words “too cheap to meter” commonly attributed to him, he was certainly persuaded that building nuclear would bring down the price of electricity. It was a judgement that he regretted in later life.

Lord Marshall, head of the Central Electricity Generating Board in the 1980s and a key influence on Margaret Thatcher, went through a similar transition to Benn. Marshall resigned in 1989 over the City’s refusal to include nuclear in the privatisation of the industry. It seemed to him that safety regulations had got out of hand, particularly in the aftermath of the Chernobyl disaster a few years earlier; and this was a primary reason why nuclear was commercially unviable.

Will this new ‘golden age’ be more successful than the last one? The question certainly needs to be asked. A dispassionate assessment of recent evidence is not encouraging. The Hinkley Point C project is reportedly 12 years behind schedule (the latest estimate of its commissioning year is 2031) and 4 times over budget. There is an echo here of the Dungeness B AGR which began construction in 1965 but only starting producing electricity in 1983.

The financial risk at Hinkley C is at least underpinned by its parent company, Electricite de France. The next European Pressurised Reactor (EPR) to be built on our shores, Sizewell C, was approved by the government in 2022. It carries no such guarantee: if it runs significantly over budget, the cost will be picked up by the British public, rather than by EDF.

So it’s a little worrying that Sizewell C was recently described – as reported in the Financial Times at the end of August – as “terrifyingly complex” and “almost unbuildable” by a former head of EDF, Henri Proglio. His view is shared by many in the industry who feel that the company’s campaign to build the world’s safest nuclear station – in the aftermath of Chernobyl in 1986, the twin towers attack in 2001 and Fukushima in 2011 – has led to a doomed design, with its excess of independent cooling systems and containment shields of reinforced concrete several feet thick.

Scarred by its experiences of the EPR in Finland and France, as well as the UK, EDF itself has moved on to a new and simpler design: EPR2. By common consent amongst both supporters and detractors of the project, Sizewell C is quite likely to be the last of its kind in the world. Opinion is divided as to whether Sizewell C will reap the benefits of replication and standardisation in the wake of Hinkley C, or whether the historical pattern of vast overrun will be repeated. The likelihood that the design will be near-obsolete, and that EDF’s attention will be largely elsewhere, doesn’t seem reassuring on the face of it.

Only time will tell who is right. But whatever the truth about Sizewell C, we surely need a more grown-up debate than another historically-blind promise of a golden age. We need a fearless and open national discussion about the cost/risk trade-off: how much should society be willing to pay to lower the risk of a nuclear catastrophe from X in a million to Y in a million?

It does not appear that we are going to get it.