Since the Starmer government came to power in 2024, it has made a series of announcements that put the UK in the vanguard of the claimed Nuclear Renaissance. The government is talking about a ‘Golden Age of Nuclear’ in the UK. However, closer examination shows that these announcements are mostly talk about what the government hopes to do and hopes might happen with no new projects in the pipeline. The onus will be on the private sector to come forward with proposals.
In terms of actual nuclear projects, there is one under construction, Hinkley Point C, one, Sizewell C with an investment decision taken that might start construction in 2-3 years and a project for three Rolls Royce Small Modular Reactors (SMRs) for the Wylfa site where it is hoped an investment decision will be taken in 2029. The Hinkley Point C project for two French European Pressurised Reactors (EPR, 3.2GW) is seven years late and 90% over budget and is at least six years from completion. The Sizewell C project is expected to use a copy of the Hinkley design, and its construction should have followed about two years behind Hinkley so the workforce could transfer smoothly from Hinkley to Sizewell, but it is now at least nine years behind. Even if the government’s central completion date estimate is met, Sizewell will not start to generate until 2039. The estimated cost of Sizewell at £40.5bn (2024 money) is 70% more than the real estimated cost of the Hinkley Point project at the point of its Final Investment Decision. This makes a mockery of claims that Sizewell would be cheaper than Hinkley because of the ‘learning’ that has accumulated at Hinkley Point. If it goes ahead, the Wylfa project will not start to generate until after 2035. If there are no more delays to these projects, it will be 2040 before the UK’s nuclear capacity will have returned to its 2015 level of about 9GW.
In 2022, the Boris Johnson government set a target of up to 24GW of new nuclear capacity in addition to the Hinkley project to come online by 2050. The qualification ‘up to’ meant the target was effectively meaningless but the Starmer government has conspicuously not adopted this target.
So why is it so difficult and does it take so long to build nuclear capacity and is the UK particularly bad at it? Research commissioned by the British government found that worldwide, on average, nuclear construction from investment decision to first power took 13-17 years. In addition to this must be added the period to get to a final investment decision. This includes: choosing the vendor and technology: assessment of the design by the national safety regulator; finding and establishing the suitability of the chosen site; and setting up a financial model to provide the finance, own the plant and buy the power. This process is unlikely to take less than five years and takes longer. So, the lead-time for a nuclear project is likely to be at least 20 years.
It has long been difficult to establish new sites for nuclear power plants. In 2010, the British government identified eight sites, all coastal, that would be suitable for new plants. All were sites with existing nuclear capacity. However, three of these have been allocated already (Hinkley Point, Sizewell and Wylfa) and two (Heysham and Hartlepool) are close enough to population centres that a change to the siting criteria would be needed. The UK’s Semi-Urban Population Density Criterion is a threshold of population density in the area within 30km of a proposed nuclear reactor and these two sites, chosen 60 years ago, do not meet that criterion. Of the eight sites, all except Wylfa are forecast to be below the annual flood level by 2050. Given how rapidly perceptions of the speed of climate change are altering, it would seem sensible to make siting processes more rigorous, not to streamline it as the government is trying to do. The government plans to open siting to new sites and for developers rather than government to choose sites but this system is some way from being implemented.
In the past, nuclear power plants have been built by utilities with an effective monopoly over generation that allowed them to pass on whatever costs they incurred to consumers. Nuclear projects have always been financially risky, but the utility’s monopoly meant these risks fell on electricity consumers. Utilities are now increasingly exposed to competition in generation markets so they cannot risk choosing generation options that are not economic. Many of the large utilities have been broken up or have lost market share to new entrants and no longer have the strength to finance nuclear projects. Even Electricité de France, the world’s largest utility, cannot finance a programme of six new reactors in France without support from government and power purchase arrangements that remove any exposure for them to competitive wholesale markets. The UK tried to get round the problem of finding financiers by bringing in institutional investors such as pension funds for the Sizewell C project. However, despite extremely generous terms with the risk falling entirely on consumers, only a little over half the cost of the plant will be met by private investors with the rest falling on taxpayers.
There are two designs that meet current safety standards, the Framatome EPR and the Westinghouse AP1000 and both have a very poor record of construction cost and time control. A third design, the Korean APR1400 would need costly upgrades from the design built so far (in Korea and UAE) to meet European standards. The design is licensed from Westinghouse who are not willing to allow it to be built in Europe except in the Czech Republic so it is not an option for the UK. All three designs date back to the 20th Century pre-dating reactor design challenges posed by the 9/11 terror attack on New York, the Fukushima disaster and the issues raised by the Russian invasion of Ukraine and the Zaporizhia site. No fundamentally new large reactor designs are being developed.
Despite their portrayal in the media as proven and economic designs, SMRs are unproven and based on some highly questionable assumptions. For example, in the past, reactors have increased in size to gain scale economies and reduce costs. Now we are told that reducing the size will reduce costs. The only commercial order for an SMR is for the Darlington site in Canada for a GE Vernova BWRX-300 but this had not started construction in February 2026.
What is conspicuously lacking in the UK are projects to follow Sizewell C and Wylfa. For large reactors there is no process established or even under development to choose technologies, to identify suitable sites, to identify investors and to develop a viable financial model that is fair to both investors and consumers. There is also a problem of credibility of the available large reactor designs because of their poor record. The logical conclusion is that no large reactor stations can come online in the UK after Sizewell C until the mid-2040s at the earliest and that Sizewell C might turn out to be the last large reactor project in the UK for the foreseeable future. This places the focus on SMRs.
It is likely to be around the late-2030s before there is any experience worldwide for any of the available SMR designs and all experience with nuclear power suggests that until there is that experience, it would be extremely risky to order SMRs. That means that even if SMRs prove easier to site and build than large reactors, they could not contribute significantly to generating capacity until the mid-2040s at the earliest.
Regardless of whether small or large reactors are built, there remains the issue of who will own them. Given the difficulty in attracting private investors for Sizewell C, it is hard to avoid the conclusion that the major or only investor for future nuclear projects will be central government, which will also provide the finance. It has the cash and can take financial risks that private companies could not contemplate. The presence of a major government stake would give minority private investors some comfort that the project will not be allowed to fail no matter how badly it is going. Even for central government this will stretch their financial and managerial capability. For example, if the UK’s 24GW target was pursued, this would require an investment of more than £300bn between now and 2050. The opportunity costs of such an investment to government would be massive.
The UK government is unwilling to face up to the challenges that are rooted in the nuclear technology these issues raise and is focused on reforms to planning and regulation. It claims that inefficiencies in these are at the root of the delays and cost overruns suffered by UK nuclear projects. It provides no evidence to back these claims – there is none.
In February 2025, the government appointed a Taskforce to examine nuclear regulation and planning processes but it seemed to have made up its mind what conclusions it expected. Starmer said “I’m putting an end to it – changing the rules to back the builders of this nation, and saying no to the blockers who have strangled our chances of cheaper energy, growth and jobs for far too long.” The final report by the Taskforce went along with this rhetoric and within a day of publication of its report, the government had promised to implement all its recommendations within two years.
The government also completed a new nuclear power plant siting policy, EN7, in February 2025 aimed at expanding the number of sites available for nuclear power plants. The onus was on the private sector to propose sites: ‘it is for applicants to bring forward proposals.’
In February 2026, the government published its Advanced Nuclear Framework. As with the siting policy, the focus was on facilitating private sector investment: ‘Government to recognise most credible advanced nuclear projects to help unlock private investment and get projects off the ground.’ It was couched in the need for a massive expansion in electricity demand driven by AI centres. It is intuitively plausible that electricity demand will grow because of electrification of cars and replacement of gas space heating with electric heating, but the demand for AI is speculative and has the feeling of a bubble that might burst. Government has consistently underestimated the extent that energy efficiency measures will reduce demand. In 2005, the UK government forecast that electricity demand would grow by about 25% by 2025. Since 2005, demand has consistently fallen by a total of 22% by 2024 with the gap between the forecast and the outturn equivalent to the output of the 24GW nuclear programme.
The narrative of the UK nuclear power sector being hamstrung by obstructive regulatory and planning processes and with the private sector desperate for the opportunity to invest in nuclear projects is a dangerous fantasy. At best, it will divert resources from options able to deliver climate change goals, quickly, reliably and cost-effectively and at worst it will dismantle the processes on planning and regulation of nuclear projects that need to be strengthened not weakened.
