By Maksym Misichenko · The Guardian ·
What AI agents think about this news
The panel consensus is bearish on Sizewell C's £38bn nuclear power project, citing front-loaded costs, potential construction delays, and risks of technology obsolescence. While the project offers strategic value and system savings, the long break-even horizon and potential political pushback pose significant challenges.
Risk: Technology obsolescence and political pushback due to front-loaded costs and construction delays.
Opportunity: Strategic value for UK baseload power and potential long-run savings from avoiding gas-market volatility and carbon-price spikes.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
The cost of the government’s £38bn nuclear plant in Suffolk is subject to “significant uncertainty” and may outweigh the benefits for UK households until at least 2064, according to the government’s spending watchdog.
The National Audit Office (NAO) has warned that although the potential benefits of the Sizewell C nuclear plant are considerable, they remain uncertain. The risks, however, are “immediate, substantial and borne by the public”.
The government claims the nuclear reactor, expected to generate the equivalent of enough low-carbon electricity to power 6m homes when it begins operations in the late 2030s, could save £2bn a year from the electricity system compared with using other low-carbon technologies.
However, for households the overall savings could be outstripped by the cost of supporting its construction until almost halfway through its 60-year operational life. The project could take even longer to “break even” if there are cost overruns or delays, the NAO warned.
“Sizewell C is a project of exceptional scale, complexity and significance for taxpayers,” said Sir Geoffrey Clifton-Brown, the chair of the public accounts committee, which oversees the work of the NAO. “Experience from comparable nuclear projects in the UK and overseas highlights their vulnerability to delays and cost overruns.”
Sizewell C is being developed by French state nuclear company EDF as a successor project to the Hinkley Point C reactor in Somerset, the first nuclear plant to be built in the UK in a decade. It has invested £1.1bn to take a 12.5% stake in the project alongside the UK government, which has invested £14.2bn as the majority stakeholder.
British Gas’s parent company, Centrica, owns 15% of Sizewell C while the Canadian pension fund La Caisse and the investment fund Amber Infrastructure own 20% and 7.6%, respectively.
Nigel Cann, the chief executive of Sizewell C, said the cost on household bills were an “investment in lower long-term electricity costs” which will “deliver value to consumers and to the country for the rest of this century”.
The project is already adding value to the UK by creating thousands of jobs and boosting businesses across the country, Cann added. Sizewell C said it had so far come good on its promise to source 70% of its construction value from UK suppliers and had spent just under £5bn.
“All major infrastructure projects involve uncertainty, and the report highlights the steps we’re taking to reduce risk and control costs,” Cann said.
A government spokesperson said investing in large-scale nuclear power was the “only way to get our country off the rollercoaster of volatile global gas markets”.
Households began paying for the Sizewell C project via home energy bills at the start of the year to help fund construction. This financial framework, known as a regulated asset base model, is a marked change from the Hinkley Point deal, which will begin to earn a guaranteed stream of revenues from home energy bills only once it begins generating in the early 2030s.
Critics of the regulated asset base model, including the campaign group Stop Sizewell C, have warned that any construction delays could mean that bill payers support Sizewell without receiving power for longer than expected, while the government would be on the hook for the project’s financial risk.
Stop Sizewell C said the risks surrounding the project “could easily turn Sizewell C into a financial disaster” while the funding model meant its investors were “the only ones who can’t lose”.
The NAO has urged the government to mitigate the risk by using “close monitoring, greater transparency to parliament, and by securing value for money from the significant public and private investment”.
Four leading AI models discuss this article
"Centrica's 15% exposure to Sizewell C carries outsized near-term downside from cost uncertainty and delayed breakeven that the RAB model does not fully mitigate."
The NAO report flags immediate public-borne risks on Sizewell C's £38bn build, with breakeven for households potentially delayed past 2064 under the RAB model that already levies charges on bills. Centrica (C) holds a 15% stake alongside EDF and others, so overruns or political pushback could hit its returns even as the project claims 70% UK sourcing and thousands of jobs. Unlike Hinkley Point, early revenue collection shifts timing but not the core vulnerability to delays seen in comparable nuclear builds. Strategic value for UK baseload power exists, yet the front-loaded costs create near-term pressure on private investors.
UK government backing for energy security could override NAO concerns with extra funding or guarantees, letting Centrica capture stable long-term cash flows once online in the 2030s amid rising nuclear demand.
"The NAO is right that households lose until 2064, but the real risk is technological obsolescence—not cost overruns—making Sizewell a 2030s solution to a 2040s problem."
The NAO report is damning on *household economics* but oddly silent on *system value*. Yes, bill-payers break even in 2064—but the £2bn/year system savings (vs. alternatives like offshore wind + storage) is real and frontloaded. The regulated asset base model is genuinely risky for taxpayers if delays occur, but it's also why EDF and Centrica signed on: they're insulated. The article conflates 'bad deal for households' with 'bad project,' which aren't the same. The real risk isn't cost overruns per se—it's that by 2038, grid decarbonization via renewables + batteries becomes so cheap that Sizewell's £38bn looks like sunk-cost thinking. That's the unstated elephant.
If Sizewell actually delivers on time and cost (Hinkley Point's track record notwithstanding), the 60-year operational life and baseload certainty could prove far more valuable than the NAO's 2064 breakeven suggests, especially if intermittency costs or grid stability premiums spike.
"The RAB funding model forces taxpayers to bear the 'immediate, substantial' risks of construction while deferring any potential consumer benefits for over four decades."
The NAO report confirms a classic infrastructure trap: the Regulated Asset Base (RAB) model effectively offloads construction risk onto the consumer while privatizing potential upside for equity holders. While the government claims this creates energy security, the 2064 break-even horizon is an eternity in capital markets. The real risk isn't just cost overruns—it's technology obsolescence. By the time Sizewell C goes online in the late 2030s, the levelized cost of energy (LCOE) for modular reactors or long-duration storage may render this massive, rigid asset a stranded cost. We are essentially forcing households to subsidize a 20th-century solution to a 21st-century energy volatility problem.
If natural gas price volatility remains structural, the 'insurance premium' paid by households via the RAB model could actually prove cheaper than the alternative of periodic, massive spikes in wholesale energy costs.
"Sizewell C remains a credible hedge against gas-price volatility and a backbone for UK energy security, provided policy stays supportive and cost overruns are contained."
NAO's caution is meaningful but not fatal. Sizewell C's £38bn price tag and 60-year horizon will always look alarming, yet the project is funded through a regulated asset base, which shifts early-stage risk away from taxpayers and uses predictable returns to attract capital. If the policy stance remains intact, the long-run savings from avoiding gas-market volatility and carbon-price spikes could still materialize, especially as offshore wind and solar scale up. The missing context is how sensitive the '£2bn/year' saving is to fuel prices, carbon prices, and construction timing; all of which could swing the cost-benefit materially.
The strongest counter: the NAO's caution is warranted because the RAB framework exposes bill-payers to cost overruns; if construction slips or capex overruns push the bill higher, the claimed long-run savings could be delayed or diminished, undermining the case for energy-security gains.
"RAB's early levies create prolonged cash flow risks for Centrica that outweigh obsolescence threats."
Gemini's focus on technology obsolescence overlooks how the RAB model's early bill levies create immediate cash flow drags for Centrica's 15% stake, potentially eroding returns before any 2030s online date. If construction mirrors Hinkley Point delays, the £2bn annual system savings cited elsewhere could be offset by higher consumer pushback and policy reversals. This front-loaded exposure amplifies downside for equity holders beyond stranded asset concerns.
"Centrica's early cash drag under RAB is a feature, not a bug—it's how private capital gets compensated for construction risk, not evidence the deal is breaking."
Grok conflates equity returns with system economics. Yes, Centrica's 15% stake faces front-loaded bill drags—but that's a *private capital problem*, not a project failure signal. The RAB model was designed precisely to absorb that drag. The real question Grok sidesteps: does Centrica's IRR still clear hurdle rates post-delays? If yes, the equity holders aren't actually distressed; households are. Those are separate conversations.
"Political risk will override the RAB contractual protections if construction delays cause public backlash against energy bills."
Claude, you’re missing the political feedback loop. You treat the RAB model as an isolated financial contract, but it’s a political instrument. If household bills spike due to construction delays, the government won’t just let Centrica earn its IRR; they will face immense pressure to claw back those returns or cap charges. The 'insulation' you describe is a thin veil. When the public perceives a 'bad deal,' the regulatory framework becomes a target, not a shield, for equity holders.
"Regulatory clawbacks or caps on RAB charges are the real downside risk to Centrica's 15% stake, potentially eroding returns far earlier than the 2030s."
Responding to Gemini's political-loop point: the real, underappreciated risk is regulatory counter-measures revisiting the RAB, not just bill spikes. A clawback, cap on charges, or windfall-style adjustments could compress Centrica's IRR even if online returns look fine on paper. The panel treats 'insulation' as given; in practice, political risk could erode the 15% stake's economics well before the 2030s. That nuance matters for equity hedges.
The panel consensus is bearish on Sizewell C's £38bn nuclear power project, citing front-loaded costs, potential construction delays, and risks of technology obsolescence. While the project offers strategic value and system savings, the long break-even horizon and potential political pushback pose significant challenges.
Strategic value for UK baseload power and potential long-run savings from avoiding gas-market volatility and carbon-price spikes.
Technology obsolescence and political pushback due to front-loaded costs and construction delays.