Exploring the Nuclear option

While spiralling budgets are likely to weigh on the UK’s vast nuclear ambitions, the value of more capacity cannot be assessed on construction cost alone.

• Despite problems with previous nuclear projects, there is a case for Sizewell C and potentially another gigawatt-scale project
• The government’s strategy to deploy up to 24GW of nuclear energy by 2050 is incredibly ambitious, and it is unclear whether a commitment of this scale makes financial or technological sense
• The National Infrastructure Council’s one-by-one recommendation for major nuclear projects is a sensible approach that avoids overreliance on a technology with a history of delays and complications
• Modular technology could well be the future of nuclear power and is worth pursuing. But without firmer details on design, regulation and cost, their benefits remain theoretical.

Across the gamut of electricity generation options the one most likely to inflame tempers in a room of clean energy advocates is nuclear power. The battle lines are firmly drawn. The two sides have names for one another, and they are not polite. Into this febrile atmosphere, the UK government recently announced the biggest increase in nuclear power generation for 70 years. 

At present, the country has around 6.5GW of that provides some 15% of total UK electricity. The Climate Change Committee’s balanced pathway to net zero assumes 10GW of nuclear capacity by 2035. Because the majority of the UK’s existing fleet is scheduled for retirement in the near future, the CCC assumes 80% of this 10GW is new-build capacity. 

The new strategy aims to deliver 3-7GW every five years from 2030 to 2044 and deploy up to 24GW of nuclear power by 2050. The nuclear industry is cautiously optimistic at this show of support – its critics are up in arms. 

The much maligned 3.2GW Hinkley C is still under construction. The strategy calls for a final investment decision (FID) on the proposed 3.2GW Sizewell C nuclear plant by the end of this parliament, and exploration of a third large-scale nuclear power project. Even if Sizewell C and a third gigawatt-scale plant are greenlit, hitting 24GW by 2050 is phenomenally ambitious. Tom Greatrix, CEO of the UK Nuclear Industry Association, says hitting that target puts the focus on three things: siting, technology and construction order. 

But first there is the question of whether the UK should be building nuclear at all.

Arguments against nuclear power are well worn, but many of them are flimsy. Critics say it is unsafe – it is not. Critics say there is no solution for nuclear waste – but there are several, and the volumes are not problematic. The life-cycle emissions of nuclear are – if anything – lower than that of solar and wind. Nuclear power has the additional advantages of providing firm low-carbon electricity with a far smaller land-use requirement than its renewable peers. The one area where critics are on firm ground is cost.

One study found the mean cost overrun for global nuclear power projects was 120%; solar and wind were 1% and 13%. 

Time is money 

Hinkley C’s Unit 1 was originally due to come online in 2025 – it looks likely to complete in 2031. The project was originally due to cost £16bn – that figure could now be up to £35bn. Cost and construction time are of course connected – the longer a project takes the more it costs. But Hinkley is not an outlier in the world of nuclear construction. Bent Flyvbjerg, professor of major programme management at Saïd Business School, found the mean cost overrun for nuclear power projects worldwide was 120%. The respective figures for solar and wind were 1% and 13%. 

Largely for reasons of time, when it comes to gigawatt scale nuclear the UK’s National Infrastructure Committee only expects Sizewell C and one other project to come online before 2050. “We’re perfectly supportive of nuclear power,” Julia Prescot, deputy chair of the NIC, tells E-FWD. “The real issue is how realistic we can be not only in terms of timescale, but in terms of investment.” 

The nuclear industry warns against judging the viability of future projects on the basis of Hinkley C – the first UK nuclear power plant to start construction since 1988. Making sure the EPR design satisfied UK regulatory standards required thousands of changes. The project had to create new supply chains, train a new workforce and essentially rebuild the UK’s capacity for nuclear construction.

The financing structure was also sub-optimal. Hinkley C’s CfD structure protects consumers from cost overruns because the project only receives payment when it starts generating. But because the investors shouldered all the construction risk, their required rate of return was a very high 9%. 

This high cost of capital was a key determinant of the agreed strike price – set at £92.50 per MWh in 2012 and rising with inflation to over £125 per MWh as of 2022. The strike price for offshore wind – even after a 66% hike in November 2023 – is only £73/MWh (and closer to £100/MWh when adjusted for inflation).

Surging energy prices in the wake of the Ukraine crisis have cast the Hinkley C strike price in a slightly better light. Nor does Hinkley’s strike price look terrible compared with floating offshore wind, of which 5GW is required for the government’s energy security strategy. In late 2023, the government set a strike price of £176/MWh for that technology. 

A better approach 

As part of its new nuclear strategy, the government has shifted to a regulated asset base (RAB) model. This will apply to Sizewell C, where the percentage return will be set by Ofgem. Consumers will be exposed to construction costs, but so too will equity holders who have an incentive to constrain costs. The RAB model pays off interest in real time so the costs do not build up. The RAB, says the government, will save consumers billions compared with a Sizewell C funded through a CfD. A fixed, reliable return on a green energy project should open up investment to institutional investors like pension funds.

Julia Pyke, joint managing director at Sizewell C, estimates the cost of the nuclear kit will be roughly 25% cheaper for Sizewell C than for Hinkley.

Sizewell C will still be expensive – estimates range from £20bn-£35bn – but the nuclear industry is confident that building projects in a coordinated, programmatic fashion using largely identical designs and technology will significantly reduce cost. The regulatory changes have been made and the reactor design is established – Sizewell C is essentially a copy.

Nuclear New Build Generation Company – the EDF subsidiary created to operate Hinkley Point C and Sizewell C – points to improvements already visible between Unit 1 and Unit 2 at Hinkley. Construction of the concrete foundation slab on Unit 2 took six months – down from 10 months for Unit 1. Welding for the steel water pools that house the reactor vessels is moving four times faster for Unit 2. The inner containment layer around the Unit 2 reactor was completed in the half the time that it took for Unit 1. “We’ve got a UK case study of the difference between doing the first one and doing the second one,” says Greatrix. 

Julia Pyke, joint managing director at Sizewell C, estimates the cost of the nuclear kit will be roughly 25% cheaper for Sizewell C than for Hinkley. “Because we’ve taken out paperwork aspects – we don’t have to re-qualify, re-demonstrate or re-design all the same pieces again,” she says.  

My models versus yours 

Not only does the nuclear industry believe it can lower costs for Sizewell C, it also decries the focus on capital and construction costs to the exclusion of what nuclear would do to consumer bills. Pyke says the additional cost to energy bills from the construction of Sizewell C will be measured in pence, rising to around £2 per household per month at the height of construction.

“Once it’s generating it will save consumers around £1.5bn each year – and that’s in ordinary times,” she says. “In an environment of Ukraine crisis-style gas prices it could save consumers multiple times that figure. Ultimately, it’s cheaper for consumers if we build nuclear power than if we don’t.”

Modelling the outlook for electricity prices in a landscape of rapid technological and regulatory evolution is complicated and the limitations are significant. But there is independent modelling that supports the argument that nuclear power will lower bills and emissions.

In October 2023, Aurora Energy Research provided modelling on average consumer bills in a 2050 net zero scenario for the NIC’s second National Infrastructure Assessment. The firm presented eight theoretical scenarios to illustrate how policy and technology choices affect consumers. 

In Aurora’s ‘base case’, the government hits its energy security strategy targets for offshore wind and solar and there is a high proportion of flexible capacity. Emissions fall to 2.6 MtCO2 by 2050 and between 2025 and 2050 consumer bills average £133 per MWh. Nuclear provides just 8GW of total electricity, which comes from extending the life of Sizewell B and building Sizewell C and Hinkley C.

In Aurora’s high-nuclear scenario – where the government manages to deploy 24GW of nuclear energy by 2050 – consumer bills are £126 per MWh – the second lowest of the eight scenarios. By reducing the need for gas plants with CCS, emissions fall even further to 2 MtCO2. But high nuclear energy also requires the highest levels of upfront capex expenditure at £429bn. Nor does the modelling make allowance for the challenges delivering nuclear on time and on budget

Consumer bills are even lower – £125 per MWh – in a ‘high flexible demand’ scenario. Here, nuclear, wind and solar are the same as in the base case, but Aurora assumes market penetration of 92% for smart charging EVs and smart heat pumps by 2050. This scenario also has the lowest upfront capex costs at £369bn. 

Of course, these scenarios are theoretical not probabilistic. How fast progress could be on alternative technologies is a matter of debate. The CCC’s balanced scenario also assumes certain adoption rates for electric vehicles and heat pumps. In 2022, data from the European Heat Pump Association indicated that at the UK’s current rate of heat pump installation, it would take the country 600 years to meet the CCC’s target.

Then there are modelling disagreements. Energy Catapult has claimed much modelling in the energy sector is overly kind to nuclear, failing to take into account cost and delays. A lack of granularity in modelling techniques, they say, also falsely benefits nuclear and wind power, underestimating the problems of combining these two technologies. Energy Catapult warns that adding both more nuclear power and more renewables will likely lead to higher costs for consumers and much higher levels of constraints for wind generators. 

Security and skills 

Nuclear advocates can point to other advantages for their technology of choice – one of which is energy security. Access to nuclear fuel is a key point, but the UK has a nuclear fuel factory and Framatome – which provides fuel for Hinkley – recently announced plans to set up a nuclear fuel fabrication facility in the UK.

Renewable advocates have called for higher domestic content in areas like onshore and offshore wind, where the vast majority of materials are not made in the UK. Pyke says Sizewell C will have 70% domestic content during construction and 90% throughout the project’s life. A bigger UK fuel industry would make that figure even higher. 

The UK’s nuclear plans are not reliant on the commitment of foreign entities. Sizewell C is no longer in the EDF Group. After a planned capital raise, the UK government is expected to hold around 50% of the project, EDF will hold no more than 20% and incoming investors will hold the remaining balance. 

In a world where politicians are under pressure to ensure the energy transition does not take away jobs, the nuclear industry can point to serious investment in skills. Some 1,300 workers have been trained as part of the Hinkley project. Sizewell C has committed to training another 1,500.

But an important distinction is that the case for Sizewell C – and potentially one additional gigawatt-scale plant – is not the same as the case for trying to build-up 24GW of nuclear power by 2050. When it comes to large-scale nuclear projects, the NIA recommends a “one by one approach” that would maintain the new UK skills base without over committing to a technology with a bad track record of delivering on time and on budget. Whether the UK pursues another gigawatt-scale nuclear plant after Sizewell C should depend largely on how well that project goes. 

The government’s strategy will also explore Small Modular Reactors (SMR). Great British Nuclear – a government-owned entity – shortlisted six designs late last year as part of the SMR technology selection process. SMRs may or may not emerge as a viable option. Building smaller, more-modular and less time-intensive projects could lead to faster cost reductions and more widespread adoption. But the sector has been plagued by broken promises and misplaced hype. 

Still, SMRs are worth pursuing – in a cautious, considered fashion – for their potential benefits. And if the technology does prove viable, then a nuclear skills base and established supply chains would be useful.

Ultimately, nuclear sceptics are right to demand solid proof of progress and efficiency – the industry knows it has to deliver when it comes to Sizewell C. But critics should acknowledge that the argument for or against more nuclear power is about much more than construction cost or strike price. 

How much weight one gives to energy security or a domestic skills base is a question of personal values. Confidence that other emerging technologies can deliver net zero faster or more cheaply should be carefully calibrated. The NIC and CCC both have hopes for green hydrogen and carbon capture as part of the net zero journey. But there is nothing like a consensus on how much hydrogen or CCS should contribute, how quickly those technologies will scale or how much they will cost.

The NIC’s one-by-one approach to large scale projects is sensible. Let the industry build Sizewell C and prove it can improve. Explore SMRs, but acknowledge the uncertainty.