Infrastructure Costs: Nuclear Edition

Building nuclear plants in Britain is more expensive than almost anywhere else

Britain used to lead the world in nuclear power. This is the country that split the atom, built the world’s first full-scale nuclear power station, and then proceeded to build nine more in the decade that followed. When Calder Hall was opened, Lord Privy Seal, Richard Butler, noted “It may be that after 1965 every new power station being built will be an atomic power station.”

But while Britain has built plenty of new wind, solar, biomass, and gas-fired power stations in the last 28 years, it hasn’t built a single new nuclear power station.

There are many reasons why this is true: from misinformation in the wake of Chernobyl around the dangers of radiation (on that, read this), to challenges in financing and complex planning rules. For example, the government initially waved ahead Chinese investment in Hinkley Point C, before later spending an undisclosed sum to buy out China General Nuclear’s stake. But, one of the main reasons why the promise of a fully nuclear grid never came to fruition is the cost of construction. Put simply, building new nuclear power stations in Britain isn’t cheap.

To try to understand what went wrong with nuclear here, Britain Remade have compiled a database of every nuclear power plant (that has reliable cost data available) constructed since 2000. In total there are 39 power plants from 16 different countries. Generating capacity of each plant comes from the International Atomic Energy Agency’s Power Reactor Information System database and all the cost data has been converted into GBP (£) and controlled for inflation, with all figures below in 2023 GBP.  

And just as we found when we looked at roads, railways, and tram systems, building nuclear plants in Britain is more expensive than almost anywhere else. Britain ranks 15th out of 16 countries by construction cost per MW of generating capacity. Only America performed worse. 

Most of the new nuclear power stations are being built in low-to-middle income countries. China is currently building 22 reactors, while India is building 8. 

But, it is an advanced economy that stands out. South Korea is able to build nuclear plants at roughly a quarter of the cost of the UK –and they build a lot of them. 

One of the most important differences in how South Korea constructs new nuclear power compared to Britain is their commitment to building fleets of new reactors. South Korea has two generations of nuclear reactors, both designed domestically. For each design, they build 8-12 reactors in a row. By building fleets, instead of individual reactors, they benefit from economies of scale and learn by doing. Construction becomes cheaper as a result, with valuable experience gained every time they build a new plant. 

Constructing a nuclear power station requires specialist technical skills in things like welding. EDF, who are building Hinkley Point C in Somerset, opened a nuclear welders academy because the UK’s skill base had atrophied in the decades since we last built one.

South Korea doesn’t have this problem. Private contractors were able to invest in skills and equipment over the long term secure in the knowledge that the government is committed to building multiple reactors.

Kepco, South Korea’s national nuclear champion, is even using the same nuclear reactor designs to construct four reactors in the UAE. Without a nuclear history and a tendency for flashy projects like the Burj Khalifa, the UAE seems like a perfect candidate for expensive construction. Yet even then, starting from scratch, the reactors will still be 2 1⁄2 times cheaper than Hinkley Point C. 

Of the 39 nuclear plants that have been constructed since 2000 with reliable cost data, Hinkley Point C ranks second from last. Only America’s Vogtle 3&4 cost more per megawatt of generating capacity. 

The Vogtle plant was being built when the Fukushima nuclear accident happened, which resulted in all work stopping as the plant’s containment building was completely redesigned. Then cost and delays led to Westinghouse, the reactor designer and project manager, declaring bankruptcy. The reactor was eventually delivered earlier this year, 7 years late and $16bn over budget.

Hinkley Point C uses the EPR-1750 (short for European Pressure Reactor with a capacity of up to 1750MW) reactor design, which has also been used in France, Finland, and China. The same design will also be used for Sizewell C in Suffolk. 

Amazingly, there is nearly a fivefold difference between how much it cost China to build the Taishan nuclear plant and the most recent estimate for Hinkley Point C’s cost, despite both using the same reactor design.

China’s cost of £2.1m per MW is cheap by global standards and may be explained by cheaper labour costs. Still, Finland was able to complete the Olkiluoto 3 plant for £5.97m per MW, and France will spend £7.24m per MW on Flamanville 3, while Hinkley Point C is set to cost £10.03m per MW. 

Reactor design is clearly an important factor in cost. But, the fact that Hinkley Point C is going to cost 39% more than the same reactor in France and 68% more than the same reactor in Finland should give us pause for thought.

It isn’t that France and Finland have been pursuing a fleet strategy like South Korea. In both cases they went at least 20 years without building a new nuclear power station. 

Likewise, it isn’t that Hinkley Point C is uniquely delayed. Finland’s plant took 17 years to build and France’s will likely also take 17 years if it comes online next year as predicted. These are among the worst performing 5% of nuclear reactors by construction time. Hinkley Point C’s currently estimated 9 year construction period is breezy by comparison. 

From Hannah Ritchie’s Sustainability by Numbers Substack

Where the differences do start to stack up is in the planning and financing stages. From being proposed in a UK government white paper, to the start of construction, Hinkley Point C took 10 years. In Finland, there was a 4 ½ year timeline between the electricity company TVO requesting that the government support a new reactor to construction starting. The actual application for a construction permit only took a year. In France, the timeline was even quicker. The board of the state electricity company, EDF, decided to support building their first EPR-1750 at Flamanville in the middle of 2004. After public consultation, this decision was confirmed in 2006. Construction then started in 2007.

It’s clear that our approach to planning and financing reactors adds significant costs. With a 44,260 page environmental impact assessment (EIA) and 2,229 written questions at examination stage, Sizewell C, the next reactor after Hinkley Point C, faced enormous expense before a spade is even in the ground. As the recent DLUHC policy paper Getting Great Britain building again pointed out, Sizewell C’s EIA is “more than 30 times longer than the complete works of Shakespeare.”

Were we always this expensive?

It wasn’t always the case that Britain was amongst the most expensive. For plants constructed before 1995, Britain averages £4.79m per MW. To put that in context, that’s half what British plants being constructed now cost and roughly average internationally. 

So what should we do about this? 

Some, like Greenpeace, have called for the complete cessation of nuclear power arguing that it is economically unsustainable. Effectively giving up and saying that the country that built the first nuclear power station is no longer able to complete big projects.

So should we listen to the naysayers and give up on nuclear energy as a costly technology of the past?

No, and there’s two good reasons why we should press on with nuclear power.

The first is the firm, reliable baseload that nuclear power provides is vital to decarbonisation. Nuclear power plants generate at full capacity 92% of the time. This reliability can underpin and complement a majority renewable grid.

Secondly, as seen above, building new nuclear power isn’t as hard or costly everywhere. By looking abroad, we can find solutions for many of the high cost factors of nuclear power construction that plague the UK. We can learn lessons from South Korea’s fleet strategy, which will reduce nuclear construction costs. To fully implement these solutions we’ll need to update our planning system so that we can approve the construction of different reactors in the same fleet at the same time, rather than repeating many of the planning steps for each identical reactor.

That’s why we should be open to the new wave of technology which is on its way. Small Modular Reactors (SMRs) have the potential to bring down the cost of new nuclear power as its components can be produced on production lines in factories and then assembled on site. Rolls Royce estimates that each of their 470MW SMR will require roughly £2bn of capital investment, significantly less than the £32bn for Hinkley Point C. At £4.4m per MW, this is roughly in the middle of international standards. Last Energy, which produces a 20MW micro-SMR suitable for heavy energy users, such as chemical plants and data centres, already has a project order book in the UK. In theory, these can be privately financed and require no state subsidies.  But as it stands, they have no route to get their reactor design approved and to find a site to build it on.

But when it comes to giga-scale nuclear power, there’s some good news. Blackpink may have stolen the headlines during South Korea’s recent state visit, but somewhat under the radar Energy Secretary Claire Coutinho signed a new Clean Energy Partnership that touted co-operation on civil nuclear capabilities. Also that week, South Korea’s Kepco signed a memorandum of understanding with Mott MacDonald to explore the potential of building a giga-scale facility in the UK.

High nuclear costs are not inevitable. If we can learn the lessons of other countries that have kept costs down, then Britain can still have a bright atom-powered future.