Nuclear has a role to play in providing reliable carbon-free electricity, but in the US and Europe it is likely to be too expensive to have more than a small role.
As I have written two books about nuclear power, people sometimes assume that I must be pro-nuclear. Well, suppose I had written two books on plague, it presumably wouldn’t mean that I thought disease was a good thing. I’m neither intrinsically pro- nor anti-nuclear, it all depends on the circumstances.
Nuclear makes up about 10% of global electricity generation and 4% of global primary energy (this data, and there is much more, come from the excellent resources at Our World in Data). Total production has been roughly flat for the last 20 years, but that masks a large increase in China and India, offsetting the closure of nuclear in Germany and the retirement in several other countries of older units.
The case for nuclear has largely rested on its low carbon footprint (even taking into account the large amount of concrete and steel used in construction, over its very long lifetime it has a CO2 per unit of electricity produced about the same as wind, and half that of solar PV, according to the IPCC).
But the Russian invasion of Ukraine also drew attention to the energy security case. Uranium is relatively easy to get from several reliable sources (1) and there is no likelihood of any uranium shortages for the foreseeable future. Bangladesh, Egypt and Turkey are all now building their first nuclear reactors. Some 30 other countries are considering new nuclear programmes, across Europe, Asia, Africa and Latin America.
And nuclear is among the safest forms of electricity generation (see the chart below).
But in the US and UK, the outlook for nuclear is more difficult, and it’s not because of worries about safety.
It’s the economics
In the past, when nuclear tended to be dismissed as irrelevant or foolish, I did try to put the case for new nuclear power in some situations. I think it is a good thing that in many Asian countries, where new nuclear power stations can be built economically and reliably, new nuclear is an important part of the current and future low-carbon electricity supply. In the UK, a densely populated country with many competing uses of land, it makes sense for nuclear to be part of a diversified energy mix, up to a point. I’ve contributed to several conferences on nuclear energy and recently taken part in a series of workshops for the Nuclear Energy Agency, part of the OECD.
But in recent years I’ve found myself having to tell a new wave of pro-nuclear enthusiasts that unfortunately, current nuclear technology is just too expensive to play a major role in the US and the UK (I’ll come to newer designs a bit later). The news that the first new nuclear reactor to be built in the US for 7 years has come on line has been greeted by some as the beginning of a new wave of nuclear stations. And if that is not the case, these enthusiasts say, why on earth not? Why don’t governments just get on with building lots more?
That new nuclear reactor is Vogtle 3, in the state of Georgia. The good news is that Vogtle 3 and its sister plant Vogtle 4, which should come on line later this year or early 2024, will produce low-carbon, safe electricity for the next 60 years, maybe more. The bad news is that the construction cost of $30 billion is far ahead of the original $14 billion budget and the reactors are seven years late. Another project using the same Westinghouse AP1000 design, in South Carolina, was abandoned after spending $9 billion. The Westinghouse company, then owned by the Japanese conglomerate Toshiba, went bankrupt. China managed to build the AP1000 more quickly but still behind schedule. A Chinese nuclear executive told me several years ago they would never buy that design again.
In Europe, the European Pressurised Reactor (EPR), a French design (originally French-German) has had an even worse construction record than the AP1000, being hugely over budget and behind schedule. The first EPR to start construction, in Finland, started operations in 2023, compared with the planned start of 2009. The second plant, in France, is 10 years behind schedule, and counting). The costs in Finland and France are a multiple of the original budget.
The twin EPRs under construction at Hinkley Point in the UK are one year behind schedule, mainly a result of the COVID pandemic. China managed to build this French designed reactor more quickly and much more cheaply than the French themselves could, with the world’s first two EPRs both starting operation in China, though the first one had to be closed for a year in 2021-22 for repairs. The Hinkley plants have been modified in light of what was learned in China.
Why China is so much better at building nuclear stations (along with the Japanese and South Koreans) might tell us a lot about what has gone wrong with the construction industry and process in the West, but doesn’t change the facts: the US and Europe have a serious construction cost problem in building conventional, large nuclear stations. A Chinese construction engineer from CGN, which owns one third of the Hinkley Point C EPR project, told me on a visit to the UK that he was baffled and dismayed by the inefficiency and short termism of the UK construction industry. Having Chinese help on Hinkley makes sense given their evident expertise in building nuclear stations. But it is obviously now very unlikely that we will see further Chinese involvement in the UK nuclear industry. Perhaps we can learn instead from South Korea, the only other politically acceptable nation with a good track record in nuclear.
The reasons for the poor construction record are a mixture of the designs themselves (the French intend to “optimise” the EPR design for their planned additional new reactors, styled EPR2) and the wider problems of civil engineering construction in the US and UK. There is research suggesting that construction productivity in the US has actually fallen. Costs for projects such as subways and railways are far higher in the US even than Europe, let alone China. On top of this, the process of getting permission to build in the first place is long and expensive, a problem that also applies to transmission lines.
If only the cost could be brought down, perhaps by a sequence of projects through which the US and UK re-learn how to build them, nuclear would offer one key advantage in a low-carbon system that is otherwise likely to be dominated by renewables, mainly solar and wind. That is, it offers continuous and reliable power, regardless of the weather. (Though one might point out that French reactors that use river water for cooling have had to close when low rainfall has lowered the river volume, making it impossible to safely discharge the waste heat into the river. If Europe faces more frequent summer droughts, this could permanently reduce French nuclear supply. British reactors are all on the coast.)
By contrast, solar is by definition only available half the time, at best. Wind-generated electricity is only available when the wind blows, which is less than half the time, especially on land, and supply varies in a less predictable way than solar.
So it might seem that a combination of nuclear and renewables is good, replacing what otherwise would be a lot of costly storage. But it’s not ideal. The most complementary electricity source for a renewables system would be a flexible and controllable one that could increase energy supply just when solar and wind were low, and then be turned down when they’re plentiful. Conventional big nuclear plants are not flexible in this sense, both in an engineering sense and in an economic sense (French nuclear plants do “load-follow” to some extent during the day, but this is different from the short term flexibility to offset hours of low wind or solar output). Nuclear economics depend on maximising output and running continuously. That is useful to have as part of an otherwise less than reliable system, but doesn’t solve the renewables intermittency problem (2).
A new hope?
What about the new nuclear design, small modular reactors? These are new designs (the International Atomic Energy Authority – IAEA – covered 70 designs in a report in 2020) based on two principles.
First, they’re small (up to 300MW electrical capacity) compared to traditional “big” nuclear (at least 1,000MW or more), and could in principle be used in a much wider range of cases, such as supplying just one industrial site.
Second, they are produced to a standard design, in a factory and delivered to the site on a truck or ship. Construction costs, the bane of current nuclear projects, would be low and predictable. Economies of scale, plus the learning-by-doing that has been such a powerful driver of lower costs in manufacturing industry in the last century or more, would lead to increasingly lower costs.
SMRs could also be flexible, or at least that’s what the UK National Grid assumes in its 2023 Future Energy Scenarios, all of which contain nuclear, mostly SMRs. SMRs have had financial support from governments in the US, UK, Canada, France and China. India is also looking at SMRs. The British company Rolls-Royce has used its nuclear submarine expertise to develop an SMR, which the company even claims would be “architecturally beautiful”. You can see its characteristically well-engineered promotional video here.
The only problem is that SMRs don’t actually exist yet. There are some small reactors around (they have routinely been used in ships and submarines for 50 years) but the idea of a modular system-built reactor is still to be proven. The first true SMRs in the West should be completed by the end of this decade. Then we will have some actual data on which to base a view of whether they can deliver a much cheaper, and ideally more flexible, form of nuclear power.
Meanwhile China, which is in most respects now the leading civil nuclear nation, has its own SMR programme, with the 125MW demonstration ACP100 due to be in operation in 2026.
I very much hope that SMRs do work as planned but I think we have to be a bit cautious before making plans for their widespread deployment. The British government is committed to procuring SMRs to meet its ambitious targets for new nuclear investment in the UK, on top of the additional EPR that is planned for Sizewell in Suffolk. There might be one or two more big nuclear plants beyond Sizewell, perhaps at Wylfa in Wales, but it’s not at all clear at this stage.
So nuclear is likely to be an important part of the low-carbon electricity system in several countries, including China and India, where the more coal use can be reduced, the better for humanity. It will remain central to French electricity production, which explains why that country’s carbon emissions per head are only slightly more than half those of Germany. But we will have to wait and see whether SMRs live up to their potential, for nuclear to have a significant role in the US and UK.
(1) One of these used to be Niger, where a military coup has rather changed the risk perception.
(2) If surplus electricity could easily and economically be stored or converted into some other energy source such as hydrogen, then this problem would disappear.