Nuclear power: no, yes, maybe, but not like this
This is a story in three acts, where we go from “trying to procure more nuclear power plants in 2024 is nuts”, to “I could see why you’d want some nuclear”, to “but if so, not like this”. This post has been quite a trip to write, where I rediscovered that writing something down is an ACE way to find out you didn’t know what you were talking about. It was also a good exercise in changing my mind a few times.
Before we start, I’m grateful to my proofreading friends, who make all the difference. I also want to thank multiple industry insiders, including from the nuclear world, for their valuable insights. It is amazing how hard it is to see our energy future, and it is good to know that very clever people are thinking about it non-stop.
It is also entirely possible that this post, which has been rewritten a number of times already, might turn out to need a further rewrite soon. Energy really is that hard to predict in 2024. I allude to a possible Act 4 at the end of this post. But here’s my best attempt right now, from a mostly European perspective.
Note: there is an FAQ at the end of this page that should cover many initial reactions or objections to this post. There are also copious links in the text to supporting documents, and even some earlier articles and podcasts. While I love to get your feedback (bert@hubertnet.nl) on if I got it wrong in this post, please do read the FAQ first or click on a few links.
![Nuclear power plant in Doel, Belgium, (C) Boudewijn Huysmans, via Unsplash](/articles/boudewijn-huysmans-IxUqE4nJ29I-unsplash.jpg)
Nuclear power plant in Doel, Belgium, (C) Boudewijn Huysmans, via Unsplash
Act 1: “it is nuts to try to order more nuclear power right now”
If you could get a new nuclear power plant today, that would be wonderful. As of 2024, only nuclear power can deliver carbon-free energy day and night, and in every season. Wind and especially solar are now more than able to provide us with all the electricity we need, at astoundingly low prices. But they can’t supply you with power if there’s no wind and no sunlight.
Nuclear can do all this 24/7, but that is about its only saving grace. Nuclear power comes with lots of baggage. You need to get rid of your nuclear waste periodically, which is tricky and causes (individual) downtime. And eventually you need to dismantle the radioactive plant, which is astoundingly expensive. Then there are of course safety/security concerns, especially in times of war. This all makes nuclear power quite the hassle. You would not do nuclear if you had any alternative way of getting carbon-free power at night or during windless weeks. New nuclear energy is also tendered at eyewatering prices compared to alternatives.
Various governments are currently attempting to order up new nuclear power plants anyhow, or are even already somewhat further down the procurement process. I think they are mostly doing so because getting a new nuclear reactor today would be very useful to rapidly decarbonise your economy (somewhat).
But we should not focus on what such a plant could achieve today. The painful truth is, you won’t get a new nuclear plant before 2038 (taking into account the years you’ll spend on tendering, permitting, lawsuits, financing, construction, commissioning). And by then things look entirely different.
Well before 2038, we’ll have other ways of creating carbon-free power whenever we want. Extremely rapid improvements in battery storage and especially hydrogen production mean that nuclear will by that time no longer be the best or even an acceptable option.
According to people who should know, multiple technologies are around the corner to do long duration energy storage (LDES), chief among these three different ways of converting excess solar and wind power into hydrogen. In effect this allows us to build lots more solar and wind than we need at any specific time, and use the excess electricity to make green hydrogen.
This green hydrogen can then be stored in nearby caverns, and later be used in gas fired power plants to also create (somewhat expensive) carbon free electricity, on demand. This would require storing dozens of gigawatt days of energy (many hundreds of TWh).
Turning electricity into gas and then turning it back into electricity is upsetting to engineers and people into thermodynamics since it is not a very efficient process. But if you look at it differently, electrical power around noon (outside winter) by 2038 will have no monetary value anyhow, so the point about efficiency is a bit moot. Also we weren’t in this to save money, but to keep the lights on.
Governments meanwhile already firmly believe in the ‘electrolysers’ that will create this hydrogen for us, and feature these in their 2030 climate strategies. Many other institutes also see this happening.
This means that once a newly ordered nuclear power plant comes online in 2038, we can be sure it will (soon) have to compete with this green hydrogen, which comes with a lot less hassle and is likely cheaper. This will probably already become apparent even while building a new nuclear plant, and I suspect many will be abandoned or never powered up. Or perhaps never refueled once the initial uranium runs out.
So in short, it appears to make no sense for governments to be attempting to get people to build new nuclear power plants right now, because by the time they come online, they are no longer needed to solve the problem of night-time, all-weather energy supply. Hydrogen can solve that for us.
But wait.. maybe things aren’t as simple as this.
Act 2: “Perhaps we do need (some) nuclear: it is all about resilience”
In our glorious carbon-neutral future, almost everything will run on electricity. Heat pumps will allow us to heat buildings using a fraction of the energy this cost us previously. Also, most ground based transport will be electrical, which saves lots of energy. Ever more of our industry will shift to electricity or hydrogen powered heating, hydrogen that we made from electricity. In addition, chemical processes that now use oil will shift to using hydrogen and (confusingly) carbon-dioxide.
In other words, if in that future our electricity fails, we are totally hosed. And in winter, quickly also rather cold.
If you feel responsible for keeping everyone warm and fed, you might appreciate having 10% - 20% of your electricity production to be always on, and not dependent on (foreign) hydrogen caverns and long distance HVDC electricity cables to countries far away. Nuclear gives you that.
Diversification is a nice thing to have, because it brings you resilience. And although in Act 1 I argued you really don’t need nuclear, and that it will be inferior to almost everything, it has the specific thing going for it that it is self-contained and, well, different. That by itself is interesting because it means that if something is wrong with your hydrogen infrastructure, you have at least something left in winter.
And this then brings into focus what you want to get out of your nuclear power plants. Not just energy when there’s no wind or sun, but also a cast iron supply of energy. This implies investing in multiple nuclear power plants that do not rely on common infrastructure, like cooling or grid connections. Also, multiple sources of uranium fuel, and especially not just Russian ones. Similarly, you’d want to not skimp on supporting infrastructure - we’ve already established that this is not about making a profit.
It is about having options, to enhance resilience.
Additionally, now that the actual thing we are worried about is diversification, we should extend that concern to our non-nuclear infrastructure. Apparently we’re willing to burn hundreds of billions of euros on nuclear. This means that similarly we should be able to spend that kind of money on extra HVDC connections, fully separated hydrogen infrastructures, far far far more wind, (pumped) hydro, batteries to tide us over any mishaps and perhaps even keeping alive natural gas or even coal infrastructure that we never use except when things are broken.
Act 3: “But not like this”
As noted, the nuclear industry currently takes 10 billion euros and 15 years to possibly deliver one working nuclear power plant, delivering a paltry 1 GW. There are plenty indications that these numbers might easily be 15 billion euros and 20 years, or worse. Also, there are almost no successful recent examples, and lots of terrible ones. Almost no one in the industry has a track record, and even companies that do claim they’ll need 16 years to deliver working plants.
To be clear, I do not hate nuclear power. In fact, I studied physics and I love the idea of nuclear power. I’ve visited multiple nuclear power plants, I cherish my copy of Introduction to Nuclear Engineering, and I find it very exciting stuff. I know there are (bad) reasons behind why it takes to long and so much money to build a new nuclear power plant these days. But that these reasons are bad doesn’t mean we can solve them in short order. There are no magic fixes. As military people like to say, you go to war with the army you have, not the army you want.
In the course of writing this piece, one former nuclear insider told me what we’d been doing wrong (in their view). The nuclear industry is failing because it is building these unique one-off reactors. If we want the industry to do better, we should do a bulk order of 50 reactors!
At first I thought this was a ridiculous idea. If the industry is failing, they should first prove they have their act together again before we pile on more money.
However, there is this thing called the experience curve. The more you do something, the more you learn how to do it well and cheaper. Currently, Framatome is the only surviving European supplier, and they are doing all these one-off projects in Finland, UK and France. Little of what they learn in each country is directly applicable to other countries.
We’re expecting this industry to learn “on the job” using single individual plants, and getting it right immediately. And they aren’t.
Currently many (European) countries are individually trying to order up new nuclear power, from many different places. But it appears we can’t treat nuclear reactors like (say) cars you can just procure. If we’d want to do this right, it is probably indeed better to not simply try to order stuff, but to engender a nuclear revival. To not simply point our fingers at Framatome and EDF and say “do better!”.
What if we actually made this a European or transatlantic project, and add the vast expertise that is still hidden within our institutes, and indeed setup a project for building 50 nuclear reactors, or more? This would allow a broad base of research that would derisk the process, so we don’t necessarily find out after 15 years of construction that the design is too complicated. And perhaps also not try to pretend that we are leaving this to the free market, but recognize this as a public activity.
Doing it like this would require governments, institutes and companies to think different, and I’m reasonably sure we can’t even get this done between a few like-minded countries. Most definitely the EU would not reach consensus on this, since Germany is fundamentally opposed to anything nuclear ever.
We might remind ourselves that the “golden age of nuclear” was in fact when we were building dozens of plants at the same time.
If you’d really wanted quality and reliable nuclear power, this would be the way to do it. But we’re almost certainly not going to.
Epilogue
In this play in three acts, in Act 1 we learned that nuclear is technically and financially not very interesting by 2038, and also a major hassle. But in Act 2 we found that betting on only wind, solar, hydrogen and some batteries might also not be everything. Especially if part of that infrastructure is far away from you. Having local nuclear might give you a lot of peace of mind in winter.
But in Act 3 we found that we are expecting miracles from the nuclear industry, miracles they can’t deliver if everyone ends up ordering a few unique plants sequentially & from different places. Things might change if we turn it into a massive project where we finance the development of 50 reactors, including a huge research commitment to make sure this works out well.
However, it seems unlikely this will happen, and that brings us back to the diversity/resilience from Act 2.
Instead, spend the nuclear trillions on other ways of getting diversity, like paying people to keep almost entirely unused natural gas power infrastructure around, and building a lot of redundancy in our hydrogen/gas/solar infrastructure, including copious long distance interconnections and lots of independent hydrogen storage and transport means.
Because if we don’t, we’ll end up with a small number of very late nuclear reactors, possibly concentrated in only a few places, surrounded by a non-resilient energy infrastructure and an anemic grid.
Act 4?
As alluded to above, writing this piece was a bit of a roller coaster, where as I learned more, my conclusions changed radically. In this post, we’ve diagnosed that one reason governments might be ordering new nuclear today, to arrive in 2038 at the earliest, is that if it arrived today it would be super useful. Imagining the future is really hard, and perhaps this article too is guilty of not seeing the obvious.
Because what we are trying to do is recreate the current energy landscape (“all the power, whenever you want it”) as precisely as possible, but now with wind, solar, hydrogen and some nuclear. But this is pretty expensive. Recently, there has been some new thinking - what if we give up on this, and deal with the fact that sometimes we have a lot of energy, and sometimes maybe you have to scale back industry and heating for a bit. This might save trillions in investments. Ensuring full energy availability for the worst 3 days of the year is surprisingly expensive.
As I wrote in 2014: “Meanwhile, wily entrepreneurs have cottoned on the new reality of highly variable energy prices, and are scheduling their business needs around them. Processes that lose money at 20 cents/kWh make a lot more sense at 3 cents/kWh! This is the future, and attempting to emulate the flat power prices of the past century is not.”
I would not be amazed if in 2038 we’ll be far more into shifting demand to when the energy is abundant, and that we’ll no longer be as interested in recreating the 2010 energy landscape.
Frequently Asked Questions
- Isn’t creating hydrogen and then burning it for electricity very expensive?
Yes, compared to solar energy, it most definitely is. The prediction is that this combination will eventually become competitive with fossil fuel based electricity. But do recall that we were in this not to save money, but to prevent carbon emissions. The correct comparison is with the eye-wateringly expensive (but carbon free) nuclear power. - Isn’t hydrogen useless?
Hydrogen is indeed overhyped, and there are many things no one should be using hydrogen for. Like powering trains, cars, or heating houses or transporting energy by ship. But for long duration energy storage, hydrogen is suitable. - The solution isn’t technology, we should simply use less energy
So I do agree to an extent, but you’d need to rejig your whole society and politics to make this happen. Meanwhile, heat pumps and electrification of transport do save whole factors in energy use. We’ll definitely end up using less energy. - But what about Small Modular Reactors?
Not one has ever been built so far. It turns out that a small nuclear reactor is not actually that attractive, as you’d need a lot of them. There are also quite fundamental reasons why it is an uphill battle - Ah but what about the BWRX-300?
Also has never been built, but there are plans to start constructing one in 2025. There are various prognoses as to when it will come online, between 2028 (!) and 2030. This would get you 300 MW of electrical power, which is not a lot. - Surely we could make cheaper reactors?
Well.. an efficient typical nuclear reactor combines 150 atmosphere pressure with high temperatures, high radioactivity and actual boric acid. It really is stupendously hard to make a reactor that is safe, efficient and reliable. Let alone cheap. Do realize that a 1 GWe reactor will have to move 5 GW of heat, which is A Lot. We won’t tech-bro ourselves out of this too easily. - Nuclear isn’t that expensive according to this study
Maybe, but do check their assumptions, like that the plant will be selling electricity for 50 years. Or that the cleanup will then be paid by someone else. - According to this study nuclear energy is the best energy ever
That could be true, and as noted the idea of nuclear energy could be quite attractive. But we don’t power the world with ideas but with concrete reactors, and as outlined above, these are unlikely to appear any time soon if we order them by ones and twos, and there is no political will to do anything different. - Renewables can’t supply the amount of energy we use today
This is called the ‘Primary Energy Fallacy’. Electrification makes transport and heating whole factors more efficient. In addition, a fossil power plant uses two to three times more input power than it generates as electricity. When taking this into account, there is no shortage of renewable energy. - What about pebble bed reactors, thorium, molten salt, (confinement) fusion etc
All very exciting, but nothing that will get us working nuclear reactors before 2038. It is hard enough already to complete a relatively boring pressurized water reactor within 15 years. - But what about the Chinese nuclear program?
This may not be what you think it is, as it turns out they too suffer from massive delays. In any case their appeal is limited since most countries do not want to buy Chinese nuclear power plants, and the last attempt to do so (in the UK) was not a success. - What about the CO₂ used for creating batteries, solar panels and wind turbines?
While it is good to take this into account, this consideration is becomining less and less important since the energy used for creating these things is by itself very rapidly becoming greener. Be especially careful when repeating 15 year old IPCC numbers on how much CO₂ is used for solar panels, these figures are absurdly out of date. It appears that solar panels now earn back their production energy within 3 years. Assuming they are still displacing coal/gas plants, that also makes them a CO₂ sink at that time. As a reality check, a 400Wp solar panel now costs less than 50 euros. This also means it can’t conceivably have cost more than 50 euros of energy to build it.
Further reading/listening
- Podcast with Jenny Chase and Michael Liebreich which discusses the current state of solar, and also the history of making bad solar predictions. Chase and Liebreich are titans in this field. They also touch on the “winter problem” also covered in this post.
- Redefining Energy Podcast on how Bill Gates and others are stuck in the past
- Dutch government position on new nuclear - stresses the carbon credentials of nuclear power, but also very much the diversity benefits
- Kernenergie: niet nodig, niet slim en niet te betalen, De Correspondent
- Liebreich: We Need To Talk About Nuclear Power, a 2019 visionary piece making many of the same points I am leisurely making in 2024. An interesting read since it has many concrete numbers and expectations from 2019, where we can compare these to what actually happened.