Britain plots atomic reboot as datacenter demand surges The UK is following the US in seeking to fast-track new atomic development, spurred on by the need to provide enough energy for its AI ambitions plus the increasing electrification of industry and vehicles. A report published Monday by the government's Nuclear Regulatory Taskforce claims that the UK is now the most expensive …
A former boss of mine was at Windscale (as it was then) during the 1957 fire and his first-hand account was sobering. I did a tour of Sellafield much later and was struck by apparently casual references to bits of the former plant that were walled up owing to previous leaks. We've seen in repeated, serious, incidents around the world that the technology has not so far been terribly forgiving and the cost of dealing with them undermines the economic case.
That's not to say technology hasn't improved and that the huge investment formerly required to build necessarily massive nuclear plants couldn't be obviated by small-scale reactors. But maybe we first need to seriously re-think who and what the electrical grid is for. We've taken it for granted that it was as necessary for supplying high energy-dependent industry as much as for keeping homes warm during the winter. Are we getting to a stage where it would actually be better for large-scale users to generate power locally themselves rather than transport it across the country?
A former colleague of mine in Belfast had also worked at Windscale at the time of the fire and told of going round the countryside collecting bits of radioactive debris using a fluoroscope to identify them. I think we've probably learned a bit since then.
I suppose it would have been about the same time that I was given a tour of the local mill. AFAICR the steam engine was being used to generate electricity. It was coal fired, of course.
The fire happened in a reactor dedicated to producing weapons-grade plutonium, not one of the civil power reactors
The British government has a bad habit of mixing up civilian and military nuclear installations which in turn results in military attitudes & safety levels being applied - which are extremely fast&loose compared to civil nuclear standards
Hence why nuclear waste was tossed down an old mine shaft.
"Small scale reactors" are subsidy sucking projects and they're mostly being pitched by outfits with a strong interest in military systems such as Rolls-Royce
You need enough steam to drive a 1GW turbine or the non-nuclear maintenace costs are very high (this is why coal/gas stations are almost all GW-class) and ideally you want that steam to be DRY or supercritical (at least 600C). LWR nuclear can produce 350C wet steam at best and that results in massive levels of turbine blade wear thanks to condensation inside the turbine causing pitting.
On top of that, LWR designs are essentially steam boilers with the hot bits actually immersed in the water - water is NOT a benign substance at these temperatuires and pressures (20-100 atm) and there have been several near misses due to corrosion. One of the nastiest was Bess-Davis in 2002 and it's worth looking up. the TLDR is that water ate its way through 9 inches of steel in 18 months, only being stopped by the pipework's chrome plating (what inspectors found inside the reactor was pretty bad too).
Engineering (burst) stresses on a boiler go up with the cube of the power. This is one of the reasons why railroads and factories moved away from steam as soon as alternatives were readily availabile - in the case of steam that meant moving to diesels only developing ~1/5 the power - 1500hp vs 8000hp)
Yes, Britian NEEDS to have a crash rollout program of nuclear power, but LWR is not the answer. The reason they're so stupidly expensive is that you need a massively strong reactor vessel (boiler) to handle the absurd energies being created (~3500MW) and last 60 years with zero maintenance, then you need an even more massively strong building to contain a potential burst.
There's also the weapons problem. LWR needs enriched uranium - which was used for the dual reasons that it was all that was available in 1953 and it allowed building a nuclear reactor small enough to fit in a submarine hull (Nautilus). The reason it was available at all is because enriched uranium is the unwanted waste product of making nuclear bombs - highly depleted uranium is processed into weapons-grade plutonium (which is what was being done when windscale caight fire). Nobody builds enriched uranium bombs because they cost hundreds of times more than plutonium ones (4-16 billion pounds apiece vs 20-60 million - weapons-grade U235 costs around £500 million/kg. Think about that when you realise there are 1-200kg in a submarine reactor. How much did those boats actually cost?)
If it wasn't for weapons nobody would bother with enriched uranium. It's perfectly feasibnle to build nuclear reactors that use natural uranium. You need more of it but not 500x more (natural uranium is ~$150/kg, 3% enriched "reactor grade" uranium is $65-80k/kg) - you get 9kg of depleted uranium for every 1kg of 3% enriched uranium and nobody's going to tolerate 89% wastage in a purely ciovil environment. In other words, where there's a reactor using enriched uranium, there's a weapons program lurking in the shadows and any handwringing about "enriched uranium" is a distraction from the real problem material.
Also: ALL designs using fuel rods are abuseable to make weapons plutonium. Fuel rods were originally used in the Manhattan Project's X10 plutonium reactors - whose design was tweaked by Alvin Weinberg to make the Shippingport/Nautilus reactors - as proof of concept laboratory demonstrators. Weinberg always envisaged something much safer for civil power and went on to achieve that in the ORNL MSRE - a design which isn't dependent on enriched uranium and was killed off in 1972 by Richard Nixon. Weinberg was kicked out of the USA Nuclear program..
Commercial operators dropped planning with LWR nuclear power in the early 1960s because it's more expensive than coal. It's even more expensive now. Weinberg's MSRE (LFTR) design was predicted to gost 20% of LWR to build and run, with 99% less waste and total immunity to the kinds of civil nuclear accidents we've seen over the decades (this was tested heavily in 1966-68). The most dangerous part of LWR nuclear power is the water and removing it makes things hundreds of times safer
Here's the fun part: at 80% cheaper than existing LWR, LFTR designs substantially undercut renewables. Britain will need ~65 nuclear plants eventualy. The reactoirs and containment are far smaller too - a reactor large enough to replace a coal station is 1/4 the size of the coal station's burners and hot enough to be drop=in replacements for those burners. It looks like China's been planning for this with its coal stations built over the last 20 years
Yes, the future is nuclear, but it will be driven by Weinberg's industrial prototype design (MSRE), not overgrown versions of his laboratory glassware one (Nautilus) - and right now the mid-late century leader looks like being CHINA (if things go the way I expect them to, they will be THE economic hyperpower from the 2040s onwards. I won't be alive to see it)
Fuelling is a separate issue, but moving to thorium would rescue the rare earth industry and allow a step change in electronics/magnetics/optics designs as currently very expensive and hard to obtain materials become cheap & common (like the way Aluminium changed after electrolysis smelting was developed)
The fundamental issue is the structure of the electricity market. To start with, it isn't a natural market. It's an artificial construct which is intended to emulate a real market but is in reality a very convoluted system of regulation intended to produce a pre-determined outcome.
What it does is optimize for short term profits rather than long term low cost or for stability or reliability. Since there is no long term security, investors optimize for short term profits. This also means that long term investments have to pay higher interest costs because they have no guaranteed market. This is the real reason why the UK has built gas turbines / wind farms (the two are joined at the hip) rather than nuclear power plants. The real money is in the subsidies and offsets, producing reliable supplies of electricity at the lowest possible cost is a mug's game.
What the UK needs to do is to admit that the "deregulated" electricity market experiment has failed and move on from it.
@thames
"The real money is in the subsidies and offsets, producing reliable supplies of electricity at the lowest possible cost is a mug's game."..."What the UK needs to do is to admit that the "deregulated" electricity market experiment has failed and move on from it."
Its not really an energy market when the government is putting fossil fuels out of business including gas when as you say gas is required for the unreliable sources. Government blocked nuclear and is subsidising the unreliables causing loads of wind to be built where it will be paid to turn off. The artificial market (agreed) was changed and skewed to make unreliable tech that didnt work the most lucrative type to build.
Unfortunately the government has decided that energy generation is not important and as a result this is where we are.
"It seems likely that datacenters are going to be mostly powered by electricity generated by a mix of gas turbines plus renewable energy from wind and solar"
That is the right way around, powered by gas, and the unreliables. it is amazing how the solution was known 20+ years ago (nuclear) especially if you believe in the MMCC Co2 theory and yet only recently the green nuts seem to be catching up.
@BadRobotics
"Investors are rightly questioning the logic of building a gas tubine, with a lifespan around 25 years, powering a datacentre that may or may not be around in 5 years."
I cant imagine it is limited to just gas turbines. We all know the AI thing is a bubble we just dont know when it will burst. But if they cant have a reliable source of power they certainly wont be relying on the unreliables exclusively. Even if such reality invites downvotes
AI may be a bubble but the reality of bubbles based around actual products is that usually 2-5 years after the bubble collapses, things are back to the same size as the bubble peak
Also, AI may be a bubble but we're going to need the power turbines anyway. With increasing electrification of transportation and the phasing out of gas/oil heating systems, electrical demand is likely to at least double (my prediction is that full decarbonisation will need a 6-8fold increase in net power generation over what existed in 2001-10) - that's a longer term demand issue though
Given Britain's housing fleet, relying on insulation is not going to achieve much more. The low hanging fruit is alrerady plucked and further energy savings will cost more for lessor results
How much niclear waste do you think is produced and how long do you think it's dangerous for?
The UK's entire nuclear waste pile will fit in a space smaller than a football field and is safe to handle (with gloves on) in about 120 years, being less radioactive than the original fuel (which is a weak alpha emitter) in ~450 years
Molten salt nuclear power can reduce THAT volume by 99% whilst being essentially inert in 450 years
If the Romans had nuclear reactors the British Museum could have the cores on display behind glass today - and the primary reason for "behind glass" is the chemical toxicity of uranium/plutonium.
Follow the money,
Although in the case of AI it is a fear of being left behind. I would be very surprised if the energy costs the datacentre operators pay covers the full cost yet they will want first access to x number of MW.
Just like Amazon buying the entire output of a wind farm so they could claim they are using renewables.
Firstly it is not 100% reliable and all that happens is the rest of us continue to use electricity generated from gas.
I am always puzzled by these energy companies that claim all their electricity is renewable. The simple answer it is not and I also suspect that the people being sold it and the resulting usage exceeds what is generated.
People keep demanding that electricity pricing be decoupled from gas costs. I am fine with that as long as those demanding it are happy to have reduced load or power cuts when demand outstrips renewable generation and short term backfill from batteries. The trouble is that they are not. They want a 99.9999 reliable source at bargain basement costs. Not that renewables are that cheap in the UK due to all sorts of other insanity like paying wind farms to no generate.
UK interest in reviving nuclear power predates the AI bubble. It's based on the goal of the total electrification of society in order to meet environmental goals.
What happened is that reality sunk in and people realized that there is no path to "net zero" which involves wind turbines. Wind turbines are joined at the hip to fossil fuels and will be forever. Solar panels are the same.
If you look at which countries in Europe have low carbon emissions in their electricity sector, it isn't the ones which depend on wind/gas. It's the ones which depend on hydro-electric and/or nuclear.
So if the UK genuinely desires to save the environment, then it either needs to find a continental scale high mountainous plateau somewhere in say Norfolk and built hydro-electric plants there, or else build enough nuclear power plants to power Britain. The latter sounds like the more realistic plan.
"The UK will need extra energy to power an uptick in datacenter construction as part of its AI Opportunities Action Plan detailed at the start of this year."
That makes the rather risky assumption that the AI bubble won't have burst by the time any of these new plants are actually ready to start generating electricity...
So what does "AI bubble will burst" mean? Do you think it means AI will go away?
All it means is a lot of people will lose a lot of the perceived value of their investments. The actual AI tech will continue to develop and spread in the hands of the familiar.
Like the dotcom bubble burst and all that happened is that big chunks of internet were consolidated into a few giant players and the "dotcom" world is now even more pervasive than it was then. So energy supply will still be required, just like hyperscale datacenters are still required.
First you have to slash the regulations around planning permission and environmental impact, then the ones around safety regulations.
Windscale opened inside 4 years of the decision to build it. Chicago Pile-1 took less than 2 years to build.
Not sure about the downvote, have checked and you are correct and frankly that is amazing. Having just read the Times article on Hinkley points £700M Fish Disco to save less than 1 salmon per year, you have to wonder how anything will ever be built again.
Environmental rules are like tax laws. Whenever someone finds a loophole it gets fixed by adding yet more rules, rather than by fixing the existing ones. The result is warts on top of warts that are ever more onerous to cope with, and inevitably add yet more loopholes. The only real solution is to start again, but no government will ever do that.
Sounds great except..who'll be building them and what will we be letting them charge for the energy?
It's just that sucesssive UK government seem very keen to let pretty much anyone build power stations here (French, Chinese, hey might as well let the Russians have a go) even when we're discussing them as mortal enemies in the next breath. Then we allow those builders to charge basically whatever they want for the power they generate. Guaranteed 200% above the average price, no problem. UK consumers will compensate you if they don't buy enough, sure help yourself. UK tax payers to pay for the cleanup no matter how much money you've made or how much of a mess you've left behind, absolutely.
UK Gov negotiates like someone with Stockholm syndrome.
Our incompetent politicians chose short term electability over long term energy security and so let our home grown nuclear industry die out, and now they and everyone else are bemoaning what was done. My Dad worked for TNPG on Hinkley B back in the 70s and saw what was coming so transferred to a US nuclear power company, but even that faded away in the 90s. Too many NIMBYs wanted cheap reliable electricity but they all wanted it generated in someone else's backyard.
Then we had Chernobyl where a poorly built nuke in country that lacked reasonable safety precautions blew it's top (non-nuclear but effectively a big dirty bomb) and now everyone wanted nuclear power banned everywhere, regardless of whether it was well built, safe, and reliable or not. In Japan money grubbing cheapskates decided NOT to listen to the engineers who told them to put the emergency generators on the roof of the containment vessels, and that (helped by the Tsunami) gave us the Fukushima meltdown.
Because of the Calder Hall/Windscale fire in the UK we were one of the first to regulate the nuclear industry and rightly so, but that regulation has now turned into a monster that needs cutting down to size. So, here we are with nearly no nuclear power industry of our own and having to buy ours from the Americans, Chinese and French. We currently have 4 sites generating power, Hartlepool, Heysham (1 & 2), Torness and Sizewell B, but 3 of those will probably end power generation in the next 5 years, which leaves us with Sizewell B. France on the other hand has about the same population level as the UK, but has 56 (soon 57) nuclear reactors generating power versus the UKs 10.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
