As AI Booms, Land Near Nuclear Power Plants Becomes Hot Real Estate 77
Tobias Mann reports via The Register: The land surrounding a nuclear power plant might not sound like prime real estate, but as more bit barns seek to trim costs, it's poised to become a rather hot commodity. All datacenters are energy-hungry but with more watt-greedy AI workloads on the horizon, nuclear power has fresh appeal, especially for hyperscalers. Such a shift in power also does wonders for greenwashing narratives around net-zero operations. While not technically renewable, nuclear power does have the benefit of being carbon-free, not to mention historically reliable -- with a few notable exceptions of course. All of these are purported benefits cited by startup NE Edge, which has been fighting for more than a year to be able to build a pair of AI datacenters adjacent to a 2GW Millstone nuclear power plant in Waterford, Connecticut.
According to the Hartford Courant, NE Energy has secured $1.6 billion to construct the switching station and bit barns, which will span 1.2 million square feet in total. NE Energy will reportedly spend an equivalent sum on between 25,000 and 35,000 servers. Considering the price of GPU systems from Nvidia, AMD, and Intel, we suspect that those figures probably refer to the number of GPUs. We've asked NE Edge for more information. NE Energy has faced local challenges getting the project approved because residents are concerned the project would end up increasing the cost of electricity. The facilities will reportedly consume as much as 13 percent of the plant's output. The project's president Thomas Quinn attempted to quell concerns, arguing that by connecting directly to the plants, NE Energy will be able to negotiate prices that make building such a power hungry facility viable in Connecticut. NE Energy has also committed to paying a 12.08 percent premium to the town on top of what it pays Dominion for power, along with other payments said to total more than $1 billion over the next 30 years. But after initially denying the sale of land to NE Edge back in January over a lack of information regarding the datacenter project, it's reported that the town council has yet to tell the company what information it is after.
According to the Hartford Courant, NE Energy has secured $1.6 billion to construct the switching station and bit barns, which will span 1.2 million square feet in total. NE Energy will reportedly spend an equivalent sum on between 25,000 and 35,000 servers. Considering the price of GPU systems from Nvidia, AMD, and Intel, we suspect that those figures probably refer to the number of GPUs. We've asked NE Edge for more information. NE Energy has faced local challenges getting the project approved because residents are concerned the project would end up increasing the cost of electricity. The facilities will reportedly consume as much as 13 percent of the plant's output. The project's president Thomas Quinn attempted to quell concerns, arguing that by connecting directly to the plants, NE Energy will be able to negotiate prices that make building such a power hungry facility viable in Connecticut. NE Energy has also committed to paying a 12.08 percent premium to the town on top of what it pays Dominion for power, along with other payments said to total more than $1 billion over the next 30 years. But after initially denying the sale of land to NE Edge back in January over a lack of information regarding the datacenter project, it's reported that the town council has yet to tell the company what information it is after.
Breeder reactors (Score:2)
Re:Breeder reactors (Score:5, Informative)
Re: (Score:3)
Re:Breeder reactors (Score:5, Informative)
You're still not incorrect, both BWR and PWR reactors are immersed in water at high pressure.
Main difference is that in a BWR - Boiling Water Reactor, the water is allowed to boil in the reactor to produce steam.
In a PWR - Pressurized Water Reactor, it isn't allowed to boil, and a coolant circuit is used to generate the steam.
PWR is a bit more expensive, but doesn't run reactor water, which can be slightly radioactive, through the turbines like a BWR, and a non-radioactive turbine is easier to handle.
Re: (Score:1)
The point of a breeder reactor is to make material for nuclear weapons. The radioactive waste they generate isn't worth the power you get from them.
If we limit our nuclear power to fuel mined and refined at current costs, used in non-breeder reactors, uranium will last less than a century, and you can only double that if you throw in thorium.
If you are willing to have power three times as expensive as current nuclear power generation rates, we could go for tens of thousands of years getting uranium extract
Stuff will last a lot longer than that (Score:3)
Yes, with current technology and current known reserves, we have "less than a century". Keep in mind by the same standard we have less than 5 years of oil left.
New technology can lower the cost. Prospecting will find more. A relatively slight increase in price will open up more deposits to profitable mining.
If you throw in thorium, it goes up by at least one order of magnitude. If you're willing to extract uranium (and thorium) from seawater, yes, it'll triple the cost of uranium (using current technolo
Re: (Score:3)
Re: (Score:3, Informative)
This BS again.
Leaving out Africa and Kazakhstan and Canada and others where there are massive deposits, just look at one where data is readily available.
Australia currently supplies about 10% of the world market. It has about 28% of the world's "available" resources. That word resources skews every discussion.
There are a few definitions. It doesn't describe how much uranium is actually available, but rather what can be mined legally and economically.
Australia has anti-uranium mining legislation at Federa
Re: (Score:2)
The point of a breeder reactor is to make material for nuclear weapons.
No. The fact that something has a possible use case such that does not mean that it's "the point of it".
The radioactive waste they generate isn't worth the power you get from them.
You definition of "worth" is a social construct. Just because they produce usable plutonium doesn't mean that it isn't worth it, it means we need to better control the waste. If you think you can rid the world of nuclear weapons by not producing plutonium you're really naive.
Re: Breeder reactors (Score:4, Interesting)
Basically yes. The original nitpick for what's green was saying no waste product but panels and turbines need to be replaced, so the line was redrawn to be no radioactive waste. Radioactive waste is annoying to deal with but I think some advanced designs highly reduce it or reuse waste in a 2nd stage reactor.
What are human timescales though. If you mean the period of time we have been a species, then breeders could likely see us for a fair longer but we have to consider accelerating change and how power demand grows at an accelerating rate. If we are space faring and seek renewable energy, well railguns launching things to space or the industrial scale to build a carbon nano fiber space elevator, then you are talking orders of magnitude more energy than we generate now. This is why the classic idea of a dyson sphere arrises.
Will humanity achieve those time scales, I wouldn't bet on it but it does kind of point to harnessing the suns energy as the ultimate objective. One might even argue wind energy is just an alternative means to harness wasted energy from the sun, as it becomes heat, and shapes a planets climate. It's a very interesting theoretical space.
Re: (Score:2)
...but we have to consider accelerating change and how power demand grows at an accelerating rate.
No, power demand does not grow at an accelerating rate. In fact, per capita, it has been reducing, even in the USA: https://www.eia.gov/totalenerg... [eia.gov].
Now, if we do decide to become space-faring, we might use a bit more, but then again, we won't be visiting Mars like we go on holiday to Spain or something.
Re: Breeder reactors (Score:2)
Can someone please explain why they have to be built closer to datacenters? Typical transmission line loss is something like 2%. If you have to pay say a 30% premium to have a closer location, why would they do it?
Re: Breeder reactors (Score:2)
I'm also guessing that high land prices are relative. For most people, being right beside a nuclear reactors probably a negative. So I'm guessing the land is cheaper than literally the middle of nowhere but not terribly expensive compared to other advantageous locations.
At least it's not cryptoscam companies buying coal (Score:3)
Like for real, you can tell the priorities when companies decide that building next to power plants is a good idea.
While nuclear isn't the worst option, the power plant has an expiry date. Building vertical data centers near the power plant solve specific problem by moving internet infrastructure closer to a reliable source of energy. So this is fine.
However if we really gave a care, we should be moving the data centers into the power plant. Like, literately double-dip on the energy generation AND the cooling. Hydro and Nuclear plants have this option, No other plants do.
Re: (Score:2)
They probably can't share cooling with a nuclear plant for safety reasons. If there is an accident or serious fault, that cooling is all that stands between a safe outcome and meltdown.
Re: (Score:1)
There are no designs in use today that cannot be passively cooled when necessary. People thought about this in the 1950s, no need to overthink it, nuclear power is safer, cleaner and has less deaths than either solar and wind. And it will last for thousands of years.
Re: (Score:2)
So how come Fukushima had a triple meltdown if it could be passively cooled? That design is still in use today.
Re: (Score:1)
Fukushima was a series of poorly trained human responses (lots of assumptions) overreacting to a contained problem. They assumed certain valves were open when they were shut and vice versa, they assumed coolants were no longer present when they were.
The reactors melted down but the problem was contained and remains contained to date. I didn't say certain things won't break, obviously you can screw up things a lot, but they won't explode like the nuclear military experiment did at Chernobyl, which required a
Re: (Score:2)
They couldn't get to the valves to check their state, and the monitoring system was down.
Anyway, there is no way I'd trust some private company with profits as its main concern to safely run a nuclear plant. You cite poor training as an issue, and in my experience it's pretty much the norm for most industry.
Re: (Score:2)
Like, literately double-dip on the energy generation AND the cooling.
You're conceptually right but you're handwaving away a lot of practical engineering challenges. It is insanely complex to design cooling systems which can at large scale handle drastically different incoming and return water temperature requirements. The requirements for a datacentre are significantly lower than those of a nuclear power station. One cooling system is virtually impossible to operate both at once in an efficient manner, and it is often of benefit in terms of cost and energy consumption to kee
Re: (Score:3, Informative)
Nothing is carbon free. In fact the mining for wind and solar release a lot of emissions as well. 12 g CO2 per kWh for nuclear. 11-12 g CO2 per kWh for wind. 41 g CO2 per kWh for solar.
Nuclear is low in carbon.
If you oppose nuclear due to its 12 g CO2 per kWh you have to also oppose wind and solar for the same reason.
Re:not really (Score:5, Informative)
No, they don't really oppose nuclear because of the CO2 released while building the plant they oppose it because it's NUCLEAR. The CO2 is just an excuse to cover their unthinking hate of any and everything nuclear.
Re: (Score:2)
Re: (Score:2)
The only way you can get a mean of 12 is by pretending that it is only maintained and fueled by EVs, at which point a wind plant goes to near zero.
Re: (Score:1)
12g CO2eq/kWh is likely an overestimate. According to the UNECE Life Cycle Assessment [unece.org], nuclear carbon intensity was found to be 5.1–6.4g CO2eq/kWh.
No doubt the numbers you cite are inflated by inclusion of much less efficient (and obsolete) gaseous diffusion enrichment plants, and probably by other means. Some of the "work" done by anti-nukes produce some rather outrageously nonsensical numbers.
Moreover, the carbon intensity of conventional nuclear is probably an order of magnitude higher than it woul
Re: not really (Score:1)
Does that include if all the employees drive coal rolling trucks to work every day 100 miles and back?
Re: (Score:2)
Re: (Score:2)
Lying? No. It's all about your estimates of mining and transportation.
Re: (Score:3)
Re: (Score:2)
It's a fact. It's the median for *different studies*, which use *different methodologies*.
Some of those, like the 288, are likely absurd. Some of them, like the 1.4, are also absurd.
The number you cite comes from a 2014 Yale methodology, that has been copy-pastad in the UN ever since.
It's less credible than an analysis of all studies.
Nobody is lying here, there are facts being discussed, facts about studies, which are all of dubious quality. You're cherry picking, and I'm not. It's that
Re: (Score:2)
Why is the 1.4 absurd? It is much closer to the IPCC 2014 12 g median value than your BS 288.
When clean energy powers more elements of a power source the g per CO2 will drop. Some countries enrich their uranium with nuclear energy which will lower the lifetime numbers to less than 10.
That's true for solar as well which has dropped from 41 to 26. Solar has become cleaner as well.
2014 Yale methodology
Why is the Yale methodology wrong?
There are more recent analyses that has nuclear lower than 12 g CO2. Such as the UNECE a
Re: (Score:2)
When clean energy powers more elements of a power source the g per CO2 will drop.
Correct. The same applies for wind and solar. Their entire co2 cost is that of fossil sources of energy in mining.
Some countries enrich their uranium with nuclear energy which will lower the lifetime numbers to less than 10.
Enrichment isn't a large cost. It's the mining and transport that is.
That's true for solar as well which has dropped from 41 to 26. Solar has become cleaner as well.
Of course. Which is why there's a wide range of estimates for nuclear CO2 impact (as well as solar CO2 impact)
There are more recent analyses that has nuclear lower than 12 g CO2. Such as the UNECE analysis.
UNECE, IPCC, are both replications of the original Yale LCA.M
288 g CO2 per kWh is not a fact. If someone wants an honest analysis/debate I'm all for that. You don't though. That's why you brought up that absurd number. Just bringing the number up is a lie. You already admitted you knew it wasn't true. You did it just to continue to muddy the waters about nuclear energy.
Quit purposefully misreading.
I specifically said the fact being discussed are the facts about the studies, not that the studies are factual. There dozens, al
Re: (Score:2)
I still haven't heard why the Yale/IPCC study is wrong.
Sovacool is an antinuclear hack. His entire career is antinuclear. Bringing up anything from him is a red flag.
Just mentioning 288 is still dishonest.
And the mining requirements for solar and wind are significantly greater than nuclear.
Re: (Score:2)
I still haven't heard why the Yale/IPCC study is wrong.
I never said it was wrong. Stop putting words in my mouth. I said it was one of many numbers.
Just as your wind number is but one of many, with a low end of 1g/kWh.
Sovacool is an antinuclear hack. His entire career is antinuclear. Bringing up anything from him is a red flag.
You don't get to write off anything you dislike as "antinuclear". That's not debating.
Just mentioning 288 is still dishonest.
What arbitrary line in the sand do you draw as honest?
And the mining requirements for solar and wind are significantly greater than nuclear.
In some ways, yes.
The mining cost of solar and wind will go down with time, nuclear will only, and can only, ever go up, comparatively.
Re: (Score:2)
You don't get to write off anything you dislike as "antinuclear". That's not debating.
I absolutely get to write off antinuclear scumbags such as Sovacool. Every time he releases a paper it gets retracted.
What arbitrary line in the sand do you draw as honest?
288 is not arbitrary. It is intentionally dishonest. How about providing a real number based on empirical data.
nuclear will only, and can only, ever go up, comparatively.
k
Re: (Score:2)
I absolutely get to write off antinuclear scumbags such as Sovacool. Every time he releases a paper it gets retracted.
No. That is an ad hominem fallacy.
That is also untrue. It is true that they have voluntarily retracted a paper after systematic errors were found in it, but that's not what you said.
You seem to have a very tenuous grip on truth when you're flinging your rhetoric around.
288 is not arbitrary. It is intentionally dishonest. How about providing a real number based on empirical data.
Of course it's arbitrary. By definition.
The number comes from Storm/Smith 2007, and is an estimate for a plant using 0.13% grade ore.
It is a real number based on empirical data. Are the deductions made incorrect? Entirely possible. Now ex
Re: (Score:2)
He retracted a paper after it was discredited. He is also an antinuclear scumbag. On Anti-Nuclear Bullshit [thebreakthrough.org]
It's not 0.13 % ore, it's 0.013 % grade ore. Very low ore grade is not used. It also assumes that enrichment was done with fossil fuels and not clean energy sources. France uses nuclear for enrichment. The number is such an outlier and is commonly used by antinuclear activists. Just that makes is suspect.
We have enough nuclear material to power to world for 4 billion years.
I'm shilling for
Re: (Score:2)
It's not 0.13 % ore, it's 0.013 % grade ore. Very low ore grade is not used. It also assumes that enrichment was done with fossil fuels and not clean energy sources. France uses nuclear for enrichment. The number is such an outlier and is commonly used by antinuclear activists. Just that makes is suspect.
You'll forgive the missed zero. Of course it's an outlier. I literally said many times it was an outlier, and Sovacool himself said so.
You claimed it was fake. That claim was incorrect.
You continue to use ad hominem argumentation, and now you're changing your tune because you've bothered to actually look up the facts.
So now you're attacking the fact that it's an outlier, which was never in dispute. You don't even know what you're arguing, you're just foaming at the fucking mouth.
You are exactly what you
Re: (Score:2)
Re: (Score:2)
I claimed it was a lie because it was.
You're misusing that word. It is not defined as you use it. That's a standard tactic of political spinsters.
You knew it it was bullshit but you mentioned anyway just to muddy the waters.
No. Leaving out numbers is what makes you you, and me me. It's what separates you, the shill, from me, the person looking to understand.
You cherry pick to fit your preconceived notion. I want the answers.
Re: (Score:2)
It's only 12g in the very best case. It can be over 100g if they don't have low carbon sources of fuel and long term storage, and depending on how they cool the plant.
Of course renewables vary depending on how they are manufactured and disposed of as well, but not as much.
The real issue is cost. They are not building nuclear plants, they are leeching off the ones that have already been built at someone else's expense. Now those plants are finding it hard to compete due to much cheaper electricity being avai
Re: (Score:3)
Re: (Score:2)
Perhaps you can see why people don't take you seriously when that's your reaction. It's not even addressing the claim being made.
Re: (Score:2)
Re: (Score:2)
> If you want to waste energy productively, use grow lights for pot farms.
If you want to waste energy productively, use regular (full spectrum) lights for pot farms.
Re: (Score:2)
I suppose you could produce a lot of hot showers while coiling it? Or at least open a laundry?
Flag on the Play (Score:2)
"energy-hungry"
"watt-greedy"
Illegal use of hyphens. Ten yard penalty from the spot of the foul. Second down.
Re:Flag on the Play (Score:5, Funny)
After a review of the play, watt-greedy was found to be used correctly as a compound adjective for the noun "AI". Energy-hungry was found without a noun to modify and was found to be superfluous.
Additionally, "Ten yard penalty" was found to need a hyphen to become "Ten-yard penalty" so the penalties cancel out. It remains second down.
And there's none hotter ... (Score:2)
Line Loss (Score:2)
Wow, a real estate boom that actually makes sense. Whodathunkit.
Umm.. (Score:1)
this doesn't sound like it will end well.
You will eat cesium. (Score:2)
Does not matter.
The (fittest) AI will survive.
And you will eat cesium.
I should work... if nuclear was cheap (Score:3, Interesting)
Anybody how has search the data knows the enormous difference between industry numbers about projected costs and numbers of recent projects at the end.
Overcosts usually turn nuclear projects from the good idea it seemed at start into financial nightmares at the end.
The main reason of raising prices usually is the speed of construction. As nuclear reactors are big, complex machines, that needs to be monitored closely about multiple concepts, any problem will turn into significant delays. And because they are so big projects, it's required to borrow a lot of money, interests on the delay can multiply the cost by a lot.
That's the main reason why last years were a lot of attention about SMR (small nuclear reactors). Instead of a big reactor that needs to be build on one go, build multiple and small, so you can turn on step by step, more like renewables, so the money return sooner and costs should be more contained, or at least, closer to projected costs.
Still, after checking about this new promise, it's still on hold. It seems that scale nuclear to smaller reactors while you pretend to maintain linear costs is not as easy as it seems. It doesn't matter if SMR can be connected faster if cost per unit power multiplies by 2 or 3. Nuclear costs then are still way beyond renewables.
As energy storage starts to become cheaper over time, nuclear industry is loosing their market opportunity. If renewable+storage turns cheaper than nuclear, nuclear just becomes unnecessary.
Re: (Score:1)
As energy storage starts to become cheaper over time, nuclear industry is loosing their market opportunity. If renewable+storage turns cheaper than nuclear, nuclear just becomes unnecessary.
One is placing bets on some magic pixie dust battery tech that's suuuurely juuuust around the corner, that'll make them cheap and scalable, other is betting on technology that's been proven for decades. hmmmmm, which I would pick, I wonder... But quick, I'm sure all those people who ran cost analyses on this investment just didn't think to build a PV farm and battery depot instead, tell them, you'll save them millions.
So GAI will eventually be promoting, ... (Score:2)
... building and expanding nuclear fission plants? And no reason to worry about radiation? Great. I see absolutely no apocalyptic science fiction scenario coming out of this. One thing *is* certain though: That real-estate sure will hot, in more ways that one.
Save energy (Score:5, Insightful)
Save energy to save the planet. Install LED bulbs, turn off the lights, don’t use AC as much.
Oh and also, we are building another datacenter filled with GPUs
Re: Save energy (Score:2)
We have to run power hungry large language models so that teens will know what to text their boyfriends.
Re: (Score:3)
Save energy to save the planet. Install LED bulbs, turn off the lights, don’t use AC as much. Oh and also, we are building another datacenter filled with GPUs
Well, yes. It's always YOU who needs to de-industrialize, you know. Somehow, it's never me ...
(Now let me go post some enviro-stuff on a worldwide computer network.)
Re: (Score:3)
Save energy to save the planet. Install LED bulbs, turn off the lights, don’t use AC as much.
Oh and also, we are building another datacenter filled with GPUs
This is slashdot, you can start by telling people they aren't teenagers anymore and that it's not cool to masturbate over their OS uptime numbers.
Now I'm leaving my desk for just an hour. The computer is going to be off during that time, because let's face it, I'm not running Windows XP on spinnin rust HDDs, and it boots up in a matter of about 3-4 seconds anyway.
Personal Responsibility is a scam (Score:2)
Carbon-free? Not really (Score:1)
Re: (Score:1)
Carbon-free usually means "no direct emissions".
Or said in another way, if that source of energy were the only one which feeds our consumption, it would be carbon-free.
It's the same in renewables.
Of course, during the transition, we still uses a lot of fossil fuels and generate a lot of indirect emissions. But it decreases as our energy mix uses more and more carbon-free sources of energy (pretty much renewables and nuclear).