Amazon To Fund 12 Small Nuclear Reactors At Former Indiana Coal Plant (techcrunch.com) 86
An anonymous reader quotes a report from TechCrunch: Amazon is funding development of a proposed 960-megawatt nuclear plant in Indiana and would be the first customer to buy its power, the companies announced Monday. Utility company American Electric Power (AEP) would operate the plant. X-energy, which designs small modular nuclear reactors and went public in April, would supply the reactors. The new power plant would consist of 12 Xe-100 small modular reactors and be located on the site of a demolished coal power plant just south of Terre Haute. The deal is the second commercial agreement between Amazon and X-energy, and it's three times larger than the first, which was announced two years ago.
Amazon has invested heavily in X-energy, leading a $700 million Series C-1 round in October 2024. The two companies have said they ultimately plan to build 5 gigawatts of new nuclear generating capacity in the U.S. by 2039. Unlike many newly announced power plants, these reactors would be connected to the public electric grid, though Amazon says the power would also serve its data centers.
The companies' aggressive timeline still depends on X-energy successfully delivering its first power plant at a Dow Chemical plant southeast of Houston, Texas. The site, called Seadrift, has recently been at the center of a lawsuit filed by the state of Texas, which accuses the plant of "habitual non-compliance" with water and solid-waste regulations and permits. X-energy has yet to break ground at Seadrift, which will host four Xe-100 reactors to provide the chemical plant with heat and electricity. The company said its construction permit is in the final phases of the U.S. Nuclear Regulatory Commission's review process.
Amazon has invested heavily in X-energy, leading a $700 million Series C-1 round in October 2024. The two companies have said they ultimately plan to build 5 gigawatts of new nuclear generating capacity in the U.S. by 2039. Unlike many newly announced power plants, these reactors would be connected to the public electric grid, though Amazon says the power would also serve its data centers.
The companies' aggressive timeline still depends on X-energy successfully delivering its first power plant at a Dow Chemical plant southeast of Houston, Texas. The site, called Seadrift, has recently been at the center of a lawsuit filed by the state of Texas, which accuses the plant of "habitual non-compliance" with water and solid-waste regulations and permits. X-energy has yet to break ground at Seadrift, which will host four Xe-100 reactors to provide the chemical plant with heat and electricity. The company said its construction permit is in the final phases of the U.S. Nuclear Regulatory Commission's review process.
Re: (Score:2)
We'll see a better TV show in a few weeks: vlad putin transferring an important intelligence asset to China.
Re: (Score:2)
Please, Putin is in no way able to ship la Presidenta to China. And besides, he'd have to pay China to take him.
Re: (Score:2)
La presidenta will ship itself on the Qatari jet on your dime. putin will only transfer the yellow folder with the dossier.
I'm honestly torn... (Score:5, Interesting)
On the one hand, the power generated may offset the CO2 emitted to build the plant, and then possibly offset even more CO2 by preventing other fossil-fuel plants from being built. Gaining experience with SMRs could be critical for the survival of our civilization.
On the other hand, I trust neither Amazon nor private utility companies. This kind of project belongs in the public sector, and for the public good, not under the control of billionaires as another scheme for extracting wealth from the people they consider to be expendable nobodies. Also, a proliferation of nuclear power has its own risks. And I'm sure we can find better uses for the power than more fscking data centres.
I guess I'd be cautiously in favour of the SMR plan if it was for the public good and not for the good of Bezos and his buds. But as it is, I hope the whole thing falls flat. We need to stop handing power - both political and electrical - to the parasite class. It's time to start shutting these fuckers down.
Re: (Score:1)
This kind of project belongs in the public sector,
Like Chernobyl?
And I'm sure we can find better uses for the power than more fscking data centres.
You are posting this on Slashdot, which is hosted in a data center.
Re:I'm honestly torn... (Score:4, Funny)
Are you sure? I thought this site ran on a dusty old 486 running FreeBSD under someone's desk?
Wait, sorry... I forgot: BSD is dead, NetCraft confirmed it.
Re: (Score:2)
The 15 of us that still use Slashdot aren't using very much resources.
Yeah, but what about the two dozen bots? Those used to be simple pattern generators but now they're expensive pattern generators, i.e. LLMs.
Re: (Score:1)
I guess I'd be cautiously in favour of the SMR plan if it was for the public good
The concept behind the SMR is making them cheaper and installing them in more locations. Neither of those things are good; the first means you'd have to cut corners and the second is obviously bad given the cost of failure.
Re: (Score:2)
True. I hadn't considered the points you raised. Thinking about them now - and knowing that we're probably going to have SMRs whether we want them or not - I'm hoping that the inevitable 'incidents' aren't too serious. I'm also hoping - for the sake of reliability and safety - that the designs evolve to the point where they're almost like kits put together from off-the-shelf parts. Complacency and carelessness will still be a problem though.
I think that, with sufficient care, SMRs can be made fairly safe ov
Re: (Score:1)
How do you even know what a SMR is? You're a cat! See the red dot? Chase the red dot! Get it!
Re: (Score:1)
An objective position based on a neutral observation.
Check the Fulfillment (Score:2)
I would feel better if these modular nuclear reactors were shipped from an Amazon warehouse and hence eligible for the over $35 dollars free shipping.
If they are sourced from an Amazon Partner, I bet they arrive with a cracked reactor core and a leaking pressure vessel?
Bezos doing atomics (Score:5, Funny)
What can go wrong?
I wonder if the control room technicians will be allowed bathroom breaks...
Re: (Score:2)
What can go wrong?
I wonder if the control room technicians will be allowed bathroom breaks...
Normal operation and monitoring is all automated. So the control room techs as such don't have any real-time or critical roles, as i understand it.
Re: (Score:3)
A shitload of corporations who are losing money hand over fist on AI and are under pressure to make it profitable are building and reopening a shitload of nuclear power plants.
I think the phrase is "what could possibly go wrong?" Followed by a laugh track and a general order of evacuation lasting
Re:Bezos doing atomics (Score:5, Informative)
This is overstated. According to X-energy’s regulatory disclosures, the Xe-100 has four main operator-controlled variables: control rod position, He circulator operation, feedwater pump operation, and turbine throttle valve position. It's more automated than older plants, but not completely automated, and older plants are also quite automated. It does use more passive safety features, though. And of course, it's not actually been built yet, so it's all theoretical at this stage.
Re:Bezos doing atomics (Score:4, Informative)
And of course, it's not actually been built yet, so it's all theoretical at this stage.
This is the most important part. Right now, X-Energy is pure vaporware. And yes, I know, all reactor designs are essentially vaporware, because their task is to heat water to turn it into vapor. But this is vapor-vapor. The time line calls for 12 small reactors to be ready by 2039, each rated 80 MW. And the previous project, as mentioned, was already set up 2 years ago, and not a single foundation has been poured yet.
This is one of those incubator projects. Amazon is investing and probably betting on a 30% chance that this will ever bear fruit. Or a 10% chance. I don't know Amazon's calculation. But this project does not mean that any of those 16 reactors (4 previously, 12 now) will ever come into existence. But if they would, Amazon had an edge on powering data centers. So they think it's worth the risk.
Re: (Score:2)
Investing is the key word here. If they need power so urgently then why not fund the reactors or energy company? More than likely this is a quick stock pump and dump.
Re:Bezos doing atomics (Score:4, Interesting)
X-energy doesn't have NRC design approval. Of course they aren't going to be pouring concrete, because they don't even know what the fuck they're building yet.
The only SMR design that does have NRC design certification is NuScale.
This is nothing but speculation and press releases at this point.
Re: (Score:2)
And if you crave really primitive back to the (radioactive) land homesteading there's always Chernobyl adjacent property. As long as you avoid chunks of radioactive heavy equipment and don't sneak in to see the Elephant's Foot area it will be fine. The chances of Russia destroying the new cont
Re: (Score:2)
It's even more theoretical than that, because there's only one SMR design that actually has NRC regulatory approval, and X-energy isn't it.
They can't even get started building without that. This is all vapor.
Re: (Score:2)
Wow, that I didn’t know! I’m not surprised though. The thing that is galling about it, is that the effort and money being put into all this theoretical stuff could instead be used to build out actual deployments of renewables, which ciould be up and running much faster. When you run the numbers on LCOE and include O&M and decomm, it’s clear that would deliver more energy for the same amount of cash.
Re: (Score:2)
This is overstated. According to X-energy’s regulatory disclosures, the Xe-100 has four main operator-controlled variables: control rod position, He circulator operation, feedwater pump operation, and turbine throttle valve position. It's more automated than older plants, but not completely automated, and older plants are also quite automated. It does use more passive safety features, though. And of course, it's not actually been built yet, so it's all theoretical at this stage.
hmmm you're right. I was thinking of "micro reactors", apparently, which are currently even more speculative.
Re: (Score:2)
So the control room techs as such don't have any real-time or critical roles
Continuous control is automated, but a control room operator very much does have a real-time critical role in plant operation as far as responding to any deviation from normal operating condition.
A start-up or shutdown of a reactor is a highly dangerous operation compared to normal running. Control room operators play a critical trouble shooting role between things running perfectly normally, and a safety system shutting the thing down, and that can occur at any moment which is why such control rooms need t
Re: (Score:1)
What can go wrong?
I wonder if the control room technicians will be allowed bathroom breaks...
Normal operation and monitoring is all automated. So the control room techs as such don't have any real-time or critical roles, as i understand it.
This answer tends to trigger a GOTO 10 response; Sooo, What can go wrong?
Re: (Score:2)
Nothing!
https://en.wikipedia.org/wiki/... [wikipedia.org]
So A Full-Self-Driving SMR (Score:2)
What can go wrong?
I wonder if the control room technicians will be allowed bathroom breaks...
Normal operation and monitoring is all automated. So the control room techs as such don't have any real-time or critical roles, as i understand it.
The "Full-Self-Driving SMR" will be announced today and may partially work reliably in 12 years time...
Re:Bezos doing atomics (Score:5, Funny)
****1/2 "Customers say that this is a great small modular nuclear reactor. They like the compact size and the fact that it's not too much power for a data center or industrial site. They especially appreciate the ease of setup and that the ultimate waste disposal question has been deferred yet again so that they don't have to worry about it. However, some customers report problems with cost overruns, vaporware, and availability of suitable fuel. Some customers note a larger terrorist attack surface per megawatt relative to full-sized reactors and question the security arrangements. There are also reports of real-world reliability problems with the novel design, with one customer reporting a catastrophic meltdown in their urban installation."
Re: (Score:2)
That was very well done, kudos!
Re: (Score:2)
Re: (Score:2)
Here's a special offer from Temu: Purchase within the next 10 minutes and you'll get it for FREE. Just pay shipping.
Re: Bezos doing atomics (Score:2)
You can get it at wholesale from AliExpress, but make sure to pay extra for shipping or they won't pack it properly and the corners will get smashed.
Always some complainer (Score:3)
who gives it one star in their review on account for the catastrophic meltdown at an urban site.
Re: (Score:2)
What can go wrong?
I wonder if the control room technicians will be allowed bathroom breaks...
On the bright side, it's Bezos. It'll take him thirty years to go from concept to spin-up. If it were Musk, I'd be worried. That bitch's move fast and break things methodology is *NOT* what we need anywhere near nuclear power.
Not that I'm cheerleading Bezos by any means. Just saying, chances he never actually crosses the line due to caution are higher than they would be with most of the other techbro guys.
What if... (Score:2)
Re: (Score:2)
A single plant is more efficient, but adds a decade (it feels) to the timeline.
Re: (Score:2)
Except that large plants actually exist and SMRs don't...
Re: (Score:2)
The problem with large plants is that they custom made and they take so long to build and last so long that by the time you need a new one everyone who knows anything about it is dead.
With a SMR you are building them one after another and they are all the same so the knowledge will stay online. If you ever actually build anything
Re:What if... (Score:5, Informative)
That's the theory, but it's not the practice. They are taking just as long as large reactors. In some senses longer, because none has actually been put into commercial operation, whereas new large reactors have come online. And they turn out to have very little actual modularity and scalability, because there's more than 150 designs out there, so no-one is winning enough of the market for repeatable learnings, and anyway, you'll never get to volume manufacturing for the parts involved, because the quantities are just too small, and everything still requires literal nuclear-grade engineering.
Take a motor-operated isolator valve that you'd find in a cooling system. You find plenty of these outside SMR applications, often 4 to 6 inch sizes, and they might cost something like 3 to 10k USD in a typical demanding industrial application. But they will cost more like 30 to 100k USD in an SMR, because they need to be made from nuclear-grade materials with documented material traceability, need seismic qualification, need nuclear-specific quality assurance, inspection and documentation, and need highly secure supply chains (no-one wants a counterfeit component showing up in a nuclear safety system). And they won't get much cheaper, because of strict configuration control, which limits the extent of any changes to components, materials and manufacturing processes.
It's also important to recognise that the economies of learning that apply afe fundamentally different and more limited for construction projects, which is what every SMR is, than for component manufacturing. That reflects the facts that (1) different places have different ground conditions, seismic risk, regulatory requirements, humidity, water access, etc, etc, and (2) huge amounts of stuff still have to be constructed on site by humans (even if assisted by machines), compared with a factory line that can be automated. This is still a problem even for solar, where the panels themselves can be cheap compared to the install costs, which is why there are various attempts being made to turn more of the deployment into a manufactured process rather than a construction process. Eg https://5b.co/en/5b-maverick [5b.co]
Re: (Score:2)
A single plant is more efficient, but adds a decade (it feels) to the timeline.
Compared to signing a product with a vapourware company who has never delivered a nuclear project? I think you're making assumptions.
Re: (Score:2)
Isn't this already not going to happen until 2039, and that's BEFORE the usual bureaucratic delays that you're counting in that "adds a decade" thing inevitably happen? What makes you think this'll actually be faster in practice? What regulators are going to allow an untried technology that's approximately equal to the tried tech faster than the tried tech?
Re: (Score:2)
It's a speculation move.
If they were serious about actually building SMRs to generate electricity, they would not be throwing money at a company that doesn't have NRC design certification, when there is another company out there that actually does have that certification and would love to start building (NuScale).
This is a PR campaign, not an electricity generation campaign.
Re: (Score:2)
Well, if they had a proven SMR design with, say, an absolute minimum of 10 year uptime experience, maybe. But they do not and anything they do will be slower than a big nuke and it can and probably will fail completely. If they keep at it, they may have something in 30-50 years or so. And that something may even be cheaper in 100 years or so.
The whole thing is people investing lots of money into tech they do not even begin to understand.
Re: (Score:2)
-and I'm just spitballing here, what if, possibly, maybe, we considered, instead of building 12 small reactors, all next to each other on the same site, we just built one big reactor instead? Has anyone ever tried that, did anyone ever do the engineering calculations to see if that was more efficient?
but but but Three Mile Island! etc.
So yeah, the industry needed a hook to be different than legacy nuclear and this is the one that seems to have less public pushback so... apparently it's what's being made. It's more expensive per kW and kWh, but at least its getting done vs not getting done.
some AI is telling me i'm being too political and while it's true that LCOE ( levelized cost of energy) will always be higher for the small reactors... the up-front cost for the big ones is SO high (tens of billions)
Re: (Score:2)
But once again: no SMRs are in commercial operation in the US. Large reactors have been built. I'm not banging the drum for large reactors, I'm pointing out that every single SMR remains vapourware at this stage. There's a good episode on this from Matt Ferrell ("Undecided").
Re: (Score:1)
-did anyone ever do the engineering calculations to see if that was more efficient?
The arguments in favor are simpler, safer, scalable, and easier to build/deploy.
https://small-modular-reactors... [small-modu...actors.org]
Re:What if... (Score:4, Informative)
On the other hand, they have more surface which can catch stray neutrons for the same amount of power, because they are smaller, and you need more of them. That means more material gets contaminated and has to regularly get replaced because it degrades over time. If they operate at the same pressure, you need the same thickness of the walls, but with smaller diameter, you need more material per volume. They need the same amount of shielding per nuclear source to get the same protection level, but you need more of them, because you have more but smaller nuclear sources.
I've been to something which could be called an SMR. It was a scaled down nuclear reactor built in the 1970ies for studies and reactor experiments. It was 25 meters high, and it had a diameter of 6 meters. And its power output was 2 Watts. No. Not 2 kW or even 2 MW. 2 Watts. Enough to light a single LED lamp. Actually, it was designed to withstand 1000 Watts of power output, because that was the theoretical maximum the uranium-vinyl-pellets to feed the reactor could create when going critical. But even then... a 1 kW engine the size of a large grain silo.
Re: What if... (Score:2)
SMR create such a tremendous amount of waste for their power output that we should have banned them from the start. We do not have the infrastructure to support these new waste streams.
Re: (Score:2)
SMRs are not simpler. They are reactors, the same as the large ones.
They can be, SMR is a general term. However the specific Xe-100 reactor of this article seem to be simpler.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
The arguments in favor are simpler, safer, scalable, and easier to build/deploy.
https://small-modular-reactors... [small-modu...actors.org]
Just build fusion instead. I mean we're talking about things which don't exist and have even be proven to work in small pilot plants, so why settle for SMRs. Don't be afraid to dream big darling.
Re: What if... (Score:2)
Build solar panels, those will work until we make a Dyson Sphere, which is basically a more ambitious solar array.
Re: (Score:2)
https://www.nrc.gov/facilities... [nrc.gov]
Re: (Score:2)
You are missing part of the "value" here - they want to push for the SMR design certification and gain operational knowledge about the system, so they can start putting them FUCKING EVERYWHERE.
You start central where you can easily monitor and improve. And then you spread out.
Do you really think Bezos doesn't want to slap down several of these things anywhere AWS has a data center?
Crawl. Walk. Run.
Re: (Score:3)
That is the old idea that still did not manage to make things cost effective. Now they are trying a clearly revolutionary new approach of going against economic reality and they are doing so with an untried design in an engineering space that is exceptionally hard and demanding, because that clearly promises better results! You just do not understand the power of wishful thinking.
Not Cost Effective (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
The Fukushima Japan reactor were USA made.
And it pales in comparison to Chernobyl. And there is the pesky detail that the reactor was not itself faulty like Chernobyl.
Chernobyl was a test scenario that went wrong.
No, the Soviet designed RBMK reactor had numerous internal design flaws that contributed to the disaster.
Now that's a SMR (Score:2)
Tech companies should have nukes (Score:3)
If tech companies can't control AI or stop hackers, then they have no right playing with nuclear reactors. Because these tech companies are 100% putting them online, and they never sufficiently harden products from cyber attacks. It's not just Amazon, but even the small reactor companies are these fly-by-night start ups funded by the same investors in tech and carry much of the same problematic attitude towards balancing profit with responsibility. (Profit maximalists, i. e. no balance whatsoever)
Slashdot is becoming ... (Score:1, Insightful)
... old men shouting at the sky - but not the ones who you think, lol
"LLMs can't program!"
"We can't do nukes!"
"Amazon is evil!"
"When is Matlock on?!?", lol
Re: (Score:2)
The problem is functional illiteracy on your side. For example, it is not "LLMs can't program", it is "LLMs cannot program well" or "getting LLMs to generate good code is more effort and takes more skill than writing good code directly".
Re: (Score:2)
The only one of those being old men shouting at clouds is "LLMs can't program." The remainder is basic observation.
Am I being facetious? You tell me. Here's an article from 15 years ago talking about building SMRs: https://hardware.slashdot.org/... [slashdot.org] and to date we've built precisely zero of them, 15 years later.
Amazon is evil? I mean... yeah unless you live under a rock, objectively yeah. Just look at literally any news article about them.
Re: (Score:2)
Re: (Score:1)
Slashdot has become Reddit.
Pretty much.
Good idea (Score:1)
I'm wondering if the location is correct? (Score:2)
No chance in hell (Score:2)
If they had a prototype with 10 years active use and no serious problems, yes. But as it is, their tech may (and probably will) still fail completely. Getting nuclear engineering right is very, very hard.
Re: (Score:2)
They don't even have an NRC-certified design, much less a prototype or even a single day of operational experience.
This is a press release for Amazon, to show they're not the big bad corporation sucking up all the electricity - this is to appease the ignorants that think "oh well that solves it then, doesn't it!" without actually looking into literally anything about this.
Maybe I'm just being cynical, but it seems that if they were serious about SMRs, they would be talking to the one company that already ha
Re: (Score:2)
Indeed. But even NuScale is a bet on mostly hot air. The thing is, the general public, and apparently most "investors" have no clue how long it takes to get complex industrial machinery from the drawing board to "works well in practice". They probably think 5 years would be long. Realistically, if we have a working, proven and economically feasible SMR design in 30 years, that would be surprisingly fast.
5 gigawatts by 2039? (Score:2)
The Internet suggests that Amazon has datacenters worth 44 gigawatts of usage under construction right now! I mean, I for sure support work towards generating more nuclear power, but this isn't even peanuts, this is "I'll gladly pay you Tuesday for a hamburger today".
$180 Billion for Fukashima Cleanup (Score:3)
https://en.wikipedia.org/wiki/... [wikipedia.org]
And here's a nice little reactor story for you: https://en.wikipedia.org/wiki/... [wikipedia.org]