• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Microsoft Plans to Build Nuclear-Powered Data Centers

Yes, but even that is not new matter gone radioactive, it's the leftovers from the initial radioactive bars.
Plus, this is about SMRs, which use very little fissionable material. Ok, they don't output that much power either...

Whatever we get from renewables, is unsteady. If you have to have a backup for renewable, SMRs are just about the best candidate.

SMR's do not use "very little fissionable material".
SMR's are about a tenth to a third the size of a common large scale commercial nuclear reactor.

The International Atomic Energy Agency (IAEA) defines 'small' as under 300 MWe, and up to about 700 MWe as 'medium'
A subcategory of very small reactors – vSMRs – is proposed for units under about 15 MWe, especially for remote communities.

The idea of having a "vSMR in every single town" scares the hell out of me.
 
Last edited:
Sure.

And at some point you need to store nuclear waste, which is radioactive for the next 30 to 60 years. Where do you store it you think?

And is MS paying for disposal of that nuclear waste too? Or do we just pass it through the enviroment?
At the same site in underground silo's, like it's done with all nuclear waste these days.

And since we're talking powering a data center the waste isn't even going to be big.
 
Very few countries have this figured out though

14 juni 2023 — Finland is on the verge of becoming the first nation to bury spent nuclear fuel rods deep underground for the long term.
Though only a tiny part of nuclear waste requires drastic solutions.
How does nuclear waste suddenly become tiny?
If you are going to distribute power generation , but have equal total power generation , then you will need the same amount of nuclear fuel and have the same amount of waste.
It is very tiny because most of it is lightly contaminated. It is also very efficient compared to anything else. While not perfect, we are yet to have any solution that it is as clean, efficient and safe as nuclear power.
 
It is very tiny because most of it is lightly contaminated. It is also very efficient compared to anything else. While not perfect, we are yet to have any solution that it is as clean, efficient and safe as nuclear power.
Wow , do you really believe that because these reactors are smaller in size , that the spent fuel is less of a danger?


Its not a new thing, disposal of radioactive waste is pretty clean - you store it.
And yet, despite that it is so simple and easy to store nuclear waste:

Incidents at the Waste Isolation Pilot Plant
 
Last edited:
SMR's do not use "very little fissionable material".
SMR's are about a tenth to a third the size of a common large scale commercial nuclear reactor.

The International Atomic Energy Agency (IAEA) defines 'small' as under 300 MWe, and up to about 700 MWe as 'medium'
A subcategory of very small reactors – vSMRs – is proposed for units under about 15 MWe, especially for remote communities.

The idea of having a "vSMR in every single town" scares the hell out of me.
SMRs have reduced fuel requirements. Power plants based on SMRs may require less frequent refuelling, every 3 to 7 years, in comparison to between 1 and 2 years for conventional plants. Some SMRs are designed to operate for up to 30 years without refuelling.
;)
 
You could do that, but then you'd be distributing fissionable uranium all over the place. Plus, the inefficiency come from transforming heat into electricity, it has little to do with the length of the lines after that.
The main advantage of SMRs, as I see it, is they don't use much fuel. Even in the case of a meltdown, they can't go Chernobyl.
It's pretty easy to manage. Not sure if this is the exact video, but this channels covered SMR and micro-nuclear a few times:

 
Cool! It’s hard to get power to bigger DCs we had a hell of a time expanding capacity and had to spend tons to get lines ran from a different circuit for one of ours. Grids that are already over stressed like California also pose challenges because not only is tje power frequently not clean blackouts are frequent enough for it to cost us. Not to mention it’s bad for the cracs.

I imagine they will subsidize this by either asking local govt for funds in exchange for power they can push back to the grid or pay out of pocket and then recoup by pushing back to the local grid.
 
SMR's do not use "very little fissionable material".
SMR's are about a tenth to a third the size of a common large scale commercial nuclear reactor.

The International Atomic Energy Agency (IAEA) defines 'small' as under 300 MWe, and up to about 700 MWe as 'medium'
A subcategory of very small reactors – vSMRs – is proposed for units under about 15 MWe, especially for remote communities.

The idea of having a "vSMR in every single town" scares the hell out of me.
Better start saving those bottle caps already...
 
Nuclear power is a fantastic idea if you're assuming the plants are going to be run by competent, safety-focused, and attentive personnel who are paid well and supervised by a management team with the same philosophy. However it becomes a bad idea when you realize that all of these companies are managed by profit maximizing, corner cutting, finger-pointing "tech bro" scumbags who will do everything in their power to avoid liability and run these things as cheaply as possible.

Also yeah there's the issue of spent nuclear material but it's fairly straightforward to stack a bunch of metal rods in concrete containers, seal them off and store them until we become efficient at performing fissile material reclamation. We're not dealing with green, glowing goop a la Simpsons, they're little silver metal pellets/rods. It's possible to re-use the remaining viable nuclear material contained within the rods and Japan has already had success with this process.
 
im no energy expert but wouldn't it take alot to cool the data center and the nuclear reactor?
if it does proceed lets hope they not build it in earthquake prone zone

I think a better question is how will a data center need THAT much power?

And at some point you need to store nuclear waste, which is radioactive for the next 30 to 60 years. Where do you store it you think?

We know how to deal with waste, throw it inside another reactor. The reason we don't do it is because this is the same process used to obtain weapons grade plutonium so to try to de-escalate the cold war and from fears of driving nuclear proliferation further Jimmy Carter put that concept in the trash. It's a well known and effective process, the only reason it's not used is because of difficult world politics.

Even so, the waste ammount is not that big and we've been dealing with it adequately for a long time.

but then you'd be distributing fissionable uranium all over the place

Not really, part of the idea of small reactors is having the reactor vessel loaded up will all the fuel it will use. When the reactor reaches end of life it's all packed and ready to go "somewhere".

Even then Chernobyl killed less people than stuff like hydro damn failures do

By official numbers yes but just as soviet dumbfuckery created the biggest nuclear disaster in history it also hid the thousands of direct and indirect casualties of the accident and clean up. It's very hard to come up with a good number but when accounting for the reduced lifespan of people in nearby areas, higher rates of cancer diseaces and so on it's obviously worse than any hydro damn failure (Zhumadian excluded)

How does nuclear waste suddenly become tiny?

Throw it inside a breeder reactor to make new fuel ;)

it's done with all nuclear waste these days

It's not. Some of it is left above ground in concrete caskets and it's fine.

Very few countries have this figured out though

14 juni 2023 — Finland is on the verge of becoming the first nation to bury spent nuclear fuel rods deep underground for the long term.

Not true, being the first to bury spent fuel deep underground is different than few nations having it figured out. The netherlands even offers visits to their storage sites
 
We should focus on nuclear power by building more nuclear power plant at least till we find a better source of power.
Unfortunately the waste is the issue. Not everyone uses it for Military purposes.
 
Ima need one for my PC. How much?
 
This makes perfect sense, and it also hardens the data centers too. MS could also temporarily make an extra buck providing spare power to the local grid until they reach their maximum operating power use from the installed SMR, then shift to building a new nuclear-powered data center somewhere else or expand the complex. Heck, if they throw in some solar panels into the complex and bank that energy to battery packs too, they could have those battery back-ups for the occasional downtime, or again sell off some of that reserve when the nearby grid needs extra power.

If this is the push needed to make SMRs and vSMRs viable and cheap, then I'm all for it. A more robust power grid with multiple SMRs would go a long ways to providing the increased energy requirements while being much safer than legacy nuclear powerplants. And since the fuel is easier to recycle and reuse for other SMRs compared to legacy nuclear, would also help bring costs down. And should the day come when everyone can buy their own 1Mw micro-SMR for their homes, we can start doing away with electrical bills and unsightly solar panel installations.
 
It's pretty easy to manage. Not sure if this is the exact video, but this channels covered SMR and micro-nuclear a few times:


There are some interesting numbers here, the cost of nuclear is over 4x more than green solar/wind energy.

I wonder how they cool these things in case of an emergency shutdown?

And where is the generator located? And the powerlines?
 
Last edited:
And should the day come when everyone can buy their own 1Mw micro-SMR for their homes, we can start doing away with electrical bills and unsightly solar panel installations.

Bad idea, interesting, but very very bad idea!
 
Sure.

And at some point you need to store nuclear waste, which is radioactive for the next 30 to 60 years. Where do you store it you think?

And is MS paying for disposal of that nuclear waste too? Or do we just pass it through the enviroment?

Storage of spent nuclear fuel has come a long way. https://world-nuclear.org/informati...torage-and-disposal-of-radioactive-waste.aspx

As mentioned, Finland is in the construction process with many nations moving forward with specified long term storage sites that will be deep underground. If there was less bureaucracy involved and fewer people trying to hamstring deployment of the tech, it would be further along.
 
Nuclear power isn't perfect, but it's by far the best source of energy we have (excluding geothermal). As for nuclear waste it's a non problem, the amounts are so tiny and when stored properly (by governments not private companies) don't represent any issue.
I agree with the idea that nuclear is the best way forward, but being in Washington and keenly aware of the Hanford Nuclear reservations... reputation, lets just say its not a complete nonissue.

Its not a new thing, disposal of radioactive waste is pretty clean - you store it.
Hanfords above average cancer rates are a thing. Casks leak, stuff happens. Its not an insurmountable issue, but it is an issue regardless.
 
Hanfords above average cancer rates are a thing. Casks leak, stuff happens. Its not an insurmountable issue, but it is an issue regardless.
Systemic issues always threaten anything in terms of security. The longer stuff exists the higher the chance it corrupts and somehow the human factor is one thing we won't ever be able to negate...

There are however also a lot of things we seem to be able to do right, and we trust in its safety. Somehow though with nuclear waste / radiation we fear it more than the safety of say, a flight on an airplane, which is an incredible amount of times more dangerous... or driving a car, yet another many times more dangerous.

"But if it goes wrong it goes really really wrong"... sure, so reduced scale also reduces the magnitude of how really wrong it can get. On the other hand, SMR's are 'scalable' :D I still think if we use more nuclear, we'll learn better how to handle it. Lots of people and countries still live with this silly idea they have a choice in the energy transition wrt nuclear. I don't think we do... especially in more population dense areas.

As long as nobody wants to acknowledge we need to switch economies to 'less' instead of 'more', we'll need nuclear for damn sure.
 
There are some interesting numbers here, the cost of nuclear is over 4x more than green solar/wind energy.

I wonder how they cool these things in case of an emergency shutdown?

And where is the generator located? And the powerlines?
Afaik, modern designs do not need special cooling in case of an emergency shutdown. Whenever power is lost, the uranium rods just fall down (they need power to stay in place), so the critical mass isn't there anymore. The chain reaction breaks then.
 
To all the nuclear naysayers, please educate yourselves.

Air pollution and climate change from burning fossil fuels are visibly, demonstrably, factually, historically-documented by organisations (including governments) as vastly more dangerous to human life than the nuclear waste/nuclear disaster per Megawatt-hour of energy produced; Acid rain damaging food sources, buildings, wildlife, respiratory issues, lung cancer, reduced life expectancy, increased onset of strokes, or other heart conditions, eye and skin allergies all provably linked to fossil fuel pollution - and that's just the side effect in the atmosphere from burning the fuels, never mind the health and environmental damage caused by extracting and refining them. There's a whole bunch of secondary crap that might be caused by fossil-fuel pollution, but not widely-accepted, or with not enough clinical trials or peer-reviews to make governments pass regulations against them yet.

How much was that "vastly" I bolded up there? Estimates vary depending on your source, but somewhere between 2 and 8 orders of magnitude. Not 2-8x more, 2-8 additional zeros on the number of times more. Whilst fossil-fuelled power won't have the horror story of Chernobyl to scare people, the total death toll from historic explosions in traditional coal and gas power plants is much higher, even when weighted by production share of global energy output. https://en.wikipedia.org/wiki/Energy_accidentsl. A good chunk of the deaths from fossil-fuel accidents are from mining/drilling and whilst uranium also needs to be mined, you don't need to do as much mining. 1Kg of uranium 235 produces 2-3 million times more electricity than 1KG of coal, for example.

Is nuclear energy completely safe and problem free? Nope.
Is coal/oil/gas categorically more dangerous to human life, both immediately, and also with less-obvious secondary dangers? Absolutely.

We do need clean energy, and nuclear isn't completely clean, but it'll have to do until renewables solve their national-scale storage shortcomings.
 
Last edited:
Most countries could produce all their electricity needs from nuclear, which runs 24/7 unlike wind which relies on there actually been wind, and solar which works less great when it is cloudy or raining. Nuclear works 24/7 be it windless, cloudy or raining. I agree it is not perfect, but is magnitudes better than any source using fossil fuels. Imo it is going to be too late by the time the world realizes our power savior (for now) is nuclear energy.

Just imagine how clean the skies and air would be if there was zero energy production using fossil fuels. The next target has to be IC engines.
 
Sure.

And at some point you need to store nuclear waste, which is radioactive for the next 30 to 60 years. Where do you store it you think?

And is MS paying for disposal of that nuclear waste too? Or do we just pass it through the enviroment?
an SMR is shut down after it’s life cycle with the fuel remaining inside.
it is left alone for half a century for the fuel to cool down enough for it to be safely handle

for preferably reprocessing and reuse or longterm storage.
 

Its not a new thing, disposal of radioactive waste is pretty clean - you store it.
We store coal and gas powerplant solid wastes too, and it's NASTY stuff in vast quantities - hundreds of millions of tons globally, 70-100 million tons in the US alone. Just one large coal power plant produces more contaminated waste that's harmful to wildlife than every nuclear plant on the planet, because that single coal plant is burning 3.3 million tons of coal every year.
(source)

Meanwhile, only a quarter-million tons of nuclear waste has ever been stored (400K tons total, but 1/3rd of that has been reprocessed):
(source)

edit: a quarter-million tons of waste, ever. In the total history of all nuclear power, going back to 1954, the cumulative waste is about the same as 5 weeks of typical waste out of a single coal powerplant. Yes, it requires special storage, but so does coal waste, of which about 42% of it is genuinely waste that doesn't get re-used and has to be dealt with and regulated by government environmental agencies.
 
Last edited:
Would this contribute to global warming?
 
They send it to poor countries, and now its their problem.
They could always cut it out with the non-proliferation BS and just reprocess the spent fuel. I think the US' stance on this is archaic.
Would this contribute to global warming?
Not in the sense that they contribute to greenhouse gas emissions with the exception of building it. Maybe a little bit from occasional use of diesel generators for backup cooling and whatnot.
 
Back
Top