Amazon just bought a 100% nuclear-powered data center(electrek.co)
electrek.co
Amazon just bought a 100% nuclear-powered data center
https://electrek.co/2024/03/05/amazon-just-bought-a-100-nuclear-powered-data-center/
17 comments
You can't.
Nuclear currently is 6x as expensive as wind and solar. Emphasis on currently.
Now nuke plants take 10 years, and alas they aren't getting cheaper. Solar / wind however is almost certainly going to get cheaper. Solar has perovskites poised to revolutionise it, and both have a huge economy of scale coming in manufacturing.
Storage wise sodium ion and longer term sodium sulfur will vastly cut the cost of grid storage.
So by the time you build a nuke plants in 10 years, it will probably be 10-12x as expensive.
I think nuclear needs a paradigm shift to reduce its costs. I think something around LFTR which scales down and produces far less waste might move the needle, to hopefully get to a projected cost of only 3-4x as expensive, but of course no design is commercialized. Maybe China will succeed.
Nuclear currently is 6x as expensive as wind and solar. Emphasis on currently.
Now nuke plants take 10 years, and alas they aren't getting cheaper. Solar / wind however is almost certainly going to get cheaper. Solar has perovskites poised to revolutionise it, and both have a huge economy of scale coming in manufacturing.
Storage wise sodium ion and longer term sodium sulfur will vastly cut the cost of grid storage.
So by the time you build a nuke plants in 10 years, it will probably be 10-12x as expensive.
I think nuclear needs a paradigm shift to reduce its costs. I think something around LFTR which scales down and produces far less waste might move the needle, to hopefully get to a projected cost of only 3-4x as expensive, but of course no design is commercialized. Maybe China will succeed.
Nuclear costs more due to construction regulation not due to op cost.
The only place I've seen quoted to the contrary is Korea, but the problem with Korea is that large corporate ventures are practically part of the government, so costs can magically be placed elsewhere. And South Korea is a lot more "shadow authoritarian" than people realize. Anyway, the US between federal / state / local simply doesn't have the Korean governmental model, to say nothing of the government <-> corporate interface.
And it doesn't matter why, the cost of nuclear is the cost, and improving the cost with solid fuel designs seems like a dead end. The problem with all the nuclear startups is they got tainted by NuScale, and I've been told that NuScale is a special case of incompetence, before it cancelled the projects it was mentioned along with other "new nuclear" startups. So what will happen when the other "new nuclear" startups rubber hits the road and can't deliver the cost savings?
So when the industry basically says that NuScale wasn't "real next gen nuclear" as soon as they fail, I take a dim view. Just sounds like an industry desperate to keep the investment money train going, not one that has a legitimate answer.
The LCOE economics of wind/solar/storage are just a brutal hammer, solar in particular, because I believe that solar will drop to 1/3 the current cost or less in another 10 years. That has to be hard to look at for other power generation forms. You cling to existing funding, grid leveling, and hold on as long as you can.
I do believe that there theoretically exists a nuclear formula that can be competitive, the power density of nuclear energy is just too many orders of magnitude of superiority.
The LFTR/MSR design, practicalities of molten salt aside, showed:
- a reactor can scale from closet sized and up
- can breed fuel from non-fissile uranium and thorium
- is virtually meltdown proof with the plug design
- uses a far higher percentage of fuel / generates far less waste
- could theoretically be used to breed/consume existing solid fuel waste
From what I've read of pebble bed, that has some of those abilities as well, maybe all of them if you could reprocess the spent pebbles economically.
But the industry, from people to support to regulations to engineering, is focused on huge massive solid fuel rod designs, and those simply aren't going to be competitive even if wind/solar never drop in price. And again, if you are in the business of trying to keep nuclear relevant, you should plan for wind or solar to drop to 1/3 of the current costs in 10 years, and plan for practical grid storage of a day or two to not be more than 10-20% premium over the cost.
And solar+storage is very very scalable. So scalable that, even with the outrageous markups on home consumer solar, is very economical.
I also think that there was a political/cultural issue in nuclear. Almost everyone I talk to in the nuclear industry is quite conservative, I think that comes from the left owning the anti-nuclear political tentpole. But nuclear was a green power solution to address global warming/carbon neutrality. If it had been aggressively positioned in the early days of global warming's scientific findings in the 80s or even 90s when IPCC reports started coming out... but the entire industry is right-wing it seems, so that was simply against the political culture of nuclear.
Being a "right wing" culture for nuclear means that the regulatory/safety aspects of nuclear are treated with scorn and derision. Something like LFTR which had safety and (allegedly) nonproliferation built-in was dismissed by the nuclear establishment early on. So as someone that loves the technology and science of nuclear, but looks into the nuclear power culture and politics, it seems like a bunch of bitter, unreasonable people set in their ways, and unfortunately, that's on the good end of evaluating things.
It is pretty frustrating as a nuclear advocate to have things like Fukushima happen, rife with bad engineering decisions, and to REALLY compound things, you get a spotlight into TEPCO management failings and corruption. And of course the nuclear industry will try to say that is Japan's management cultural issues. But... I think those management issues exist in bland US power companies as well.
I just thing the entire "new nuclear" industry needs a total reboot with different people, a different culture, a different approach. Because the previous culture simply doesn't work. And that probably can't happen until the lights shut off on old nuclear plants, but for now I support keeping those running for grid leveling, because the alternative is natural gas turbine or (shudder) coal.
I think the US national labs should start doing some moonshot nuclear research. I see that MSR research is being funded now (probably because China has been doing it). What would solve all of nuclear's problems is a cheap scalable reactor that has little waste and is meltdown proof that is provably just 2-3x as expensive as wind/solar.
And it doesn't matter why, the cost of nuclear is the cost, and improving the cost with solid fuel designs seems like a dead end. The problem with all the nuclear startups is they got tainted by NuScale, and I've been told that NuScale is a special case of incompetence, before it cancelled the projects it was mentioned along with other "new nuclear" startups. So what will happen when the other "new nuclear" startups rubber hits the road and can't deliver the cost savings?
So when the industry basically says that NuScale wasn't "real next gen nuclear" as soon as they fail, I take a dim view. Just sounds like an industry desperate to keep the investment money train going, not one that has a legitimate answer.
The LCOE economics of wind/solar/storage are just a brutal hammer, solar in particular, because I believe that solar will drop to 1/3 the current cost or less in another 10 years. That has to be hard to look at for other power generation forms. You cling to existing funding, grid leveling, and hold on as long as you can.
I do believe that there theoretically exists a nuclear formula that can be competitive, the power density of nuclear energy is just too many orders of magnitude of superiority.
The LFTR/MSR design, practicalities of molten salt aside, showed:
- a reactor can scale from closet sized and up
- can breed fuel from non-fissile uranium and thorium
- is virtually meltdown proof with the plug design
- uses a far higher percentage of fuel / generates far less waste
- could theoretically be used to breed/consume existing solid fuel waste
From what I've read of pebble bed, that has some of those abilities as well, maybe all of them if you could reprocess the spent pebbles economically.
But the industry, from people to support to regulations to engineering, is focused on huge massive solid fuel rod designs, and those simply aren't going to be competitive even if wind/solar never drop in price. And again, if you are in the business of trying to keep nuclear relevant, you should plan for wind or solar to drop to 1/3 of the current costs in 10 years, and plan for practical grid storage of a day or two to not be more than 10-20% premium over the cost.
And solar+storage is very very scalable. So scalable that, even with the outrageous markups on home consumer solar, is very economical.
I also think that there was a political/cultural issue in nuclear. Almost everyone I talk to in the nuclear industry is quite conservative, I think that comes from the left owning the anti-nuclear political tentpole. But nuclear was a green power solution to address global warming/carbon neutrality. If it had been aggressively positioned in the early days of global warming's scientific findings in the 80s or even 90s when IPCC reports started coming out... but the entire industry is right-wing it seems, so that was simply against the political culture of nuclear.
Being a "right wing" culture for nuclear means that the regulatory/safety aspects of nuclear are treated with scorn and derision. Something like LFTR which had safety and (allegedly) nonproliferation built-in was dismissed by the nuclear establishment early on. So as someone that loves the technology and science of nuclear, but looks into the nuclear power culture and politics, it seems like a bunch of bitter, unreasonable people set in their ways, and unfortunately, that's on the good end of evaluating things.
It is pretty frustrating as a nuclear advocate to have things like Fukushima happen, rife with bad engineering decisions, and to REALLY compound things, you get a spotlight into TEPCO management failings and corruption. And of course the nuclear industry will try to say that is Japan's management cultural issues. But... I think those management issues exist in bland US power companies as well.
I just thing the entire "new nuclear" industry needs a total reboot with different people, a different culture, a different approach. Because the previous culture simply doesn't work. And that probably can't happen until the lights shut off on old nuclear plants, but for now I support keeping those running for grid leveling, because the alternative is natural gas turbine or (shudder) coal.
I think the US national labs should start doing some moonshot nuclear research. I see that MSR research is being funded now (probably because China has been doing it). What would solve all of nuclear's problems is a cheap scalable reactor that has little waste and is meltdown proof that is provably just 2-3x as expensive as wind/solar.
I tried to look up how many potentially used nuclear powered ships are out there, since you could just drop an anchor and a ton of power all over the world with one, but it seems like a surprisingly short list outside of US navy. I guess it would make sense due to nonproliferation. What a shame though. Imagine if every other container ship had one of these power plants how many remote areas would probably meet a great deal of their energy needs from a battery array in their shipping port, rather than the current paradigm where they must provide dirty fuel for shipping companies that goes on to pollute the harbor.
Imagine a nuclear powered container ship captured by pirates.
Would not do them any good, without the skilled manpower to operate and maintain it
As I understand it, naval reactors are cooled by seawater. Couldn't the coolant be diverted, causing the reactor to overheat and eventually meltdown? How dangerous a naval reactor meltdown would be I have no idea.
Definitely not good, but wayyy less dangerous for humans and the environment than a huge oil spill
Why are they so risky? I thought it was mostly about the gigantic price tag.
It is about the overall build time, giant price tag and later on risks involved.
I am intentionally oversimplifying a bit, but for a new solar park, you ram metal poles into the ground, run metal beams cross those, put panels on those and cable all of these into a transformer with a lot of electronics. If that transformer detects a problem, it cuts power to the grid and that's it. Then you dig up a short caused by moles, replace it and turn it on again. Wind turbines aren't far off of that thing as well.
Wih nuclear you have for example the question of getting fuel. There is little nuclear fuel available outside of ex-soviet control. Efficient use of that fuel is also tricky, because if you tried to enrich nuclear fuel to make use of nuclear waste, you're 90% the way towards nuclear weapons and suddenly it's globally political.
You have the build time. We can stamp down a lot more solar parks and wind turbine parks in the time a single nuclear plant gets accepted, at least in europe. And also the maintenance time. Due to the planning and build costs of nuclear plants, these need to be planned to last 20 - 40 years. Especially with climate change and climate change turning chaotic - who can say some location is safe to start building a nuclear plant for 15 years, and say safe for 20 - 30 years? Especially if we consider that against renewables, and power storage improvements over these 15 years?
Like don't get me wrong. I also hate it that germany exited nuclear before exiting coal. Which, most likely, was lobby'd by the fossil energy companies. But building new nuclear facilities is too late.
I am intentionally oversimplifying a bit, but for a new solar park, you ram metal poles into the ground, run metal beams cross those, put panels on those and cable all of these into a transformer with a lot of electronics. If that transformer detects a problem, it cuts power to the grid and that's it. Then you dig up a short caused by moles, replace it and turn it on again. Wind turbines aren't far off of that thing as well.
Wih nuclear you have for example the question of getting fuel. There is little nuclear fuel available outside of ex-soviet control. Efficient use of that fuel is also tricky, because if you tried to enrich nuclear fuel to make use of nuclear waste, you're 90% the way towards nuclear weapons and suddenly it's globally political.
You have the build time. We can stamp down a lot more solar parks and wind turbine parks in the time a single nuclear plant gets accepted, at least in europe. And also the maintenance time. Due to the planning and build costs of nuclear plants, these need to be planned to last 20 - 40 years. Especially with climate change and climate change turning chaotic - who can say some location is safe to start building a nuclear plant for 15 years, and say safe for 20 - 30 years? Especially if we consider that against renewables, and power storage improvements over these 15 years?
Like don't get me wrong. I also hate it that germany exited nuclear before exiting coal. Which, most likely, was lobby'd by the fossil energy companies. But building new nuclear facilities is too late.
Wih nuclear you have for example the question of getting fuel. There is little nuclear fuel available outside of ex-soviet control.
Rosatom (Russia) controls less than half of world uranium enrichment capacity, and none of the other ex-Soviet states have enrichment plants:
https://world-nuclear.org/information-library/nuclear-fuel-c...
Rosatom had 46% of world enrichment capacity in 2020.
If you mean uranium mining, ex-USSR states (Kazakhstan, Russia, Ukraine, Uzbekistan) account for slightly more than half of production:
https://world-nuclear.org/information-library/nuclear-fuel-c...
As of 2022 those 4 countries accounted for 55% of mined uranium.
The bigger issue that you correctly identify is that building nuclear plants takes a long time. Also, unlike a wind farm or solar farm, the reactor must be completely finished before it can start earning revenue from electricity generation. Many wind and solar projects start supplying power to the grid when they are less than halfway complete, since they're made up of turbines or racks of panels that work independently.
Rosatom (Russia) controls less than half of world uranium enrichment capacity, and none of the other ex-Soviet states have enrichment plants:
https://world-nuclear.org/information-library/nuclear-fuel-c...
Rosatom had 46% of world enrichment capacity in 2020.
If you mean uranium mining, ex-USSR states (Kazakhstan, Russia, Ukraine, Uzbekistan) account for slightly more than half of production:
https://world-nuclear.org/information-library/nuclear-fuel-c...
As of 2022 those 4 countries accounted for 55% of mined uranium.
The bigger issue that you correctly identify is that building nuclear plants takes a long time. Also, unlike a wind farm or solar farm, the reactor must be completely finished before it can start earning revenue from electricity generation. Many wind and solar projects start supplying power to the grid when they are less than halfway complete, since they're made up of turbines or racks of panels that work independently.
> There is little nuclear fuel available outside of ex-soviet control.
That's by design of the federal government as they strong armed all other countries into Non Proliferation Treaties and made nuclear power a bad word. If they want to go green, they can start by defanging the watchdogs.
That's by design of the federal government as they strong armed all other countries into Non Proliferation Treaties and made nuclear power a bad word. If they want to go green, they can start by defanging the watchdogs.
I'm talking about financial risk btw, not safety risk.
Any large project like a dam or nuclear plant is risky from a financial perspective. You're putting up a lot of capital and probably paying a lot of interest. Larger projects in general tend to run over schedule and over budget too. Back in the day, the US was better at building big things and using public funds for them so it wasn't as risky. Add in the mountain of paper work that the regulators now require for everything nuclear and you have a very high risk project.
It doesn't have to be this way. Other countries have shown that you can consistently build these projects on time and in 4 to 5 years if you have an experienced work force, solid supply chain, and a reasonable regulator. Japan holds the record for a gigawatt scale reactor built in just over 36 months.
Any large project like a dam or nuclear plant is risky from a financial perspective. You're putting up a lot of capital and probably paying a lot of interest. Larger projects in general tend to run over schedule and over budget too. Back in the day, the US was better at building big things and using public funds for them so it wasn't as risky. Add in the mountain of paper work that the regulators now require for everything nuclear and you have a very high risk project.
It doesn't have to be this way. Other countries have shown that you can consistently build these projects on time and in 4 to 5 years if you have an experienced work force, solid supply chain, and a reasonable regulator. Japan holds the record for a gigawatt scale reactor built in just over 36 months.
[dupe]
More discussion over here: https://news.ycombinator.com/item?id=39597184
More discussion over here: https://news.ycombinator.com/item?id=39597184
In many places the answer has been that the state manages the project since sovereign states cannot go bankrupt. But maybe the answer in the US is that the monstrously large tech companies who need 24/7 (and more recently, clean) energy become the anchor customer.