I have a modest set of solar panels on an entirely ordinary house in suburban London. On average they generate about 3,800kWh per year. We also use about 3,800kWh of electricity each year. Obviously, we can't use all the power produced over summer and we need to buy power in winter. So here's my question: How big a battery would we need in order to be completely self-sufficient? Background …
I can dismiss the the other solutions that are worse then pumped hydro because pumped hydro is actually the best case scenario for grid-level storage and it requires A LOT of space. Anything else, batteries, pneumatic mines etc etc are going to be worse in terms of space by orders of magnitude, not to mention the actual costs. Hand waving the need for grid-level storage by saying we would us hydro shows you don’t understand the scale of the problem.
That excerpt from that engineer is great, but WHERE IS THE STORAGE? Show it to me on a map. You can’t because it does not exist. New Nuclear plants are being built, finally, but there is a reason that no grid-level storage exists. It’s literally not possible today. There exists a pilot battery plant in Australia, and there exists a few megawatts of storage in Scotland, but these are few and far between and none of them are suitable for massive deployment.
I can dismiss the the other solutions that are worse then pumped hydro because pumped hydro is actually the best case scenario for grid-level storage and it requires A LOT of space.
It’s like you didn’t even read the bit about how HVDC makes this a non-issue.
. . . but WHERE IS THE STORAGE? Show it to me on a map. You can’t because it does not exist.
It’s in every hydro dam that’s already built in between Arizona and New York. If we even do need more, there is plenty of land to use.
How about this: I throw out everything I said about synergizing different solutions. We just have solar and storage. No long distance transmission or wind. How much does that cost to power a city?
That study has been done. Going by Lazard’s levelized cost of energy 2025 report, the most optimistic cost to build new nuclear is $141/MWh–and keep in mind that I’m giving nuclear the best case scenario here. A solar+storage solution that would provide 97% of the power needed for Las Vegas would cost $104/MWh. “But that’s sunny desert with lots of empty land around it”, I hear you say. The bigger deal is that Washington DC could have 81% of power done at $124/MWh. Northern city where it snows a lot, and it’s still more viable than nuclear.
“But 81% isn’t 100%”. No, please stop. You get to 81% before you get to 100%. This isn’t even the best way to get to 100%.
This study has a comprehensive wind/water/solar solution fighting with two arms tied behind its back, and it’s still kicking nuclear’s ass.
. . . New Nuclear plants are being built, finally
Nope, not in the US, they aren’t.
Here’s a map of NRC licenses. The green pips are the ones where licenses are already approved. Here’s the list and where they are at:
William States - Licensed to go ahead in 2016. Canceled in 2017 with a contributing factor being the bankruptcy of Westinghouse (which itself happened because of cost overruns at the Vogtle nuclear plant build)
Turkey Point - Licensed new builds in 2018. No news on actually going forward.
North Anna - Licensed new builds in 2017. No news on actually going forward.
PSEG - Issued an early site permit, but not the full license. The ESP was set in 2016 with no movement noted since then.
Fermi - This was licensed just in the past few months. They want to have it in operation by 2032, which, lol, no it isn’t.
That’s not a list of success stories. Add the Vogtle debacle to the list and it’s all a bucket of failure.
The AP1000 design at Vogtle was supposed to prevent the need for botique engineering that had been a problem with reactors in the past. You could use one design everywhere. That was hoped to prevent all these cost and schedule overruns. It didn’t. In addition to Vogtle, it was also built in China at the Sanmen and Haiyang plants. Like Vogtle, Sanmen went over budget and over schedule, but managed in the end. There’s less information about what happened at Haiyang, but the timeline of beginning construction and reaching first criticality is roughly the same as Sanmen; we can assume it went about the same.
There’s a very clear reason why this is happening, and it comes down to this chart:
This is a list of megaprojects and their tendency to go overbudget. Everything from rail to mining to airports. The third worst budgetary offender is nuclear power at a mean cost overrun of 120%. It managed to be better than Olympic Games, at least. The very worst is the related issue of nuclear storage at a whopping 238% mean budget overrun.
Way down at the bottom, you will find solar, power transmission, and wind. Solar projects have a mean overrun of 1%, energy transmission 8%, and wind 13%.
That should make it very clear why the list above has approved licenses with no actual movement. Who the hell would want to put their money into that? You can invest in wind or solar, have a very good chance of it staying within budget, and it will be making revenue within 6-12 months. You put that in nuclear, and you better hope that other investors will pitch in when the budget doubles, or else you have to do it if you hope to see your money again. In the very best case scenario, you’re not going to see a cent of revenue for at least 5 years, but probably more like 10.
Meanwhile, old nuclear is being taken offline because it’s too expensive. If it’s not even worthwhile to keep what we have, what hope is there for building new?
It’s in every hydro dam that’s already built in between Arizona and New York. If we even do need more, there is plenty of land to use.
This is the key factor I’m talking about. There is not “plenty of land” for hydro storage, and flooding the amount of land required to provide grid level storage is an ecological disaster. Plus your analysis of mega-project like nuclear plants going over budget and over-time absolutely applies to any grid-level storage project you would need to go 100% solar/wind.
But just for fun, how much space would the grid level storage projects take up? I’ll let you use Hydro because it’s the best case scenario that exists today as far as energy density.
But beyond that what is your point, that humans shouldn’t build big projects, and any attempt to do so is “boneheaded?” Capitalism can’t build big projects I agree, but the problem isn’t the projects themselves it’s the profit-motive.
I can dismiss the the other solutions that are worse then pumped hydro because pumped hydro is actually the best case scenario for grid-level storage and it requires A LOT of space. Anything else, batteries, pneumatic mines etc etc are going to be worse in terms of space by orders of magnitude, not to mention the actual costs. Hand waving the need for grid-level storage by saying we would us hydro shows you don’t understand the scale of the problem.
That excerpt from that engineer is great, but WHERE IS THE STORAGE? Show it to me on a map. You can’t because it does not exist. New Nuclear plants are being built, finally, but there is a reason that no grid-level storage exists. It’s literally not possible today. There exists a pilot battery plant in Australia, and there exists a few megawatts of storage in Scotland, but these are few and far between and none of them are suitable for massive deployment.
It’s like you didn’t even read the bit about how HVDC makes this a non-issue.
It’s in every hydro dam that’s already built in between Arizona and New York. If we even do need more, there is plenty of land to use.
How about this: I throw out everything I said about synergizing different solutions. We just have solar and storage. No long distance transmission or wind. How much does that cost to power a city?
That study has been done. Going by Lazard’s levelized cost of energy 2025 report, the most optimistic cost to build new nuclear is $141/MWh–and keep in mind that I’m giving nuclear the best case scenario here. A solar+storage solution that would provide 97% of the power needed for Las Vegas would cost $104/MWh. “But that’s sunny desert with lots of empty land around it”, I hear you say. The bigger deal is that Washington DC could have 81% of power done at $124/MWh. Northern city where it snows a lot, and it’s still more viable than nuclear.
“But 81% isn’t 100%”. No, please stop. You get to 81% before you get to 100%. This isn’t even the best way to get to 100%.
This study has a comprehensive wind/water/solar solution fighting with two arms tied behind its back, and it’s still kicking nuclear’s ass.
Nope, not in the US, they aren’t.
Here’s a map of NRC licenses. The green pips are the ones where licenses are already approved. Here’s the list and where they are at:
That’s not a list of success stories. Add the Vogtle debacle to the list and it’s all a bucket of failure.
The AP1000 design at Vogtle was supposed to prevent the need for botique engineering that had been a problem with reactors in the past. You could use one design everywhere. That was hoped to prevent all these cost and schedule overruns. It didn’t. In addition to Vogtle, it was also built in China at the Sanmen and Haiyang plants. Like Vogtle, Sanmen went over budget and over schedule, but managed in the end. There’s less information about what happened at Haiyang, but the timeline of beginning construction and reaching first criticality is roughly the same as Sanmen; we can assume it went about the same.
There’s a very clear reason why this is happening, and it comes down to this chart:
https://energyskeptic.com/wp-content/uploads/2024/03/Why-large-projects-fail-Flyvberg.jpg
This is a list of megaprojects and their tendency to go overbudget. Everything from rail to mining to airports. The third worst budgetary offender is nuclear power at a mean cost overrun of 120%. It managed to be better than Olympic Games, at least. The very worst is the related issue of nuclear storage at a whopping 238% mean budget overrun.
Way down at the bottom, you will find solar, power transmission, and wind. Solar projects have a mean overrun of 1%, energy transmission 8%, and wind 13%.
That should make it very clear why the list above has approved licenses with no actual movement. Who the hell would want to put their money into that? You can invest in wind or solar, have a very good chance of it staying within budget, and it will be making revenue within 6-12 months. You put that in nuclear, and you better hope that other investors will pitch in when the budget doubles, or else you have to do it if you hope to see your money again. In the very best case scenario, you’re not going to see a cent of revenue for at least 5 years, but probably more like 10.
Meanwhile, old nuclear is being taken offline because it’s too expensive. If it’s not even worthwhile to keep what we have, what hope is there for building new?
It’s not a matter of regulation, either. The industry would really like it to be, but they’ve been putting their thumb on that scale for a while now. Even with that, nobody wants to finance this shit.
It’s not just that nuclear is expensive. It’s a boneheaded thing to drop money into at all.
This is the key factor I’m talking about. There is not “plenty of land” for hydro storage, and flooding the amount of land required to provide grid level storage is an ecological disaster. Plus your analysis of mega-project like nuclear plants going over budget and over-time absolutely applies to any grid-level storage project you would need to go 100% solar/wind.
But just for fun, how much space would the grid level storage projects take up? I’ll let you use Hydro because it’s the best case scenario that exists today as far as energy density.
But beyond that what is your point, that humans shouldn’t build big projects, and any attempt to do so is “boneheaded?” Capitalism can’t build big projects I agree, but the problem isn’t the projects themselves it’s the profit-motive.
We already built it. Good bye.