[epistasis]

You have that exactly backwards: solar + storage is what will give us energy abundance at less money than we could ever imagine from nuclear fission or fusion.

China is winning the AI Cold war because it's adding solar, storage, and wind at orders of magnitude more than nuclear.

I'm not sure who's doing your supposed "envisioning" but there is no vision for cheap abundant energy from fusion. Solar and storage deliver it today, fusion only delivers it in sci fi books.

Nuclear is 20th century technology that does not fit with a highly automated future. With high levels of automation, construction is super expensive. You want to spend your expensive construction labor on building factories, not individual power generation sites.

Building factories for solar and storage lets them scale to a degree that nuclear could never scale. Nuclear has basically no way of catching up.

[noosphr]

China has been building out nuclear capacity at 5% a year for 25 years.

Solar and wind capacity had shot through the roof in the last five years because they can't sell hardware to the west any more.

The other big item is hydro power, which China has a ton of untapped potential for. Unfortunately for the West every good river has already been damed so we can't follow them there.

[epistasis]

> Solar and wind capacity had shot through the roof in the last five years because they can't sell hardware to the west any more.

"can't sell hardware??" hah! I've never heard that weird made-up justification, where did you pick it up from?

China installed 277GW of solar in 2024, capacity factor corrected that's 55.4 GW of solar power. That's equivalent to the entire amount of nuclear that China has ever built. One year versus all time. And then in the first half of 2025, China installed another 212GW of solar. In six months.

Nuclear is a footnote compared to the planned deployment of solar and wind and storage in China.

Anybody who's serious about energy is deploying massive amounts of solar, storage, and some wind. Some people that are slow to adapt are still building gas or coal, but these will be stranded assets far before their end of life. Nuclear fusion and fission are meme technologies, unable to compete with the scale and scope that batteries and solar deliver every day. This mismatch grows by the month.

[cyberax]

> China installed 277GW of solar in 2024, capacity factor corrected that's 55.4 GW of solar power.

The problem is not just the mean capacity factor, but the capacity factor in _winter_. It's terrible for China, less than 15%. And more importantly, you can have _weeks_ with essentially zero solar power when you need it most.

[noosphr]

This is not an issue in China as they overprovision demand by 50 percent. Their grid can run off baseload generation alone in their 2060 plan.

Trying to explain that a grid build by electrical engineers, rather than financial engineers, has resilience build in to people whose whole idea about electricity generation is greenwashed bullshit from McKinsey and Co is at best a waste of time and at worst an excellent way to raise one's blood pressure.

[ben_w]

> It's terrible for China, less than 15%.

55.4 GW per 277 GW is an (annual) capacity factor of 20%, so the response here is "yes, and?"

> And more importantly, you can have _weeks_ with essentially zero solar power when you need it most.

Half the country is a mid-latitude desert. What makes you think the whole country has "weeks" with zero solar? And it does have to be the whole country in this case, because one thing a centrally planned economy can do well is joining up the infrastructure, which in this case means "actually make the power grid the USA and the EU keep wringing their hands over".

[cyberax]

> Half the country is a mid-latitude desert. What makes you think the whole country has "weeks" with zero solar?

The "whole country" is irrelevant. You can't transmit arbitrary amounts of power across the large geographic areas, most of energy has to be generated in a reasonably close proximity.

> And it does have to be the whole country in this case, because one thing a centrally planned economy can do well is joining up the infrastructure

Transmission lines are expensive, regardless of your ideology.

[ben_w]

> The "whole country" is irrelevant. You can't transmit arbitrary amounts of power across the large geographic areas, most of energy has to be generated in a reasonably close proximity.

Only technically correct because you said "arbitrary": it's well within China's manufacturing capabilities to make a grid that can transmit 3 TW over 40,000 km, with a conductor cross section so thick it only has 1 Ω resistance.

As in: all the world's current electricity demand, the long way around the planet.

I have, in fact, done the maths on this.

> Transmission lines are expensive, regardless of your ideology.

"Expensive" but not "prohibitively expensive".

All infra is "expensive". Nations have a lot of money.

[ben_w]

> Solar and wind capacity had shot through the roof in the last five years because they can't sell hardware to the west any more.

They can't sell as much as they would like, specifically to the USA, due to tariffs/trade war, but there's a much bigger world out there than just the US, and the overall exports are up over the last five years: https://www.canarymedia.com/articles/solar/chart-chinas-sola...

There's a Chinese-made Balkonkraftwerk sitting a few meters away from me on my patio, it cost €350, of which €50 was delivery and another €50 was the mounting posts, the remaining €250 got me 800 W of both panel and inverter.

> Unfortunately for the West every good river has already been damed so we can't follow them there.

For generation, yes. For storage, no.

[bryanlarsen]

> Unfortunately for the West every good river has already been damed so we can't follow them there.

You don't need a river for hydro power storage. All you need are two reservoirs with a height difference between them. Typically one of the two reservoirs is preexisting and the second is constructed. ANU identified ~1 million potential sites.

https://re100.eng.anu.edu.au/global/

[pfdietz]

> sci fi books

I blame these for the unquestioned belief that fusion is desirable. It's a trope because it enables stories to be told, and because readers became used to seeing, not because science fiction has a good track record on such things.

The fact that the volumetric power density of ARC is 40x worse than a PWR (and ITER, 400x worse!) should tell one that DT fusion at least is unlikely to be cheap.

With continued progress down the experience curve, PV will reach the point where resistive heat is cheaper than burning natural gas at the Henry Hub price (which doesn't include the cost of getting gas through pipelines and distribution to customers.) And remember cheap natural gas was what destroyed the last nuclear renaissance in the US.

[formerly_proven]

It's hard to imagine a form of energy production less desirable than fusion.

Okay, sure, burning lignite and using the exhaust as air heating in the children's hospital. You got me.