Comments by "jeppen" (@jesan733) on "Is Chinese Nuclear Tech Better Than the US? || Peter Zeihan" video.

The Chinese are building 5-ish reactors per year, which is actually a bit slow and won't get them to French levels of nuclear penetration anytime soon. But they are building and researching very broadly, so they are getting ahead technology-wise. They build, research and improve SMR, district heating reactors, pebble bed, thorium reactors, conventional LWR and much much more. The US can't compete because of ridiculous bureacracy. So "if-if-if" the future is nuclear, it'll likely be using Chinese models. This may give them quite a bit of geostrategic and commercial leverage.
26
@heyho4770 In 2023, Chinese coal generation increased more (341 TWh) than their non-hydro RE (296 TWh). Nuclear increased just 17 TWh. But nuclear is rather long-lived and doesn't have the same limitations as VRE. What the Chinese endgame is remains to be seen. They still seem to be taking nuclear very seriously. Lots of research, lots of education, lots of workers being churned out.
6
"If you look at the history of reactor development they started "modular" and then when they got them working well the first thing everyone did was make them as big as they could because it makes the energy so much cheaper." Well, at least they thought so, and perhaps it worked for some time, but was it the right long-term choice? Large cores are harder to cool, and safety considerations have been driving costs. Large constructions also give less experience accumulation and series production. It also give longer construction times, which add to interest, and add financial risk, which also increase interest rates. The pain points today seem very much connected to the enormous sizes. But yeah, I agree the government should make sure we build a bunch of each of promising designs. Then they can compete.
4
@ninefox344 of course I agree that big LWRs are safe enough, but the bureacracies do not. Sweden built in 4-5 years a pop 40 years ago. We need to get down to 4 years or less ideally, but we're up at 12 years or something like that. I think SMRs can help with that. It may be for the wrong reasons, but we need to do whatever works in the world of constraints that we're in.
4
@patrickweaver1105 end of life costs are negligible in an LCOE calculation. The costs and delays in building the things are what matters. SMRs give other economies of scale. If the cost reduction from industrial learning effects is 20% per doubling of produced items, SMRs offer 2-3 additional doublings and could thus halve the cost that way.
2
The NRC has afaik agreed that NuScale emergency planning zone doesn't need to extend further than the site fence. We can put them close if we want to. Also NuScale puts 4-12 reactor modules in the same building, so they've kindof solved putting many together already within the design.
2