Consider that rising populations and standards of living, as well as shifts from fossil fuels, mean increasing total electrical generation likely by ~500% in the next 50 years or so.

The alternative is a large number of very disappointed people.

True, so we have to divide "centuries" by 5 or more, that's still long enough to help with the transition. Not sure if that's what you meant though? Please correct me if I misunderstood.

200 years / 5 gives about 40 years useful life. Some technologies (thorium MSR) remain unproved, so there's R&D lead time which is likely prohibitive --- we're going to go into this with the designs we have now.

Construction and planning still takes 10--20 years, and commercial viability remains dicey. (Illinois is looking to shut down two nuclear plants not for political/public opposition but economic reasons: https://www.chicagotribune.com/news/environment/ct-exelon-nu...)

Meantime wind and solar keep getting cheaper. They carry few long-time-horizon risks, and are as safe as nuclear within measurement error, a fact nuke boosters tend to omit. Even hydro, as horrific as past incidents have been, is safe when well-managed, and again lacks nuclear's civilisation-spanning risk time horizons.

If presently proven nuclear technologies can smooth the path, then maybe there's a role, but we're still looking at swapping out all but a minuscule fraction of nuclear within a generation (which will mean trying to smoothly phase out an entrenched, powerful, politically-connected, but dying industry, much as today's coal industry), bearing resulting near-term risks, addressing (all but certainly via government subsidies) unfavourable economics, and creating a long-term risk liability spanning millennia.

Meantime, nuke boosters generally continue to paint renewables as the enemy, misconstrue risks, and lowball opportunities, none of which does much for engendering institutional trust in the nuclear community itself.

Don't get me wrong: the challenges are huge, there's plenty of bullshit and opportunism on all sides, and we're going to need to make some really unpleasant decisions and trade-offs here. The Democratic party platform announcement is probably a good thing: it's another option on the table.

But this isn't smooth sailing, and going in with a very realistic view and eyes wide open is critical.

> 200 years / 5 gives about 40 years useful life.

Well, no, it's not that simple. Some advances would require new reactors, such as breeders, but even in a closed system with breeders there would be maybe a 2:1 ratio of traditional burner reactors to breeders (producing fuel for the burners). So even in such a scenario existing LWR's wouldn't be a stranded investment.

But even if we forget breeding for a moment, an increase in usage of uranium would make it economical to exploit poorer deposits (that 200 year figure is at current prices and with current extraction technology), increasing the total economically recoverable deposits (this is similar to practically all other mining, as technology progresses and price increases the economically recoverable deposits also increase). And at some point we will hit the crossover point where, like in the SA article you linked to earlier, extraction from seawater becomes economic. Giving an additional 60000 years of deposits with current usage.

So between breeding (about a factor of 200 improvement in resource utilization) and seawater extraction, we have enough uranium for at least 60000*200 = 12 million years with current usage. Even if we increase usage by a factor of 100 (say, getting rid of fossil fuels thus increasing electricity usage by about ~3x, and increasing the share of nuclear to, say, 50% of the total electricity usage, and still have quite some room for increasing consumption), we still have fuel for 120000 years. Probably more than enough time to get fusion working, or if not, we're hosed anyway in the long run.

> we're going to go into this with the designs we have now.

I agree with that. While there's exiting things on the horizon, it makes sense to start with what's available and mature now.

That, however, doesn't mean we're locked into that forever and ever. We can start deploying traditional LWR designs today, while at the same time commercializing breeders, seawater extraction of U, Th, and whatnot.

None of this means that a "nuclear future" will be inevitable, certainly. I just don't think that the availability of fuel is the limiting factor. AFAICS, the biggest threats to nuclear energy is an inability to deliver on time and on budget combined with political opposition killing the industry.

If wind, solar, geothermal, storage, smart grids etc. manage to kill of the fossil industry without the help of nuclear, hey, I'm ecstatic. I just think having the nuclear option on the table makes success more likely.

Main point is that "200 year supply" is not all that impressive. Demand growth can dominate that rapidly.

When coal was first becoming a mainstream fuel in the US, reserves were estimated as sufficient for 1 million years. Current estimates are for about 100-250 years (latter from BP's Annual Statistical Review for 2019). Demand increased somewhat, and consequences emerged. Now 150 years into the million we're phasing out coal as quickly as possible.

Breeders, thorium (a breeder fuel), and uranium seawater extraction are possible but have proved challenging or limited to date. Seawater extraction particularly presents formidable challenges.

> Main point is that "200 year supply" is not all that impressive. Demand growth can dominate that rapidly.

Sure. But if we keep exponentially increasing energy usage, at some point we'll boil the oceans. And long before that we would have wrecked the earths ecosystem. So I'm assuming that the exponential increase in energy usage will stop at some point.

The 120000 year figure I arrived at in my previous post included a factor of 100 increase over current usage. Even if we add another factor of 100 increase (a total of a factor of 10000 increase over current usage!), we'd still have fuel for 1200 years, surely more than enough to get fusion working.

> When coal was first becoming a mainstream fuel in the US, reserves were estimated as sufficient for 1 million years. Current estimates are for about 100-250 years (latter from BP's Annual Statistical Review for 2019). Demand increased somewhat, and consequences emerged. Now 150 years into the million we're phasing out coal as quickly as possible.

Sure, but ~200 years ago our understanding of geology was quite rudimentary compared to today. And yes, starting from more or less zero we did exponentially increase usage for a couple of centuries. I don't think it's realistic to continue at the same exponential rate for several centuries more, regardless of where the energy comes from (maybe in the far future if humanity starts to look at interstellar travel we would have a usage for such truly stupendous amounts of energy).

Furthermore, we're not phasing out coal because we're running out of it, but due to climate/pollution/economics. From a climate perspective, it unfortunately seems we have more than enough coal left to wreck the climate if we would burn all of it.

> Breeders, thorium (a breeder fuel), and uranium seawater extraction are possible but have proved challenging or limited to date.

The main reason is economics. Uranium is currently just so cheap that it doesn't make economic sense to deploy breeder reactors and reprocessing yet. And without volume deployment, they remain expensive and underdeveloped compared to the current once-through cycle. However, there is no question whether the technology works. It does.

Earlier in the nuclear age (say 1950-1970 or so), uranium was believed to be scarce, isotope enrichment by gas diffusion was very expensive and energy consuming, and we believed the world would soon be powered by atomic energy. So a lot of effort was made to develop breeder reactor and reprocessing technology. However, all of this turned out to be incorrect. Uranium turned out to be quite plentiful, centrifuges made enrichment a lot cheaper, and nuclear power expansion ground to a standstill.

But the earlier R&D showed that the technology is viable, so if uranium prices would start to drastically increase, the option to deploy breeders at scale still exists.

Boiling the oceans would indeed be bad, and I suspect we'll see deviation from long-term growth trends well before that point. But, again, for the third time, readily available fissionable fuels are much less abundant than is commonly understood (or you assert), with a supply very likely much below 200 years at best, and quite possibly only a few decades.

What's changed with regards to coal is not geology or geological knowledge, but consumption. Not mereely increases in existing uses but new applications of the fuel.

Breeders failed commercially for a number of reasons, fuel costs being only one factor:

The story of the fast breeder reactors is a story about a technology embraced by large enthusiasm, which never realized its expectations, at least in the expected time-frame. The reasons for this are many. Partly, it was technical problems, as those with cooling and safety. There were also economical difficulties since the price of uranium did not develop as expected. There were military implications: the problems of handling and transporting plutonium. Also, there were social or ideological complications, since public opposition rose against breeder reactors. The political consequences were that the financial support from the governments disappeared.

https://static.sys.kth.se/itm/wp/cesis/cesiswp186.pdf