[cyberax]

> Nuclear is not really suitable for load-following, at least not the installed capacity.

That's not quite the case. You can load-follow with nuclear, but it requires reactors to be designed for that. France does this, for example.

You also can simply keep reactors working at a constant level and just dump excess power into their cooling system. This is not as bad as it sounds, because fuel is just about ~5% of the total cost of the produced nuclear energy.

Most nuclear power plants do not do this because they don't need to do it.

[pydry]

>This is not as bad as it sounds, because fuel is just about ~5% of the total cost of the produced nuclear energy.

Therein lies the problem. Capital costs dominate nuclear plant costs and they are high.

If you load followed such that you kept the reactor at an average of, say, 50% nameplate capacity that would lead to a levelized cost per MWh of about 2x$168 = $336.

(LCOE listed here is $168: https://www.reuters.com/breakingviews/green-surge-is-circuit...)

For reference, Bhadla solar park sells a MWh for roughly $30, so even if you charged $300 to store and retrieve it you could still provide cheaper electricity than a load following nuclear power plant running at 50% of nameplate.

[cyberax]

I question your estimation of the capital cost. LCOE for most nuclear power plants is way below that: https://www.oecd-nea.org/lcoe/

For the US it's $33 per MWh, so doubling it still gives reasonable cost.

As for wind, it simply can not compete right now for guaranteed capacity. The adequacy rating for most wind power plants is around 10%, so you need 10x overbuild to even compete.

[pfdietz]

Many of the entries in that table (LTO) are for 20 year life extensiom of existing plants, not new plants.

[cyberax]

And so what? Modern nuclear reactors are licensed for 60 years and are expected to last 80-100 years with maintenance (reactor vessel annealing, mainly).

This is exactly a point in favor of nuclear.

[btreecat]

So where's the "solar spills" or "wind contamination?"

One of these solutions is encumbered by problems of safety, the other capacity. It's easier to scale up capacity than it is to scale safety.

If I had to make long term bets, radioactive materials will continue to be radioactive, green capture, storage and transmission will get cheaper and more reliable.

Speaking of transmission, that's another huge problem. You can't go plopping powerplants just anywhere, and power generation needs to be somewhat close to those consuming it.

Distributed collection and storage helps reduce challenges around transmission infrastructure in low density or hard to travel areas.

Consider Puerto Rico, their investments in solar have skyrocketed, especially with the need to rebuild so much infrastructure. They had nuclear, and they were cleaning up contamination for decades after shutting it down. Also part of the issue with power there is transmission. You don't get the efficiency out of tiny boilers to make them cost effective for these folks.

I think you're pushing a bit hard for a tech that has a lot of problems and while it could play a role in our future, it's unlikely to be a dominant force.

[cyberax]

> So where's the "solar spills" or "wind contamination?"

At the factories that produce silicon and composites for the windblades.

> One of these solutions is encumbered by problems of safety, the other capacity. It's easier to scale up capacity than it is to scale safety.

So far, no large country has managed to move to 100% carbon-free renewable generation. And I'm not seeing this changing.

> Distributed collection and storage helps reduce challenges around transmission infrastructure in low density or hard to travel areas.

If we're talking about Europe, they are facing the problem of Dunkelflaute - long periods of no wind, no sun, and low temperatures in the middle of the winter. A worst-case once-in-century scenario would require around a _month_ of storage.

So far no technology is even close to that.

[pydry]

France's are struggling to last 60.

Extending their lives to 100 will not only raise maintenance costs even further it will raise the risk of catastrophe quite significantly.

This would be a catastrophe for which e.g. American plants are insured only up to the level of 0.05% of 1 Fukushima level event.

[cyberax]

You know that material science has improved a lot since 1970-s when most of French nuclear reactors were designed, right?

In particular, annealing had been developed fairly recently: https://www.neimagazine.com/features/featurerenewal-by-annea...

[pfdietz]

If we are trying to determibe the cost of power from new nuclear plants, looking at the marginal cost of upgrading/refurbishing existing plants makes no sense, as that is much cheaper.