[pydry]

They're good until 2030 and then they're all going to start aging out at more or less the same time.

Nuclear power is extraordinarily capital intensive but fairly reasonably priced to run once it's built.

This also creates a strong incentive to run plants beyond their operating lifetime, which France may do. That takes us into exciting new territory safety-wise.

[Trombone12]

And then it becomes comes time to dismantle them, and they turn capital intensive again.

[api]

This tail end expense is maybe the largest problem with nuclear power. Economic systems and corporate structures are not set up for this. The whole economy is used to things costing nothing at the end of their life, or nearly so.

Done with a car? Scrap it, strip it, recycle the recyclables, and toss the rest. Done with a building? Implode it and cart it away. Same goes for almost all other pieces of machinery and capital.

Done with a nuke? Now you have another cost almost as high as building it to begin with! It's like having to buy your car and then spend almost the same amount to un-buy it at the end of its life span.

Yes such things can be priced in, but it requires a tremendous amount of foresight and discipline. There is a constant temptation to cut corners on any cost that won't be incurred for a long time, especially if that time is beyond the term of a politician or the career of a corporate bureaucrat.

Done with a solar panel? Send it to a bulk electronics recycler. Done with a wind turbine? Scrap it like any other piece of heavy machinery.

There are other industries with non-trivial decommissioning costs, like chemicals, oil and gas, etc., but at least the time frame is reasonably short. Nuclear decommissioning costs drag on and on, theoretically many times longer than the plant's useful life span.