While decommissioning is indeed expensive and as far as I'm aware still mostly based on estimates since there's little experience with the full process, I don't think this statement is correct.

What's really, really expensive is building the plant in such a manner that it is both secure and maintainable.

The problem with securing it is that a critical event is so damn expensive (cf. Fukushima) that you need to secure it against all imaginable problems - and as you can imagine, that list is growing over time. And still, you can't buy insurance for a plant. So the government carries the bulk of the risk.

> you need to secure it against all imaginable problems

It's not that hard if you avoid greedy power plant owners. TEPCO was repeatedly warned many years ago that they had to build a barrier against tsunamis, and all other plants of the area followed the advice.

True, but nuclear energy has two important benefits. It has a stable output (so you don't have to invest in energy storage) and it takes up much less space (so it doesn't visually disrupt the landscape as much).

Demand is not stable so you still need energy storage or something else to fill the peaks. The only alternative is not running the plant at full capacity, which will make it even more expensive.

Less space: that's true. But you still need a fair bit of space to the nearest city. :)

That is not "the only alternative" in practice. If you're in a market (which most companies owning nuclear generation are) you'd sell your excess energy, you wouldn't lower the output of your nuclear generation resources.

Yeah, but with nuclear you only have to deal with demand variability, not supply variability.

Unless the 1/1000 failure event happens and all the land for miles around is rendered uninhabitable and unfarmable.

We could have many, many more Chernobyl and Fukushima style disasters, and nuclear would still be safer than real-world coal.

(But that's an easy target to beat.)

All failure events happened due to incompetence and/or greed, and increasingly unlikely due to new designs and improved protocols.

And even assuming those accidents are unavoidable, coal still produces more health problems and deaths.

I'm glad we've abolished greed and incompetence in this future world.

(I'm also not sure why so many people are playing the nuclear/coal dichotomy when we got into this discussion from wind power and the default backup for renewables is now CCGT?)

The situation is completely different now. Reactor design is much safer, operational safety is at another level and safety risks are more clearly understood.

We have had six nuclear accidents with more than local effects in the last sixty years, with hundreds of reactors running. None of them could have happened with modern designs.

In modern nuclear reactors, safety is not an issue.

So intentional circumventing of safety measures and (cost) cutting corners is not possible sich the new tech?

Almost in every industry i can think of there are countless examples of how profit is traded for safety. Why should this be any different?

Safety measures which cannot be intentionally circumvented (except by planting a bomb, I suppose) predate Chernobyl by many years. The RBMK was a horrible design which threw away passive safety[0] for convenience and cost savings.

https://en.wikipedia.org/wiki/Passive_nuclear_safety

x2.

Modern designs for things that can fail dangerously incorporate features that cause the system to go down a branch that ends in a safe failure mode high up the tree rather than allowing the possibility of a less safe failure many more steps down the tree.

It's like how Uhaul overbuilds the living crap out of it's trailers then puts relatively (compared to the rest of the system) light duty tires on them. Better to have idiots blow out tires than other less safe failure modes.

Not sure what your point is.

My point is many technologies were thought of being safe untill you encounter this one error, iceberg, negligence.

Yeah, but at this point, I think the risks of modern reactors are understood well-enough to say with confidence that safety is not a major argument against nuclear energy.

When you say that, does it take into account nuclear where we reuse fuel?

Fuel sourcing and disposal is a small part of the cost.

Edit: a projected cost breakdown for Hinkley Point C! https://www.nao.org.uk/wp-content/uploads/2017/06/Hinkley-Po... (figure 2)

So fuel is not inconsiderable; 6.2bn "fuel costs", 1.8 "fuel management", 2.6 "fuel disposal". From a total of £54.8 bn.

Add 10-50% to actual construction and decommissioning costs for overruns.

And yet still there are so many people that believe that nuclear is the only pragmatic path to a green energy future.