Why are nuclear power construction costs so high?

All I see is a whole pile of unsupported assertions, let's take them one by one:

> When they accept an expensive out-of-spec part, it's because they can safely redesign the less expensive parts.

'They' presumably being the engineers, and because they are engineers it is automatically assumed that the less expensive parts can always be redesigned. But in the case of a reactor vessel it isn't clear at all how and if that is even possible without compromising on something that apparently originally wasn't to be compromised on. And because the nuclear industry isn't exactly known for their transparency when it comes to such defects it is hard to have visibility on whether what is now no longer the original design really is as safe as what went before.

I think we can agree on at least this simple fact: if the original spec was presumed to be optimal and the new change is still a costly one that there is some pressure to allow a solution to pass that maximizes the economic equation, in this case to redesign the rest of the parts that closely interact with the part that is out of spec. But because the interaction with the reactor vessel is one that is closely based on the operating parameters of the complex as a whole and funds are limited there will be pressure on to compromise. In your world such a compromise would never happen. In mine there is ample evidence that it is and I highly doubt that the nuclear industry is exempt from such pressure to compromise.

The very fact that they did not simply demand a vessel made to spec spells out exactly such a compromise.

> The only thing that hurts is their long-term profitability because it'll be less efficient than they hoped.

That is exactly where the pressure comes from any further compromise will limit that efficiency and hence the profitability (including subsidies) of the plant. So there is pressure to minimize the costs of such a redesign to ensure that the economic damage is limited. The question whether or not that is possible within the original safety margins is an open one, and for at least one reactor I'm aware that safety margins were exceeded on more than one occasion and yet the plant remained open, simply because of a continuous redefinition of what was deemed to still be acceptable. Something that in your world, again, likely is an impossibility:

https://www.laka.org/nieuws/2015/wat-met-het-scheurtjesonder...

Just one example, there are many more (sorry, this one is in Dutch, it is about the plants that I know most about). So there is clear evidence (at least, clear enough to me) of this 'normalization of deviation' that you claim does not exist in this context.

> Big, expensive parts are always treated as one-of-a-kind articles with unique requirements. It's how bridges and roads get built, for example.

Indeed. And bridges never collapse and roads never have problems... In terms of our knowledge about bridges we are still learning new things. Not that long ago that a completely safe and well designed bridge ended up with a whole slew of patches to deal with various resonances in the steel cabling that held up the bridge when the wind was strumming those cables causing massive deflection of the bridge deck, far in excess of what the design originally allowed.

Engineering complex, one-off installations is hard. Reliability and reproduction go hand in hand, only by iteration over a design across many cycles and learning from various defects and errors does engineering progress. It's not just a matter of plugging numbers into formulas, there is a significant amount of feedback from the field about how the assumptions hold up that drives engineering forward and in the case of a one off design that loop doesn't exist. If there is only one reactor vessel and it does not end up being to spec the real effects of that change won't be known until the reactor is decommissioned. Until then we're on ice that we hope is thick enough but that we can not be 100% sure of, see that article linked above. There too engineers ended up being quite surprised at their findings when analyzing the reactor after it had been in service for a while.

> With nuclear, they have special teams that do nothing but manage every type of risk you could ever think of for every piece of nuclear material in the country. > It's very advanced, and many industries are trying to use those techniques themselves now.

This is again a claim that essentially creates an elevated class of engineers who are above making mistakes and who lead the way in ways that I can only assume is through magic. Because in my world engineers do make mistakes, they miss elements in their risk assessment and they make mistakes in their assumptions and sometimes even in the design itself.

But for a one off reactor vessel with a material defect there is no 'plan B', the job could not be called off, so instead of scrapping this vessel and getting one that was built to spec we now work on a cascade of changes. And your claim essentially is that because all these people are so good at what they are doing that they can make this all work without further consequence other than some financial adjustments. My claim is that this isn't a bolt or some other simple part of the reactor and that the safety implications of such a change will not be known until either one of two things happen: the reactor serves out its lifespan, is decommissioned and after analysis of the vessel it is proven that there was no material difference between this one and the one that they originally wanted to have. Or we do find such a difference and in that case we conclude that we were lucky. The third alternative we'll leave unspoken.

> Nobody is normalizing or bending anything.

I don't think you realize that you've essentially made the case for doing just that far more eloquently than I ever could: you are normalizing the deviation by making the claim that it can always be done safely. But how do you know this? In the long, long chain from the QA inspector that faulted the vessel, to the recommendations, to the engineers that redesigned the other parts to the management surely there is pressure from above to solve this, just like there was pressure on NASA administrators to launch and that pressure worked its way downward. And I fail to see the difference between rocket scientists and nuclear power plant engineers. Both are very capable people with very extensive backgrounds in the fields that they are operating in. And if it was just the scientists I'm pretty sure they would have ordered a new vessel and left it at that.

But because there is a political element to this (and politics driven by financial considerations at that) you end up with the exact environment that can lead to this thing called 'normalization of deviation' and that way accidents can and do happen. There is a mountain of evidence for this and I'm not going to close my eyes to that on your say-so. And what goes for the USA may not hold for other countries with less capital and possibly even higher pressures on the management to deliver.

> Every potential problem is thoroughly addressed on an individual basis.

I'm sure it is. Just like in aviation, right? And of course the regulators are not in bed with the likes of GE.

> There is no deviance from the safety requirements.

Blanket unsourced statement. How can you make this claim with such certainty?

My claim is that safety requirements are violated routinely, by people who believe that they are in control of the situation and who have the best of intentions. They're people, after all. And I do have some evidence for my claim:

https://abcnews.go.com/Politics/us-nuclear-power-plants-safe...

> Nuclear is decades ahead of every other field in terms of risk management.

That is not a reason to be super happy about nuclear, but it is a source of worry for all these other fields. And this is probably one of the few things where we agree: that risk management is a field that is still very much under appreciated. And I see that reflected in my practice almost every week.

Er, but it's still vitrified and stable as a glassy material that's less prone to leaching fission products into any water right? As far as I understand it, the whole point of vitrifying it is to trap anything and everything aside from Xenon and Tritium into the vitrified end product.

Ideally, we could have started building out solar in 1989, and would have had today's PV prices by the '90s, and would be well along to mitigating the climate catastrophe now. But we chose otherwise.

Starting nuke construction now would just be kicking the can down the road again.

It would have happened somewhere else, instead, maybe worse. The same flaws were everywhere.

Even that did not prevent Fukushima. Never mind Chernobyl.