Although I'm not a huge proponent of nuclear power over renewables, I'm not sure the overruns in those projects are a good argument. These projects become hard to cost and understand up front because so few are built. If the UK built 10 then the costs would come down and the knowledge and experience would grow.

That's the problem: the cost doesn't really go down. You can only operate nuclear if you guarantee the prices a decade ahead. That's just not realistic and the end result is that you'll end up subsidizing ever KWh produced and then you still have to factor in decommissioning costs. Nuclear is fantastic technology, but we can do so much better.

They're hard to cost, which means that the numbers that are given are more aspirational than realistic. It's in the interest of those touting projects to be as optimistic as is tolerable when the estimate costs -- and everyone expects them to do this, so if they didn't and were more realistic they'd go nowhere.

Cost overruns are a great argument even if you ignore the massive time overruns, which you can't. All those cost overruns factor into the electricty price forever. Flamanville has a 60 year lifespan for this reason and even then has an estimated cost of €138/MWh [1]. Compare that to Cestas at €108/MWh [2].

Go further south and Spain has recently been paying €25/MWh [3].

[1]: https://www.powermag.com/flamanville-3-europes-hard-won-nucl...

[2]: https://www.reuters.com/article/markets/currencies/new-frenc...

[3]: https://ratedpower.com/blog/spanish-government-solar/