Hacker News new | ask | show | jobs
by colordrops 2585 days ago
Exactly, so how is everyone so sure that "modern nuclear reactors" will encounter no problems?
2 comments

To use one example: Nuclear fuel is in liquid form, in a vessel that has a frozen plug of material being actively cooled at the bottom. If the reactor ever loses power, the plug melts, and the fuel drains into a collection pool, where due to simple geometry it cannot reach criticality.

Something bad (earthquake, tsunami, etc) could still happen that causes a plant shutdown. But there's no pressure vessel (other than regular steam on the generating side) to worry about breaching. And unlike current generation reactors, where even a full scram of control rods still requires some hours of active cooling to prevent core meltdown, if the liquid-type loses power, the core drains, and that's the end of it in terms of potential radiological release. That doesn't mean it isn't a complex, difficult system to design and operate. But it's a much, much safer design by its very nature.

I'm sure there was some explanation as to why superscalar CPU architectures were secure and foolproof as well. The point is that no human mind is capable of simulating reality and accounting for every single possibility. There are design flaws and unexpected surprises that make even the most well thought out plans and explanations seem foolish in retrospect. The point is determining the worst case and assuming it will happen.
I think Asimov's The Relativity of Wrong applies here.

( https://chem.tufts.edu/AnswersInScience/RelativityofWrong.ht... , it is an enjoyable essay to read. )

You are correct that flaws and surprises lurk in all human endeavors. But that does not mean that all potential errors are in the same scale or class. We do not know everything, but we know far more than nothing!

I have never read anyone say that ever.