Comment by Llamamoe
> The problem with nuclear is not the ultra-low probability of incidents, but the potential size of the incidents.
This is also not as bad as people think. Chernobyl was bad, but the real effect on human health was shockingly small. Fukushima is almost as well-known, and its impact was negligible.
Even if we had ten times as many nuclear disasters - hell, even fifty times more - it would still be a cleaner source of energy than fossil fuels.
Meanwhile the amount of overregulation is extreme and often absurd. It's not a coincidence that most operational nuclear plants were built decades ago.
> This is also not as bad as people think. Chernobyl was bad, but the real effect on human health was shockingly small. Fukushima is almost as well-known, and its impact was negligible.
Yeah the final outcome was pretty negligible, especially if we ignore to huge exclusion zone that can’t be occupied for a few hundred years.
But even in those disasters, we often got a lucky as we got unlucky. The worst of the disasters was often avoid by individuals taking extreme risks, or even losing their lives to prevent a greater disaster. Ultimately all of the disasters demonstrated that we’re not very good a reliably managing the risks associated with nuclear power.
Modern reactor designs are substantially safer and better than older reactors. But unfortunately we’ve not building reactors for a very long time, and we’ve lost a huge amount of knowledge and skill associated with building reactors. Which drives up the cost of nuclear reactors even further because of the huge cost of rediscovering all the lost knowledge and skill associated