No it isn't; when dams fail villages get swept away and everyone dies. Bad mining disasters can wipe out ~50 people too.
When Fukushima failed, basically nobody got hurt, and there is a lot of inconvenience and resources have to go to cleaning the thing up.
Obviously it is still awful for the people involved, but I'd personally rather go through a nuclear accident than a dam failure or have a mine fall on me, because I'd still be basically ok after a nuclear accident. The argument might be that it is uneconomic after accounting for the costs, but it is demonstrably safe and so a far better failure mode than anything else we've tried.
I mean, Fukushima has basically proven that the only question is economic cost/benefit.
Do you have a source behind this? From what I can tell this statement seems wrong. E.g.:
https://www.epa.gov/ghgemissions/global-greenhouse-gas-emiss...
https://ourworldindata.org/energy-production-and-changing-en...
Coal is a huge source of energy. And the dirtiest available. And if electricity+heat production accounts for 25% of global greenhouse gasses, and transportation 14%, there's a decent chance coal power plants contribute just as much to global warming as all of transportation does, and basically zero chance that it plays a "very minor part" when compared to ships, planes, and cars.
I'm not saying that coal is good or nuclear is bad or vice versa. As I said in my original comment, there is no one practical (in terms of safety, efficiency, cost etc) global energy solution right now. I don't believe any coal-powered plants should be built. Should existing ones be replaced by nuclear? Depends on the actual area, geo-political situation (which changes over time) etc.
The failure scenario is demonstrably better than everything we’ve tried. More people have died installing solar and wind than have from nuclear.
My point is that failure of nuclear energy provision is demonstrably worse than the failure of any other options we've tried. Does this mean it's better/worse? Who knows.
But you simply cannot say that something is safe if <xyz>, especially when history says otherwise.
I'm flabbergasted at the number of people talking unironically about "human error" as if it's historically insignificant...