BUT, the potential for harm from nuclear accidents is vastly higher, and also lasting (though one can make a competitive case for fossil fuel impact on global warming).
In any case, "the big one" hasn't happened yet. It may not ever happen, but if it does, ONE incident can cause incalculable harm.
I think potential for harm needs to be assessed in addition to just death per kilowatt hour.
Edit : I am specifically replying to the phrase "the impact to human life is minimal" when referring to Fukushima. I am not comparing harm of nuclear versus harm of fossil fuels, etc. Fukushima's consequences were, and still are now to a very large number of people, including displaced, farmers, fishermen, and even the effects on commerce and relations between Japan and neighbouring nations ( Taiwan's food ban, etc ), anything but minimal. How much cancer cases increased is not the relevant measure point here.
Nuclear is by far the lesser of the two evils, providing baseload level amounts of carbon neutral energy with nearly a century lifespan, highly regulated and being far, far safer than any competing energy source.
To get Hinkley Point C built, UK taxpayers have to guarantee an electricity price well above market rate, and the taxpayer has to underwrite all the loans, and the taxpayer will be on the hook for any accident cleanup costs above the reactor's $1bn insurance, and the taxpayer will be storing the waste, and guess who'll be paying for the decommissioning? That's right, in all probability it'll be the taxpayer.
In nuclear is the future, we're in for a very expensive future.
This was in part (or, from what I read some time ago, mostly) caused by irrational fear and overrreaction to the incident, causing an evacuation much larger than necessary.
> Chernobyl and Fukushima represses pretty much the worst case...
And yet, even now people are still trying to piece together the scope of the damage that was done. These accidents were already incalculable in the harm they've done.