Even accounting for the times things have gone “catastrophically wrong”, nuclear is many orders of magnitude safer per unit of energy than every other energy source except solar.
The death rates might be a difference in units; the Forbes article is using deaths per trillion kWh, the other might be deaths per thousand/million kWh.
The difference in ranking might be down to how they model deaths from nuclear power accidents. One may be using the linear no threshold model, and the other may be using something else. We don't have an agreed upon model for how likely someone is to die as a result of exposure to X amount of radiation, which causes wide gaps in death estimates.
E.g. Chernobyl non-acute radiation death estimates range from 4,000 to 16,000, with some outliers claiming over 60,000. That's a wild swing depending on which model you use.
Sure, in deaths per unit energy. But the real risk of nuclear is financial. The tail risk is huge for any producer on their own, which makes insurance extremely expensive, and which means that usually only nations bear the full financial risk of nuclear.
These are death rates with the current saturation of plants. If we wanted to cover all of Europe, a much more densely populated area, with nuclear, the deaths (and other negative consequences) would be far greater, no?
The thing about nuclear is that the land area consumed per unit of energy is, like the deaths per unit of energy, extremely low. You can “cover all of Europe” without needing to put very many people (if any) in the potential exclusion zone.
Even with that being said, those safety numbers have held even with China building large numbers of reactors in relatively dense areas. I'd be surprised if European reactors turned out to pose much of a higher risk.
https://ourworldindata.org/grapher/death-rates-from-energy-p...