I am no power system expert, but enough interesting links on the theme pop up that I may as well record them.
Vaclav Smil et al.
Fun with solar
Solar! who hates solar?
Apparently there is an argument that it does not need to be expensive: Why has nuclear power been a flop?, which asserts that nuclear power suffers from cost disease. I am not qualified to assess that.
Without knowing much about nuclear power, my instincts are that I should be persuaded that nuclear reactors at the current rate will only destroy a town every few decades on average (Wheatley, Sovacool, and Sornette 2017) which is an acceptable toll compared to coal power, for example.
I would require more persuasion that they could be nearly so safe in the kind of deployment the world would need to use nuclear power for climate change reduction. If the number of nuclear fission reactor deployments scaled up massively and was concentrated in countries which most need power, which also tend to have have a poor track record of corruption and active terrorists there are greater base risk rates. I imagine the rate of home-made nuclear weapons and accidental meltdowns might be higher in this case. These risks are rather hard to quantify.