More specifically, when nuclear works well, it works very, very well. When it fails, however, it fails especially spectacularly.
As I said: IPCC's SRREN gives mortality/GWh statistics, and nuclear is not the safest, by that measure. Data included Three Mile Island and Chernobyl, but not Fukushima.
Actually, I've previously discussed both here:
https://www.reddit.com/r/Futurology/comments/267e5e/what_are...
NB: Conca is a rather partisan pro-nuclear advocate. I've found his credibility wanting and bias quite evident in past.
The IPCC's report: http://srren.ipcc-wg3.de/report/
Risks are in one of the appendixes (I'm not on a device which makes sorting this out trivial at the moment).
And yeah, I know it's long....
And p. 745: http://srren.ipcc-wg3.de/report/IPCC_SRREN_Ch10.pdf
History shows that nuclear power is very safe as long as you don't build the plants in tsunami-prone seismic zones, operate them for decades beyond their intended service life, or allow a bunch of unhinged Russians to play "Hold my vodka and watch this" in the control room.
...and waste.
We've also learned that disposal of radioactive waste isn't a problem as long as you mix it with carbon-combustion products and spew it into the air, like coal plants do. Out of sight, out of mind appears to be the optimal way to deal with deadly pollution.
Not all harm involves glowing green blobs of radioactive doom on CNN.