Secondly, to Noname, the only two nuclear disasters, Chernobyl and Three Mile Island, were caused by a violation of the protocol for the operation of nuclear power plants. So the operation of nuclear power plants is pretty safe when the rules are followed.
Sorry, but you are mistaken there. Undoubtedly Chernobyl was the worst accident in the history of civil nuclear technology and Three Mile Island was one which could have become much worse and got a lot more media attention than many other cases.
For example there was the Kyshtym disaster, in a Russian facility in Mayak in 1957. As much or even more radiation was set free there than even in case of the Chernobyl disaster. The Soviet government was very eager (and mostly successful) to hush up the disaster and prior to 1990 there was no official confirmation for the incident. The main difference to the Chernobyl disaster was that due to the conditions (both of the disaster and the meteorological conditions) most of the fallout and contamination occurred in a comparatively small region around. Due to the hushing up policy over many decades it is impossible to determine how many people exactly died, but to this day a high percentage of people living in the area are suffering from leukemia and other lethal illnesses that can be traced back to exposure to high levels of radiation.
Also in 1957 there was a fire in the British plant at Windscale during which radiation was set free (fortunately in a much lesser degree than in Mayak). On the seven level scale of severity of nuclear disasters (Chernobyl being the only that got rated 7 so far) this one was rated 5. Reluctance from the government to investigate in detail at the time again makes it impossible to determine the exact number of fatalities caused from exposure to radiation set free in the accident, but a 1982 report wrote of at least 40 fatalities that can be directly linked to the accident. Other reports of later times had higher numbers, but due to the initial hushing up and the time lapsed since then it becomes difficult to provide exact numbers.
In 1961 there was an accident in a reactor of the Idaho national Laboratories. Three workers were killed by the radiation and were so contaminated they had to be buried in lead coffins. Most of the radiation was limited to the close surroundings of the place, but even in a distance of 80 kilometers the radiation was more than twice the normal level (and that's a rather lucky outcome).
In 1964 a facility worker in Charlestown / Rhode Island was killed by a lethal amount of radiation. Since it was "only" this one man and no radiation was set free this is basically classified as a "normal" work accident, but it is still one out of many similar cases.
I could go on with the list, but the main point is that there have been quite a number of accidents or near accidents where the worst was only barely prevented. So long things only just turn out to be not quite as fatal as could be we are quite ready to accept the risk. But come the day a nuclear power plant in a more densely populated area fails disastrously we will not be able to pretend that we had no idea of the risk.
But even if we ignore the chance of avoiding the worst case scenario there is still the unsolved problem of what to do with the nuclear waste which radiates for hundreds or thousands or (in some cases) many more years? To this day there hasn't been any satisfactory solution to the problem.
PS: I mentioned the scale from 1 to 7 earlier in this post. This scale is the International Nuclear Event Scale.
7 - Major Accident - one case (Tschernobyl)
6 - Serious Accident - one case (Kyshtym disaster)
5 - Accident with wider consequences (three cases including Harrisburg)
4 - Accident with local consequences (this is the lowest stage of the scale in which people outside the plants may be affected or on which people die. I didn't find a number on how many cases have occurred. Wikipedia lists six cases as "examples" and searching for a number I found reference to other cases not listed on that list of examples so there seems to be a certain casualness about this one.
3 - Serious incident
2 - Incident
1 - Anomaly
Those three last cases apparently occur on such a regular basis that I couldn't find the exact numbers anywhere (though plenty of references to individual cases). Given the fact that all those 4 to 7 cases started out with level 1 anomalies I didn't find this particularly reassuring.
