This is a companion post to the Generating Electricity without Fossil Fuels series. For many people, safety is the primary concern for nuclear power. I should address nuclear safety before analyzing it as one of the main possible strategies for moving away from fossil fuels.
Prerequisites: This post goes with Generating Electricity without Fossil Fuels.
Originally Written: March 2021.
Confidence Level: Common knowledge amount nuclear engineers.
There are several important risks to think about for nuclear power:
Proliferation
A country that has nuclear power plants can use them or the associated technology to build nuclear weapons.
Nuclear weapons are terrifying, city-destroying monsters. Hydrogen bombs are even worse. The largest one ever tested, Tsar Bomba, released twice as much energy as the eruption of Mount Saint Helens. We could make even bigger hydrogen bombs, but then individual bombs would have a measurable impact on the global climate. Nuclear weapons could not only make wars much more destructive, they are a threat to all of humanity.
Proliferation is a major risk in the developing world. In the developed world, nuclear weapons already exist, regardless of whether we build nuclear power plants. If the US government wants a nuclear weapon, it’s not going to make one using a civilian power plant. It’s going to get one from its stockpile.
The Non-Proliferation Treaty recognizes the US, Russia, China, the UK, and France as nuclear weapons states. US nuclear weapons are also hosted by Germany, Italy, the Netherlands, Belgium, and Turkey. India, Pakistan, North Korea, and probably Israel all have nuclear weapons outside of the Non-Proliferation Treaty. Most current greenhouse gas emissions come from countries that have nuclear weapons. Other developed countries could build nuclear weapons, but chose not to for reasons independent of how they generate electricity.
Proliferation risk is a valid concern for expanding nuclear power into more countries in the developing world, but not against building more nuclear power plants in the developed world. If eliminating nuclear weapons in the near future were plausible, this argument would be relevant again.
Terrorism
Terrorists could attack a nuclear power plant and cause tremendous destruction. To prevent this, nuclear power plants need to have security. They could also be located outside of major cities and away from major landmarks. The electricity can be transported to the cities.
Our stockpile of nuclear weapons is much more vulnerable to terrorists. Nuclear weapons are mobile and designed to explode. Nuclear power plants are immobile and designed not to explode. The level of security needed for nuclear weapons is more than sufficient for the level of security needed for nuclear power plants.
The US currently has about 5,000 nuclear weapons. We would need about 500 nuclear power plants to produce all of our electricity. The UK currently has about 200 nuclear weapons and would need about 40 nuclear power plants for electricity. Weapons stockpiles are much bigger targets than civilian power plants.
Accidents
Natural disasters or accidents could cause a nuclear reactor to melt down. People think of Chernobyl or Fukushima when they think of nuclear power.
These disasters are uncommon, preventable, and not as bad as disasters from other power sources.
At Fukushima, a 40 year old nuclear reactor was hit by the fourth strongest earthquake in recorded history and then flooded by a tsunami. The radiation released might kill a few hundred people, but this is unlikely to be measurable. At Chernobyl, the reactor was built in a building not designed to contain radiation and the plant manager decided to disable some safety equipment to run a test. The resulting meltdown killed 54 people directly and the radiation released caused about 4,000 deaths by cancer. These are far from normal operating conditions – or even from common crises. No other civilian nuclear power plant has killed anyone.
As a comparison, dam collapses occur more frequently and can be much worse. The worst dam collapse in history was the Banqiao Dam failure in 1975. The death toll is uncertain because the Chinese Communist Party concealed it until the 1990s, but estimates range from 26,000 to 240,000. The worst dam collapse in the US was the Johnstown Flood of 1889, which killed 2,209 people.[1]It also was the first major event for the Red Cross and significantly changed American liability law. Dam failures still occur every few years in the United States, although most do not kill anyone.[2]The most recent fatal dam failure in the US was in Kauai, Hawaii, in 2006. Johnstown, PA has also had a second lethal flood, in 1977. Source.
Poorly designed dams sometimes collapse, but that does not mean that we should abandon hydroelectricity. It means we should build better dams. Poorly designed nuclear reactors sometimes meltdown, but that does not mean that we should abandon nuclear power. It means we should build better nuclear reactors.
Modern nuclear reactor designs, especially molten salt reactors, make meltdowns even less likely than their already good past record indicates.
Forbes has put together an estimate of how dangerous different power sources are. Globally, nuclear is about 1,000 times safer than coal, the most common source of electrical energy. Coal in the US is 100 times more dangerous than nuclear globally. Wind energy is half again as dangerous as nuclear, while solar is about five times as dangerous as nuclear.
Nuclear does not deserve its reputation as an unusually dangerous power source.
EDIT [2022-02-06]:
Our World in Data has also weighed in on this question. They discuss what are the safest and cleanest sources of energy (as opposed to just electricity), and get consistently lower numbers than Forbes. Here is a comparison of their numbers of deaths per terawatt-hour:
| Forbes | Our World in Data | |
| Coal | 100 | 24.6 |
| Oil | 36 | 18.4 |
| Natural Gas | 4 | 2.8 |
| Biomass | 24 | 4.6 |
| Hydropower | 1.4 | 0.02 |
| Nuclear | 0.09 | 0.07 |
| Wind | 0.15 | 0.04 |
| Solar | 0.44 | 0.02 |
I don’t want to do a detailed dive into the sources. Normally, I would trust Our World in Data more than Forbes. However, Our World in Data seems to have put more effort into making sure that they got all of the nuclear data than they put into making sure that they got all of the data from other sources. If their data base included all of the nuclear accidents, but only some of the accidents for other sources, that would also explain the differences seen here. The hydropower numbers are especially different because Forbes includes the Banqiao dam failure, while Our World in Data seems to have excluded it.
I will treat these as error bars. The actual death rate might be hard to measure, but it is most likely between the two numbers. The error bars for nuclear and renewables overlap, so it is hard to know which is actually safer. Both renewables and nuclear are orders of magnitude safer than fossil fuels or biomass.
Nuclear does not deserve its reputation as an unusually dangerous power source.