Modern society runs on electricity. It lights our cities, runs our factories, and powers the computer that I’m typing this on right now. And it will only become more necessary as we electrify various parts of society that still use oil and gas (cars and trucks, obviously, but also stoves, furnaces, and so on). Power plants, however, are also a huge contributor to climate change, not to mention miscellaneous pollution that is harmful in other ways. But it turns out there’s a power source that doesn’t release any greenhouse gases or pollutants, generates loads of power, and has been in use since the 1960s. That power source is, of course, nuclear power. If the world switched to nuclear (or, more accurately, if we hadn’t stopped switching to nuclear in the 1980s), we would be able to solve a pretty big chunk of climate change. But for some reason, we didn’t.
Some might view the decline of nuclear power as evidence of a successful environmental campaign to reduce contamination. The Indian Point nuclear reactor in New York was shut down just a week ago on April 30, to celebration from some environmental groups. But how much of nuclear power’s reputation as a dangerous menace is justified? And why has nuclear power been in decline in the US? To answer that, we’ll have to start from the beginning, at how electricity is generated on a fundamental level.
Pretty much all power plants (except for solar panels) operate under the general principle of “heat some fluid up and use it to spin some kind of turbine.” Coal, oil, and gas plants burn fuel that heats water into steam that makes a turbine spin; geothermal plants do the same thing with already-heated water underground. Hydroelectric dams wait for the sun to heat up some water and carry it up a mountain in the water cycle, then make said water go through a turbine on the way down. Wind turbines take advantage of the sun creating temperature and pressure differences in the atmosphere, which causes air movement which spins a turbine.
Nuclear power plants are fundamentally the same thing as coal, oil, and gas power, except that instead of lighting a fire, they use magic warm rocks, better known as “enriched uranium.” How nuclear reactors actually work on a physics level is beyond the scope of this article, but suffice to say that once constructed, a self-sustaining nuclear reaction is initiated, which generates heat that heats up lots of water that can be used to turn turbines and generate electricity. The first point of comparison would be detonating a nuclear bomb in extremely slow motion, but that gives the wrong impression, because a nuclear power plant has absolutely no potential to become a nuclear bomb. Nuclear bombs must have their nuclear material concentrated in a very specific way to explode, such that it’s pretty much impossible for a nuke to go off by accident — they’re not like conventional explosives, where a single spark can blow everything up. Comparing a nuclear reactor to a nuclear bomb is kind of like comparing a candle to a hand grenade.
“But don’t nuclear plants generate a bunch of dangerous waste products?” Well, kind of, but it really depends on what “a bunch” means. A lot of conversations about nuclear power seem to assume that if we were to switch to pure nuclear energy, we would be faced with a massive logistical challenge of storing gigantic quantities of radioactive waste. But there just really isn’t that much nuclear waste to worry about. In 2010, it was estimated that 250,000 metric tons of what we commonly think of as “nuclear waste” are stored in various repositories around the world. (This is “high level waste,” which makes up only about 2-6% of what is technically classified as nuclear waste; the other 94–98% is “low level waste,” e.g. protected clothing that was worn in proximity to the reactor; this might be faintly initially radioactive but can be safely disposed of as trash in a few years.)
Anyway, 250,000 metric tons is an Extremely Big and Scary Number, except that it’s only about 1.25 times the cargo capacity of that ship that ran aground in the Suez Canal back in March, which has a deadweight tonnage of almost exactly 200,000 metric tons. To be clear, this isn’t an annual rate — all the dangerous waste we’ve produced since we started doing nuclear power would fit into a bit more than one very large container ship. For another bit of scale, 250,000 metric tons is roughly the amount of cement produced worldwide in the last thirty minutes. Obviously it’s not just something we don’t need to worry about — radioactive waste needs some special storage procedures (including the extremely interesting question of how exactly to communicate the dangers of nuclear waste to a potential culture 10,000 years in the future) — but it absolutely isn’t an impediment to expansion of nuclear power.
Okay, so nuclear waste isn’t as much of an issue as it seems. But what about accidents? A single plant accident could kill tens of thousands of people with radiation, right? Not really. There’s three events people think of when they think about “nuclear power plant accidents”: Fukushima Daiichi in 2011, Three Mile Island in 1979, and most famously, Chernobyl in 1986. But the role these events have in the popular imagination vastly outweighs the harm they caused.
The Fukushima Daiichi power plant, for example, was hit by a magnitude 9.0 earthquake, followed by a 45-foot tsunami. The reactors shut down as designed, but a failure of the backup generator pumps caused a meltdown and subsequent hydrogen explosions which released a plume of radiation into the atmosphere. The total death toll linked to radiation exposure? A single worker, who died of lung cancer in 2018.
The plume of radioactivity is estimated to have delivered a dose of 12–25 millisieverts (mSv) to members of the population in the evacuation area. (For comparison, the average American absorbs about 6.2 mSv annually from natural sources and medical scans, and the annual exposure limit for power plant workers is about 50 mSv.) The dose of 12–25 mSv is estimated by a WHO report to have caused people within the exclusion zone to go from a 35% lifetime risk of organ cancer to about a 36% risk — and even that is probably too high of an estimate, for reasons I’ll get to in a moment. Meanwhile, the evacuation itself resulted in at least 573 deaths, mostly from stress, interrupted medical care, and loss of support networks. In other words, overreaction to the accident caused more death than the accident itself, by a large margin.
The Three Mile Island incident was similar. On March 28, 1979, due to a chain of failures including poor control panel design, operator error, and a stuck relief valve, a reactor at the Three Mile Island plant in eastern Pennsylvania went into partial meltdown and vented a small amount of radioactive material. This release caused the affected population to receive approximately 10 µSv (about 0.01 mSv) of additional radiation — a negligible amount, considering that the average person receives a 20 µSv background dose every single day. In terms of harm to the population and to plant workers, it was practically a non-incident.
Finally, let’s look at Chernobyl. Chernobyl was a terrible disaster, orders of magnitude worse than both Fukushima and Three Mile Island. The exact mechanics of how and why it happened are kind of complicated, though Wikipedia has a well-written explanation if that’s of interest. As a summary, an extremely long chain of management, operator, and most importantly design failures caused a massive explosion at the Chernobyl nuclear power plant on the morning of April 26, 1986, releasing a large quantity of radioactive material into the environment. The disaster killed 28 plant workers and firefighters in the following weeks due to exposure to massive amounts of radiation (along with two workers who were killed immediately in the blast), and the resulting radioactive plume exposed a large portion of the population of the area to notable quantities of radiation. But estimating the increased mortality that was caused by said radiation is extremely difficult. A 2005 WHO report puts the total expected death toll at about 4,000 deaths from cancer among emergency responders and people living in the affected areas; some reports allege higher counts. But according to a 2008 UN report (p. 64), the only deaths directly caused by the disaster are those 30 plant workers and firefighters plus 15 deaths from thyroid cancer that were linked to contaminated milk. (You might be wondering about other ecological effects and stress to populations across Europe — I’ll get to that in a moment.) So, what explains this discrepancy? And why is the estimated risk from Fukushima likewise probably too high?
The problem is that reports of deaths from low amounts of radiation are based on a model known as Linear No-Threshold. According to LNT, the risk of cancer and cell damage after being exposed to radiation is linearly proportional to the dose of radiation sustained, and there is no minimum safe dose; someone exposed to 5 mSv of radiation would have 10% of the increased cancer risk of someone exposed to 50 mSv. This assumption is controversial, to say the least. Since it’s established that radiation damages DNA, supporters of LNT believe that any single instance of DNA damage has the same small chance of slipping past the body’s built-in defenses against DNA corruption and causing cancerous growth. Critics of LNT contend that the body can repair DNA damage extremely well at a certain rate, and it’s only beyond that rate that the system gets overwhelmed and causes cancer — as such, the rate of dosage matters as much or more than the absolute amount. Empirically, it’s extremely difficult to get good data one way or another.
The lowest dose for which we have clear evidence of increased cancer risks is about 100 mSv, but going below that threshold is tricky because of the massive sample sizes needed to pick up such small increases above pure statistical noise or longer term effects. A nonzero number of people may have died after Chernobyl (and perhaps after Fukushima) due to cancers that they wouldn’t have had otherwise. But estimating the exact number is difficult and might be impossible. The UN report on Chernobyl explicitly refused to provide a civilian death toll, stating that “any radiation risk projections in the low dose area should be considered as extremely uncertain, especially when the projection of numbers of cancer deaths is based on trivial individual exposures to large populations experienced over many years” (p. 66).
That said, even taking the WHO report of 4,000 deaths from Chernobyl at face value, and even counting the 573 evacuation-related deaths in Fukushima as deaths from nuclear power, this results in a total of a few thousand deaths over half a century of nuclear power. And indeed, every single one of those deaths is a tragedy. But nuclear power, compared to equivalent fossil fuels, actually saves lives. According to Our World In Data, for every terawatt-hour produced by nuclear power, about 0.07 fatalities occur. (A single terawatt-hour is enough to power 100,000 American homes for a year.) Meanwhile, a gas power plant will kill 2.82 people, mostly from air pollution; a coal power plant will kill 25. And, obviously, coal and other fossil fuels release greenhouse gases which will kill many more through climate change in the coming decades.
But there’s a bigger problem with the LNT model beyond its potential miscounting of deaths. A LNT-based model of nuclear safety caused untold amounts of trauma and suffering in the wake of the Chernobyl disaster and continues to cause issues with nuclear regulation. For example, the disaster was shown to have no effect on rates of developmental disabilities in children across Europe and in individual countries. Roughly 2,500 mothers in Greece, however, terminated their pregnancies following the accident based on faulty advice from their obstetricians. Indigenous Sámi reindeer herders in Norway, who happened to receive a relatively high portion of the Chernobyl fallout, suddenly found their traditional hunting practices peppered with dosimeters and government scientists (not to mention articles with titles like “Radioactive Reindeer Roam Norway”). Without any of these mitigation measures or disruption, their annual exposure in the most contaminated areas would have been an additional 3 mSv per year, roughly equivalent to that of airline pilots. Similar disruptions occurred to sheep farmers in the United Kingdom until 2012; if these had not occurred, particularly heavy consumers of lamb would have received additional annual doses of about 0.05 mSv. So it’s true that the Chernobyl disaster caused untold amounts of trauma and suffering in Europe, but much of it appears to have stemmed not from the disaster itself but from an overzealous response.
But burdensome measures in the name of safety aren’t limited just to the Chernobyl response, and the response was somewhat understandable considering the lack of knowledge in the initial stages of the accident. However, LNT-based models cause even larger issues when they’re applied to nuclear regulation, an area with no such excuse. As I mentioned previously, the lowest radiation dose clearly linked to cancer is about 100 mSv; probably for this reason, the annual radiation exposure limit for American nuclear power workers is half that at 50 mSv. But according to the LNT model, there is no minimum safe dose of radiation. So, while the maximum dose is 50 mSv, nuclear facilities are expected to keep doses “as low as reasonably achievable.” Which sounds good, right? Except what does “reasonably achievable” mean? Generally, it means that if the cost of nuclear plant construction is similar to other modes of power, the precautions are reasonable. But this means that it is literally impossible for nuclear power to ever compete on price with any other form of power, by definition! Any efficiency advantages nuclear has over any other forms of power will immediately be countered by ever-more-onerous regulation that raises costs not because it is proven to actually make workers any safer, but simply because the agency is required to impose it. And if you’re a power company, why would you ever build a nuclear power plant when you could build an equivalent coal or oil power plant for the same price and much less resistance from the population?
To be clear, regulation is generally a good thing. I’m not suggesting we just throw the rules away completely — after all, that’s what led to the Chernobyl disaster. But some of the regulations placed on the nuclear industry are comical. For example, while transporting a container from a spent fuel storage pool to the reactor at an Idaho nuclear facility, a forklift dripped a small amount of pool water on the blacktop. The water in a fuel storage pool is not radioactive whatsoever. In fact, swimming in one involves receiving less radiation than a normal background dose, because the water blocks naturally occurring cosmic rays. But nevertheless, to keep radiation as low as reasonably achievable, all the asphalt along the forklift’s path was dug up and buried as nuclear waste at taxpayer expense. The area was then repaved with asphalt naturally slightly higher in thorium, making it more radioactive than the material that had been dug up!
So where does this leave us? Nuclear plants are being decommissioned all over the country and across the world. These plants are typically being replaced with coal, oil, or gas power, which emits carbon dioxide, pollutants, and, in the case of coal, releases more radioactivity than an equivalent nuclear plant. (Both of the quantities are tiny, though — perhaps a maximum of 0.02 mSv per year from coal, versus 0.0002 mSv from nuclear. Remember, background radiation exposure is about 6 mSv.) Part of this is due to local political opposition, because not very many people want nuclear plants in their backyards. But most of it is price. Over the past ten years, the cost per megawatt-hour of electricity from new power plants has gotten cheaper for almost every form of electricity, because as time goes on we learn how to build power plants more efficiently. Coal went from $111 to $109. Natural gas went from $83 to $56. Solar panels dropped massively, from $359 to just $40 (which has bigger implications that I’ll get to in a moment). Nuclear, the only exception, increased from $123 to $155. It’s little surprise that no new nuclear plants have been completed in the US since 1996. Regulation and irrational fear have basically doomed the industry.
Considering the power potential, safety, and cleanliness of nuclear power, the answer is theoretically a no-brainer. Make the regulations more sensible, build a bunch of nuclear plants, and enjoy cheap, pollution-free, and incredibly safe electric power. But is it really that simple? Here’s the thing. In general, I try my best to consider things logically. And I’m probably considerably less skeptical of technology than the average person. I don’t have any fear of flying, I’ll happily eat genetically modified food, and when lab-grown meat becomes available I’ll be first in line. And I did a lot of research on nuclear safety for this article. I understand, on a quantitative level, that the risk of cancer from living across the street from a nuclear power plant is probably lower than the risk of getting hit by a car while crossing said street. But would I feel comfortable living there? I really don’t know. The Simpsons, Chernobyl, atomic bombs, Godzilla, news stories about minor leaks, and a million other facets of culture have permanently implanted a fear of Nuclear Anything deep into my lizard brain, not to mention the fundamental cultural psyche of pretty much the entire human race. And if I can’t get over my fear, why should I expect anyone else to? I don’t think it’s possible, and it might not even be desirable, to adopt nuclear power as our solution to climate change and pollution.
Does that mean we’re all doomed from climate change? Not really. As I mentioned above, the price per megawatt-hour of solar power from new installations dropped 89% from 2009 to 2019. Prices for onshore wind power dropped 70%. Those (fully renewable!) power sources are now tied for the cheapest sources of electricity, 25% cheaper than their closest competitor (natural gas) and a third of the price of coal. Thanks to lower cost, renewable energy accounted for 72% of all new electric plants worldwide in 2019. And there’s every reason to think that those prices (especially for solar) will keep falling. In 2050, our future selves might look at our current solar prices and power production the same way we look at computers in 1990, back when they had 16 MHz processors, 1 MB of RAM, and cost $15,000. (For a modern comparison, this is a fully functional computer with a 1000 MHz processor, 512 MB of RAM, and built-in Wi-Fi. It costs $10.) Now, the drop in price doesn’t mean we’re out of the woods, and the response to climate change is perhaps the greatest challenge humanity has ever faced. But it’s hard to overstate how much of a game-changer cheap renewable energy really is. Ten years ago, the solutions to climate change all involved massive cuts to prosperity and living standards, cuts that would never have been accepted by developing countries. But with cheap, clean electric power and further technological innovation, it’s now more than possible to solve climate change and increase everyone’s living standards at the same time. So perhaps it’s time to lay the entire nuclear debate to rest. Even without nuclear power, there is a light at the end of the tunnel — a light powered by clean, renewable, and safe energy.
If you have any further questions, would like to see a column on a specific topic, or think that I got something wrong, feel free to email me at firstname.lastname@example.org.
The opinions expressed in this article are solely the author’s and do not reflect the views of The Phoenix Editorial Board.