Uday Mehta

Engineer | Writer | Podcaster

What is Chernobyl and How Will It Kill Me?

Editor's Note: Hey, it's Uday, I'm also the editor because I don't have the money to hire an editor.  I recently wrote a piece "Why is Nuclear Power Not in the Green New Deal?" which presents nuclear power as an economically and environmentally viable solution to the energy crisis and climate change. To help people understand the subtleties of nuclear power, I'm launching a set of companion pieces in a series I'm calling "How Will It Kill Me?" Previously, we talked about radiation and nuclear waste. Today, we talk about accidents.


HBO's miniseries Chernobyl is a masterpiece. Beautifully shot and wonderfully acted, it shows one of the darkest moments in modern human history, when a new dangerous force had been unleashed upon the Earth for the first time. It's didn't even have to be substantially dramatized or modified because the actual story is just that gripping and unpredictable – most of what is depicted in Chernobyl is legit. 

That's not to say that everything was real – people bleeding from radiation poisoning, a blue skybeam emanating from the core, and the English accents probably aren't up to snuff, but who cares. The failing robots, animal genocide, and ejected graphite are all spot-on. They even mention things like positive void coefficients, concepts that would be overused sci-fi jargon in a superhero TV show but are quite relevant in this context.

Regrettably, the series has one glaring flaw. Though it has no obligation to live up to any such standard, it would have appropriately contextualized this accident within the wider discussion of safety around nuclear power. Things like this have an impact – if you watch this show, you're likely to come away with the (inaccurate) notion that nuclear power isn't safe. The events at Chernobyl were not caused by something inherent to nuclear power, but faulty engineering practices, buried information, and improper oversight – a point that can get lost among the stunning and horrifying visuals.

When considering nuclear power going forward, the discussion around safety is likely to be its biggest hindrance. The fact that you're using nuclear power does not make it any more likely that an accident will happen, or mean that such an explosion is inevitable. So let's understand why this is the case. As Professor Legasov says in the opener, "We are dealing with something that has never occurred on this planet before."

A quick caveat – I've tried to make this as science-lite as possible, but yes, there will be science here. Don't just gloss over these points, because if I included it, it's something you need to know.


What makes a nuclear power plant dangerous?

It's the presence of nuclear material, specifically material that is radioactive. Usually, this material is kept inside several enclosures to make sure it can't escape.  If there is no barrier between the nuclear material and human beings, radioactive gases and particulate could be carried by the wind, whereupon people could ingest or inhale them. (I have 3000 words on radiation for you if you need a refresher). On this note, there are a couple of mischaracterizations in the show that we should clarify.

Radiation poisoning is not contagious, because that's not how radiation works. Remember that radiation is just energy. If you get a dose, it means that energy has harmed cells in your body, and you're experiencing health problems because of the damaged cells. The only way you can 'pass' radiation from one person to another is if there is actually radioactive material (particulate) stuck on someone's clothes or skin, and it gets stuck to someone else. Think of it like a burn wound. You get burned by being close to an intense source of heat. Once you're burned, you can't pass that burn on to someone else just by physical contact.

While radiation can be damaging to an unborn baby, the relatively low doses received at Chernobyl weren't enough to affect fertility, stillbirths, or otherwise affect the health of the child. The fear of having a sick child did lead to a massive wave of abortions by women in that region.

So it's the same effect as a nuclear bomb?

No. NO! This is one of the few unfortunate dramatizations that the show indulged in. There was never actually a serious threat of a 4-megaton blast because of… melted fuel being dumped into radioactive water. There was a risk, but the actual scale was something like a couple of Tomahawk cruise missiles.

Bombs and reactors are fundamentally similar in the sense that you're harnessing the same scientific principles, but fundamentally different in the sense that the goal is different. With reactors, you're actively trying to control the reaction, so you don't have that much nuclear material (fuel) to begin with. With bombs, you're intentionally starting a chain reaction that you want to make as big as possible, which is why you pack a lot of material into a small area and let the reaction run wild.


The percentage of the fuel that is made up of the isotope Uranium-235 is called the enrichment. For example, if I have two atoms of U-235 out of ten total atoms, my fuel is 20% enriched. Reactor fuel is typically less than 20% U-235, and Chernobyl used fuel enriched to only 2%. Weapons need an enrichment upwards of 80%, where the Fat Man bomb was at 96% Plutonium. So even if someone were to steal fuel from a reactor, they'd have to have an enrichment plant set up, which would include highly sophisticated machinery and a ton of energy. So no matter how bad it is, a reactor accident is *nowhere* close to a bomb.


Still scary. So what are the chances?

There's something called the International Nuclear and Radiological Event Scale (INES), kind of like the Richter earthquake scale, but for nuclear stuff. It goes from 1-7, where anything above 5 means there was some dose to the public. There have been seven total accidents that got a 5 or greater on the scale.

All of this is in 17,000+ reactor-years of operation. That is, if you added up all the times that every nuclear reactor in history has been in operation, it would be over 17,000 years, and there were only a handful of major accidents. Only two of these – Fukushima (the only 5+ event since 1987) and Chernobyl – ended up having significant dose to the public.

Let's be clear – this is not to dismiss the seriousness of these accidents. We're not fucking around here. Rather, the point is to highlight that risk is a product of frequency and consequence. Yes, the consequences of a major nuclear accident can be high. But the frequency is so low that it makes the overall risk negligible. It's like flying on an airplane – if something goes significantly wrong, it probably means that most people on board are dead. And the fact that plane crashes are still news (remember the two Boeing 737 Max jets from the past year?) means that they're far from normal. Think of how many flights there are every single hour of every day. It's rather remarkable that plane crashes don't happen *more* often than they actually do. That's a testament to just how safe the technology is.


Why do big accidents happen?

A few reasons. You can have a meltdown, which is when the temperature just gets too hot. This can be due to a loss of coolant, decay heat, or criticality.


Maintaining a nuclear reaction centers around the idea of the neutron population. Remember, the idea of the chain reaction depends on the fact that when you shoot a neutron at an atom of U-235 (fuel), it will split apart (fission) and release 2.5 neutrons on average, each of which can then split more Fuel atoms. Criticality refers to how the neutron population is changing over time. If you were to freeze the system at one point in time and count all the neutrons, and then freeze the system at some later point in time and count the neutrons, you could see whether it was increasing (supercritical), decreasing (subcritical), or staying the same (critical).

So if you started a reaction, it would intuitively be supercritical. You put in one neutron, you get 2.5 out. In reality, you can lose neutrons due to leakage (they escape the system). These neutrons also need to be travelling at a certain speed. When fission shoots out neutrons, they're travelling very fast, and a moderator material is used to slow them down. Some neutrons might get slowed down too much, so you lose those neutrons too.

Of course, you want to be able to control the criticality very precisely. After you get to your desired power output, you want to remain exactly critical, you need to find a way to remove 1.5 of the 2.5 neutrons you get from each fission, so you have one neutron in, one neutron out. This is done through the use of control rods, which have materials that will absorb the neutrons upon impact instead of reflecting (scattering) them back into the system. Inserting the rods into the core will take neutrons out of the system and decrease the criticality, and raising them out will keep more neutrons in the system and increase the criticality. In an emergency, you can drop all of the control rods all the way into the core to stop the reaction completely.

Decay Heat

When Uranium fissions, it actually splits into… stuff. This stuff can also decay and release heat (decay heat) and neutrons. So even after the reaction stops, you can have heat being generated from the stuff that Uranium splits into.

Recap: When you're generating power, you want to keep things critical, which means that the neutron population stays the same over time. The moderator slows down neutrons to make sure they're at the right speed to induce fission. The control rods are used to guide the speed of the reaction, and hence the power output. They're also the fail-safe in case there's a power surge.


Why did the big accidents that happened happen?

Spoiler alert! Well, not really, because it's… history. So it's really your fault if you consider this a spoiler. Here's the tl;dr: there were some flaws in the reactor's design from the beginning, the people involved hadn't been properly trained, crucial information had been ignored, certain safety features were purposefully disabled, and most of the systems had human decision points instead of automated ones.

No, I want to know! Tell me about Chernobyl.

The final episode of Chernobyl actually gives you a good recap (skip to Legasov's testimony at around 28 minutes in), which is actually more technical than the recap I'll give you now. If you're feeling a little science sickness, skip the next couple of sections.

The operators recognized there was a big safety flaw in the reactor. In the case of an emergency, there were cooling pumps tied to emergency power, but they would have taken a full minute to get to maximum power. So if there was a power loss, you wouldn't be able to cool the reactor for a minute, during which core meltdown could occur. The reactor had (somehow) been running for two whole years without this critical feature! The operators wanted to see whether they could compensate for that lack of power in another way, so they decided to run a test. From there, a bunch of stuff went wrong. Here are the most important tidbits to consider:

  • Personnel: The test got pushed back, so the night shift (who were unfamiliar with the actual procedures for the test) was on duty.

  • Buildup of Poison: Sometimes, fission produces neutron poisons, which suck neutrons out of the system and slow the reaction down. If the reactor is operating at a high enough power, these will get burned off; if not, the poison will build up.

  • Coolant Flow: The coolant was moving faster as a part of the test, meaning that it spent less time in the cooling tower, so it was hotter than normal when it came back to the core, and couldn't do as good of a job at cooling.

  • Voids: Let's say that the coolant (water) boiled and turn into steam. These steam bubbles are called 'voids', which are effectively empty spaces where there should be coolant. On top of being a coolant, water likes to absorb neutrons. With more neutrons in the system, more fissions can occur more quickly, and this faster reaction is going to produce even more heat, which will boil off more coolant, create more voids, and so on.

  • Just the Tip: Regardless of all the mistakes that can possibly be made, there's always a fail-safe – the SCRAM, when all the control rods drop down into the core as fast as possible to bring the reaction to a halt. But unknown to the operators, the tips of the control rods were made of Graphite which is a *moderator*. Remember, a moderator is what slows down neutrons to make sure they're the perfect speed to cause more fission. So the thing that is designed to completely stop a reaction has a tip that does exactly the opposite (which was because it was cheaper to manufacture them that way). So when all the control rods were inserted, there was complete moderation, and the power spiked. Even using the mechanism specifically to control an uncontrollable reaction, it was impossible to control the reaction.

Here's what happened: operation at low power → buildup of poison → slow reaction → withdrawal of control rods to speed up reaction → increased coolant flow → hotter coolant → boiling temperatures → coolant burns off → creation of voids → less cooling & less absorption → more neutrons in the system → a faster reaction → more coolant burns off → even less cooling → exponentially faster reaction → control rods are dropped in to stop the reaction → graphite tips increase the reaction even more → immense steam buildup → steam explosion → exposure of reactor core to the outside.

This isn't just ancient history. What happened at Fukushima?

Though Fukushima happened in 2011, it was also old technology, having been built in the late 1960s. It's also worth noting that there were established safety concerns dating decades back – from a feature that was inexplicably changed during construction, the flooding of a backup generator in 1991, multiple ignored tsunami studies in 2000 and 2008, and warnings over the susceptibility to earthquakes in 2011 itself.

The events here are much more straightforward than what happened at Chernobyl. There was an earthquake, upon which the reactors did automatically shut down. Of course, there was still some decay heat, which required consistent cooling (in essence, the same theoretical situation that the Chernobyl operators were trying to test). Crucially, the plant primarily ran on the power it generated itself, so upon the shutdown, it had to resort to emergency sources. The subsequent tsunami breached the seawall and flooded and disabled the plant's emergency generators, in addition to the switching station which would have switched power to additional backup generators located offsite. The overheating caused a reaction that produced hydrogen, which then ignited (after unsuccessful venting efforts). So… boom again.

Wasn't there something in the States too?

Yup. The worst nuclear accident in United States history (a 5 on the INES scale) is actually a testament to how good the technology was. The situation was a result of a part failure that led to a loss of coolant, which was misinterpreted by a reactor operator who then manually overrode the emergency cooling system. So it would have been fine if there was no active human interference. Despite the incompetence, the radioactive release was confined within the facility. The nearby public only got a radiation dose that was equal to half a routine x-ray at the doctor's office, all because the containment systems worked as intended. Yet, concerns over this accident led to a slowdown in the historic growth of nuclear power in the States, setting back the industry by at least a decade.


Won't this inevitably happen again?

No! Not at all. And that's the point. Not only was it a perfect storm of stupidity that allowed something like this to happen, but it's also a sign of the times that it was physically *possible* for it to happen.

It bears repeating that the test that was being run at Chernobyl was to see if they could overcome a safety design flaw that had existed for *two years*. And the operators there were able to (at will) disable certain features just because they wanted to. Things like that are unimaginable given today's regulatory environment. Now, instead of active systems, we rely on passive systems , which don't require human input to work and activate by themselves. And, there's no more override switch either. For example, consider the Positive Void Coefficient, the thing that caused the cycle of temperature increase and power increase. Modern reactors have a Negative Void Coefficient, meaning that if temperature increases, the power decreases. This is done by using the same material for the coolant and the moderator, so if your coolant burns off (which would speed up the reaction), your moderator would also burn off (which would stop the reaction).

While time doesn't necessarily heal all wounds, it does mean something that Chernobyl (aka the worst nuclear accident in history) happened over thirty years ago. It's the only reactor accident where people died from radiation exposure. Two employees died at Fukushima, but from the earthquake, not radiation or anything nuclear-related.

Remember how we talked about the parallels to flying, and accidents where people die from a plane crash? If you heard about a plane crashing, that might freak you out for a bit, but you'd inevitably take a plane the next time you needed to. Now imagine if someone told you that 50 planes crashed… back in 1987. Oh, and 10 planes crashed in 2011 too. Those numbers would probably be meaningless to you. After all, the technology, standards, training, regulation, knowledge, and overall sophistication within the airline industry have all surely increased since then, right?

But… aren't other power sources safer?

Quite the opposite. The number of deaths due to nuclear accidents since Chernobyl is less than sixty… combined. Deaths of coal miners are in the double digits almost every year. Things like oil spills are so commonplace that last year's Sanchi tanker collision in the East China Sea (the worst spill in terms of volume since 2010's infamous Deepwater Horizon incident in the Gulf of Mexico) went largely unnoticed.

Let me put this another way: to produce the same amount of power, a wind farm would take up more than 500 times the space a nuclear power plant does. In a wind farm, all that land becomes unusable. So even in the worst case of a catastrophic nuclear accident, the potential danger zone would take up the same amount of space that a wind farm would from the very beginning.


If nuclear is like flying on a plane, other power generation methods are like driving cars. The last fatal commercial airline crash was in 2009. In 2017, car accidents resulted in over 100 deaths *per day* nationwide. Per day! but you don't think twice about getting in your car and driving somewhere, because we've been desensitized to how frequently something can go wrong. Plane crashes, due to their horrifying visuals and larger symbolic meaning (see: 9/11) capture our imagination, exactly like Chernobyl did. Just because it looks scary doesn't mean that we have to (or should be) be scared.


Top ten list, please. What are the takeaways?

This isn't Buzzfeed, asshole. But if you insist…

1. Reactors CAN'T be weaponized into bombs, even by purposeful sabotage.

2. Radiation poisoning is NOT contagious.

3. Only Chernobyl and Fukushima have had a significant dose to the public.

4. Chernobyl was a due to human incompetence coupled with a covered-up engineering flaw.

5. Fukushima was because of willful ignorance towards warnings and a unique combination of natural disasters.

6. Three Mile Island resulted in a dose the equivalent of getting half an x-ray at the doctor.

7. The human component has largely been removed from modern reactor technologies.

8. Nowadays, there are redundant, independent, and diverse systems to ensure a defense-in-depth approach to safety.

9. Nuclear is (by far) the safest power industry, in terms of both accidents and fatalities… by a mile.

10. In over 17,000 combined years of operation, there have been fewer than a handful of public exposures to radiation.

Class dismissed.

*Note: The primary purpose of this piece is to communicate qualitative themes about nuclear power, and hence intentionally does not cite the sources for the statistics presented. Readers are encouraged to research these important topics, particularly in peer-reviewed scientific journals, for confirmation.

World-class journalism in your inbox.

* required