Let’s talk about Chernobyl. We enter our time machine, and roll ourselves back to the start of the cold war. We’re nuclear engineers in the Soviet Union charged with getting as many reactors operating as soon as possible. Every bit of enriched Uranium is going to bomb manufacturing, as is all the available heavy water. So, all we have is a bunch of unenriched Uranium and regular water to build our reactors. Hmmm.

Let’s think back to our earlier reactor design talk:

Hey, something nifty! Water is both a good coolant and moderator! No moderator, no chain reaction, right? So, if you use water as your coolant and moderator, your reactor has an intrinsic safety feature. If you lose coolant, you lose moderation and the chain reaction stops. We all live! Thus, almost all nuclear reactors in operation today use water as a coolant and moderator….
But what water, and what fuel? Here’s the trade-off: The better the moderator, the crappier your fuel can be… The rarer versions of fuel (U-235 is better than the common U-238) or moderators (heavy water is better than regular water) tend to be the better. Building a reactor typically requires enriching for either the better moderator or the better fuel.

But, we don’t have enriched Uranium or heavy water. Well, we aren’t required to use water as both our moderator and coolant. What if we stick with water as our coolant, but use something else to moderate the neutrons? Graphite will do the job and it’s cheap and plentiful. So, we’ll build our reactor with unenriched Uranium (or even better, depleted Uranium coming out of the atomic bomb plants) as the fuel, regular water as the coolant and graphite as the moderator. What do we have to lose?

In this design, when the water coolant is lost, the graphite moderator stays and the chain reaction continues. In fact, it speeds up, as the regular water isn’t around to absorb some of the neutrons anymore. Added bonuses? The graphite will tend to chemically explode when it heats up enough. So, lose coolant, the chain reaction speeds up, the reactor quickly overheats and explodes with both chemical and nuclear power as energy sources. Neat. Even better? If the fuel is old, and thus filled with all sorts of highly radioactive waste atoms, the massive explosion will be sure to spread all these atoms far and wide–a sort of gigantic dirty bomb. The resulting mess will result in far far far more radioactivity than dropping an atomic bomb. At least in the bomb, most of the radioactive atoms are consumed to produce the explosive force.

I’d like to imagine the following exchange, between a middle manager in the Soviet Union and us, some plucky nuclear engineers, when planning these plants:

Middle manager: “You have my plant design?”
Us: “Yes, but it is incredibly dangerous!”
MM: “But it will work without any Plutonium, enriched Uranium or heavy water?”
Us: “Yes. In fact, it produces Plutonium as a waste product!”
MM: (Claps hands) “Excellent. We shall have such nice dachas when I tell everyone of this plan.”
Us: “It is far to dangerous to build. I refuse to do it!”
MM: (Laughs. Then pauses.) “Oh. You’re serious.”
MM: (Considers his boss, probably some one-eyed, one-armed veteran of Zhukov’s Berlin campaign in the Great Patriotic War, who won’t be sympathetic to concerns about hoards of irradiated civilians after asking why his reactor isn’t operating yet.)
MM: (Points to us.) “Guards, shoot this man.”
Us: (Shot in the head)
MM: (Turns to our assistant) “So, ready to build the reactors?”
Assistant: “Let’s just pick some places in Ukraine, Romania and other shitholes to build ’em, yes?”

I have no doubt similar disagreements happened in the United States, when faced with similar shortages. American middle managers couldn’t resolve disputes with engineers by the bullet-to-the-head trick, probably tipping the debate outcome.

Chernobyl was this sort of reactor, a reaktor bolshoy moshchnosti kanalniy (aka RBMK, aka insane). The inevitable eventually happened. Coolant was lost. Does it really matter why? Ok, fine. This particular fiasco wasn’t caused by normal operations. They were messing with the plant, trying an “experiment.”

Like most reactors, these must be shut down periodically to swap out the spent fuel rods for fresh ones. So, shutting down the plant, by putting in all the control rods, is pretty typical. During one of these routine shutdowns, the plant operators got an idea. The water pumps, supplying coolant to the reactor, are powered by the reactor’s own turbine. Backup power is available from on-site diesel generators, but these take 40 seconds to power up. The plant operators asked themselves, if we shut off the steam to the reactor’s turbine, will the turbine keep spinning long and fast enough to keep the water pumps operating until the diesel backup gets up to speed?

Great question, guys! I’ll give you a hint to the answer: No. Kaboom!

Why on earth would they think to try something so nuts, to shut off coolant to a reactor that will run out of control upon the loss of coolant? Most were conscripts from coal power plants elsewhere in the empire. They didn’t know better.

What exactly happened? The wikipedia article on the Chernobyl disaster does an admirable job breaking it down clearly. In short, the coolant ran out, the reaction sped even more out of control, melting the reactor and exploding the graphite. Huge amounts of highly radioactive waste was spread all over Europe.

Something to help you sleep at night? Reactors of this design, admittedly modified to be a little safer, are still operating in the former Soviet Union.

When it comes to “disasters” at nuclear power plants, this is only one deserving the title of disaster. Compared to Chernobyl, the Three Mile Island is a pathetic also-ran. In this contest, the US takes a distinct silver to the Soviets’ uncontested gold.

The reactor at Three Mile Island was far more typical, using pressurized light water as both a moderator and coolant, and slightly enriched Uranium as a fuel. What happened? Again, the Wikipedia article on the TMI accident is superb. The short of this one?

Poorly designed controls and displays for a reactor made it virtually impossible for the plant operators to respond properly when a coolant pump failed. There was no reliable way to see how much high-pressure water remained in the reactor, no reliable way to tell if emergency valves were opened or closed, no way to tell when things were running out of control. The operators made poor decisions based on poor information, resulting in the high pressure water eventually all flashing to steam and the reactor core melting. Still, even with all these bad things in a row, only a teeny, tiny amount of radiation escaped into the environment.

Why on earth would such poorly designed reactor controls allowed? Because, brilliantly, almost every reactor in the United States is different from every other one. Thank the lowest-bid culture, in which almost every commercial power plant in the US was guaranteed to be different from any other built before. With such a system, one cannot even train experts to work around shoddy design choices. Each must be discovered individually at a particular plant. Urk!

Not every nuclear power plant system is run this way. The French, in contrast, have almost all of their reactors one of three basic designs. Yes, they have quirks. But at least the quirks are shared among all the reactors of the same type. The net result? The French get about three quarters of their power from these nuclear plants and have had no accident even approaching the severity of Three Mile Island.

The possibility of a nuclear plant accident, however prominent a role it plays in the public imagination, should be a secondary concern to the problem of nuclear waste. Chernobyl, without a doubt, was a fantastic failure. While there will inevitably be future accidents at plants, they are vastly more likely to be more Three Mile Islands rather than Chernobyls.