If you remember only one thing from my work, let it be that nuclear plants following best practices today don’t just wreck half of Eastern Europe on a whim. You gotta get in there and bork some atomic bullshit good and hard for that to happen. And if ever a nuclear plant was thoroughly goddamn borked to fuck and back, it was the Vladimir Ilyich Lenin Nuclear Power Plant, better known to history in the English speaking world simply as ‘Chernobyl.’
Today’s Moment of Science… Chernobyl, Part One: The Accident.
RBMK reactors were developed through the 1960s and commissioned in nuclear plants across the Soviet Union starting in 1973 in Leningrad. The first of four RBMK reactors for the Chernobyl plant was commissioned in 1977, followed by baby’s first meltdown just five years later in 1982. Coolant loss led to fuel overheating and a partial core meltdown, releasing radioactive particles into the chilling Soviet atmosphere. This was suspiciously similar to an accident at the Leningrad plant in 1975.
Studies conducted in 1980 showed that accidents were “likely” in an RBMK reactor even under the most ordinary of conditions. According to KGB documents released by Ukraine in April 2021, Chernobyl was understood to be “one of the most dangerous power plants in the USSR” by 1983. There was another accident involving the third and fourth reactors in 1984. The USSR politely spared us any concern by keeping word of that last incident secret until the middle of a global pandemic.
While they could have done the boring thing and fixed the issues, they chose a more thought-provoking option and made some helpful updates to the user manual. This was the ultimate ‘In-Soviet-Russia-nuclear-solves-you’ solution that would allow any goddamn idiot to safely use a faulty nuclear reactor.
So, the not at all ominous previously failed safety test.
At 1am, April 25th, 1986, plant workers began preparations for a test on reactor #4 to check that, in the event of a power outage, their turbines could produce enough power to pump coolant for a bit over a minute until emergency generators took over.
You know, to prevent a meltdown.
They slowly lowered the reactor’s power level, but then got a call asking “more power plz, the tiny comrades are home from school and Tetris doesn’t just power itself.” Which maybe would have been okay, but this pushed responsibility to the overnight shift workers who weren’t trained to conduct this test.
Which is where we should have a quick discussion about the RBMK’s… quirks.
Nuclear reactors are, and I say this with the benefit of not having to design one myself, relatively simple. A neutron and a uranium atom fall in love and bang, leaving waste products, an imperial assload of energy, and more super pissed off neutrons as evidence of their brief volatile affair. Certain conditions have to be met to continue this chain reaction. A moderator is used to slow down neutrons enough to react with their target atoms. Couple that with a neutron absorber to keep too many collissions from making things more ‘splodey than you’d like, toss in a coolant and a turbine and you’re practically a nuclear engineer just from reading this.
The RBMK used graphite for a moderator surrounding the channels. Light water (known to us plebes outside the nuclear community as… water) was used as a coolant, but its capacity both as a moderator and a neutron absorber needed to be taken into account as well. Boron carbide control rods absorbed neutrons and were inserted into the massive apparatus from above and below to soothe all the reactor’s sore spots.
When the overnight shift got to work, they reduced power as planned. However, it plummeted below target levels seemingly without warning. Xenon-135, a fission waste product and annoyingly strong neutron absorber, built up in the reactor when it was on low power and poisoned it to a stop. It was going to take time to get the reactor back online safely.
But there was a test to be done, so.
They quickly pulled out nearly all the control rods, and finally the power was… still not quite where they needed it for the test but, eh, fuck it. With nothing holding the reactor back, power would go up. Somehow.
To finally start the test at 1:23 am on April 26th, the water pumps were powered by a wing, a prayer, and a turbine that was running on fumes until emergency generators kicked in. As the turbine’s inertia began to fail it, so too did the momentum behind the remaining water that cooled the reactor. Voids in the water from the steam didn’t absorb neutrons to the degree that liquid water did. It’s a nuclear grade feedback loop; fuel heats the water causing steam, leading to an excess of reactive neutrons causing more reactions, causing more heat and more steam. Lather, rinse, duck and cover.
At 1:23:40, the AZ-5 button was engaged. It’s also called ‘SCRAM’ or the ‘stop this fucking nonsense right this instant’ button. It should have inserted all the control rods, stopping the reaction in its tracks.
At 1:23:45, Reactor #4 tore a hole in the sky.
This has been part one of this three part MOS on Chernobyl, reminding you that I do not have HBO’s budget.
To get the Moment of Science delivered to your inbox every time I declare the fact checking to be complete, head to patreon.com/scibabe.
Also important — this little design oddity about what parts of the control rods were made of and how things went as they were inserted; first they sped it up, and a few inches or millimeters later they exercised some control.
https://www.google.com/search?client=firefox-b-1-d&q=positive+void+coefficient+rbmk
PS, while looking for a cite on the “tip effect” control rod problem — as I recall the tips were something that did not absorb or slow neutrons and that virtually transparent section had to pass through the reactor followed by the rest of the control rod which was made of something more appropriate.
But while looking that I stumbled across this
https://allinnet.info/history/the-dark-accident/amp/