Just in case anybody is curious about my sources, I spent twelve years in the Navy as a surface warfare officer on nuclear powered ships. Additionally, I've assembled a synopsis of the events which took place at Chernobyl, just to dispel any rumors of the potential for a repeat in the west. This is very long, and I've broken it into three posts. I'll caution you that this is not the complete story, and I've simplified some very complex interactions. Just take it for what it's worth. Anyway, here we go: You wouldn't use a toothbrush to unstop a toilet, would you? Sure, with some effort you can probably make it happen, but that doesn't mean there's any wisdom to it. In pursuit of this your friends would be quite justified in saying "what the fuck are you doing? Oh never mind. I'll just go get you some hemlock." Strangers couldn't be berated for making the same observation. Even more alarming would be watching you put the toothbrush back in its holder by the sink after an admittedly hard-fought success. Much the same is true about Chernobyl, only in this case the Soviets were trying to brush their teeth with a plunger. It's no surprise that they wound up eating a lot of shit. Chernobyl was a quagmire from the start. The reactor, called an RBMK type reactor, was designed as a quasi-breeder to add some measure of plutonium production in order to supplement the Soviet weapons program. This called for a lot of design elements that would have otherwise not been necessary. Chief among these elements was reliance on a fixed graphite moderator. And now for some reactor theory The moderator does a couple of things for a reactor, but mostly it acts as a speed brake for neutrons. Most neutrons produced in a reactor are traveling too fast to interact with another fissionable nucleus. They’re called “thermal” neutrons because they’re hotter than usual, and on this scale being fast, hotter, and heavier all mean the same thing: more energy. To reclaim these neutrons they need to be slowed down, or thermalized, so that they can cause more fission. In a weapon, the rest of the plutonium or uranium does this naturally because its cross section for collision is very high and the reaction mass is set up for it geometrically. This is not the case in a reactor; they’re laid out to prevent the fuel from thermalizing the neutrons. If they weren’t, the reaction would essentially be in a steady-state critical reaction and controlling it would be very difficult. Instead, the process of slowing the neutrons is left to a moderator. In a reactor, if the moderator were to stop working, the reaction will stop in just a few hundred fission generations since the neutrons being produced are too high energy to interact with the fuel rod arrangement. You can read more about fast neutron moderation here The RBMK design at Chernobyl had a graphite moderator built into the core. There are tubes or shafts of graphite surrounding the fuel rods to moderate the reactor. As with any reactor, careful moderator engineering allows you to control a large number of variables, not the least of which are fuel density, response times, and a bunch of natural circulation properties. RBMK's are a quasi-breeder design that runs off 2.4% enriched uranium and breeds a small amount of plutonium-239. It's not a full breeder like the liquid metal fast breeders, but it still does carry on a small amount of the fast-neutron plutonium reactions. This necessitated that the moderator must return some higher energy neutrons to the fuel, but moderate most of the others. Without getting way too far into how that one works, just take it on faith that graphite moderation was the best option to be breeding plutonium in the RBMK. The graphite moderator gives RBMKs what's called a "positive void coefficient." This means that any air pockets in the cooling water causes an increase in the reaction rate. When the water around the fuel rod is displaced, neutrons are allowed to flow more freely into the moderator (this has to do with some of the neutron absorbing qualities of water), and with a permanent graphite moderator wrapping the fuel rods, there's nothing to buffer the reaction and it continues to grow. Additionally, the fast neutron reaction necessary for the breeder qualities would interject a wide range of neutrons energies that, without the water to absorb them, would be moderated back into the fuel and accelerate the reaction. And more so, when operating at very low power levels the neutron-producing spontaneous fission of the relatively high levels of plutonium in the core can unexpectedly start fission cycles that are many hundreds of generations long, and the presence of a void space can cause dramatic and uncontrollable power spikes. Another peculiarity of reactors that contributed significantly to the disaster is what's called xenon poisoning or xenon smothering. A common by product of U-235 fission is Iodine-135. This iodine, of its own right, is relatively unremarkable, and has a rather short half-life of 6.7 hours, decaying into Xenon-135. Xe-135 is extraordinarily good at absorbing Neutrons; in a reactor it has a cross section of neutron absorption almost a million times that of Uranium. As a result, the presence of a small amount of Xe-135 can smother a reaction by depressing the neutron flux (the Xenon is absorbing the neutrons rather than the Uranium). Once Xe-135 absorbs its neutron it becomes Xe-136 and its cross-section of absorption becomes very much smaller and its interaction with neutrons is no longer of consequence. Xe-135 levels are usually managed by an equilibrium of the reaction rate. At any given power setting Xe-135 is going to be produced at certain rate. Reactor designers are well aware of this, and simply allow enough excess neutrons to be produced that they can be lost to the Xenon without a significant performance impact. But the Xenon smothering becomes very important when changing power settings. Imagine that you had run a reactor at 100% power for several days. The Xenon levels would have long ago established an equilibrium and would have been fairly constant, and every thing is happy. Now say you command a power decrease to 50% power. You drop your control rods in to absorb some neutrons until you get down to 50% in a few minutes. Thinking that everything is happy again, you go on about you business. The next day you notice that the power level has risen to 65% with out having ordered any changes. This is because at the time you started the power decrease the core was saturated with Xenon, and as you lowered the rods the xenon was still being produced at the 100% power level. Remember that the Xenon is a decay product of Iodine, and it takes several hours for Iodine to decay. In the few minutes it took you dropped your rods in, and for the several hours late, the Xenon production was as if the reactor was still at 100%. About 10 hours later the Xenon production level will fall as the Iodine finishes decaying, and it'll take about 40 to 50 hours for it to fully settle out at the new level. Xenon has the effect of masking the real power setting of the reactor until the equilibrium is restored. It acts as a temporary absorber of neutrons, and once any excess is consumed, the power level will change. This becomes critical to events at Chernobyl when considering an ordered increase in power. In the mean time you can bore yourself to tears with Xenon transients here.