Dr. Zoomie, what with all the zombie/plague/end of the world movies and shows I started wondering what happens to the world’s nuclear reactors in an apocalypse. Not a massive one that destroys the Earth, of course – more the smaller ones that might strike. Would the reactors continue cheerily running along, providing energy to survivors and helping them to rebuild civilization? Or would they all eventually melt down, poisoning the planet? Hmmm…might make for yet another movie!
Hmmm…I guess it depends on the pace of the apocalypse. A fast-moving event (i.e. nuclear war or a smallish asteroid impact) would give a lot less time to prepare than a more leisurely event (plague or zombies). I guess an alien invasion might fall somewhere between these two extremes, depending on the aliens’ attention to detail and patience. But we’ll let that wait a bit….
Let’s start with the difference in…let’s call it the pace of an apocalypse. The exact amount of time isn’t the issue – the issue is how much time the nuclear reactor owners and operators have to prepare for the loss of the power grid, because that’s the limiting factor in what happens to the reactor and how bad things might get. The reason for this is that, after a reactor shuts down – and in an apocalypse, they’re all going to shut down eventually – the radioactive fission products continue to pump heat into the reactor fuel and that heat has got to be removed or the fuel is going to melt down. The way that most reactors do that is by pumping water through the core to remove the heat from the fuel, then through a heat exchanger to remove that heat from the water. Then the cool water flows back through the core to start the cycle again. In some newer reactors, this can happen without coolant pumps, using natural circulation (hot water rises into a heat exchanger and the cooler water sinks back down into the core), but even these reactors need to keep the heat sink filled with cooler water. So having a source of power for the coolant pumps – or for pumps to keep the heat sink filled with cool water – is a big deal.
Let’s start with what’s going to happen when the power grid goes down and the reactors lose off-site power. Emergency generators will automatically start up to make sure that the coolant pumps can continue to shove water through the reactor cores and, in many cases, the plant will automatically scram (an emergency shutdown) or the operators will manually shut it down by inserting control rods into the reactor core. There are a number of automatic scrams – if there’s a loss of reactor coolant, if there aren’t enough coolant pumps running to keep the core cooled at the current power, if power is lost to the control rod drive mechanisms, and so forth. Anytime reactor conditions trigger an automatic scram the control rods are inserted into the core, absorbing neutrons, and brining the nuclear chain reactor to a halt. And even in the absence of a scram, the reactor operator can shut down the reactor more deliberately, by slowly inserting the control rods until the reactor is no longer critical. If necessary, the operator can then go on to cool down the reactor plant – in a day or so they can get the temperature from 500 degrees or so all the way down to room temperature. But, with the radioactive fission products still decaying away, without cooling it won’t stay at room temperature for long. Anyways – back to our apocalypses!
A “slow” apocalypse (e.g. plague, zombies, attack by easily distracted aliens) is one that unfolds over several weeks, months…maybe even years. With enough time, the plant owners and operators can build extra fuel storage capacity to keep their emergency diesel generators running for months or years before the fuel runs out. Maybe they can install an additional generator or two as well, to prepare for long-term mechanical failure. With this much time, they might even be able to refuel from time to time, perhaps extending their emergency power supply for years to decades. The longer they can keep cooling water flowing through the reactors, the longer they can avert a meltdown; if they can keep the core cooled long enough (a decade or two), they might avert a meltdown altogether. Advanced reactors that have a passive cooling system, “powered” by natural convection will last much longer since there are no moving parts and no need for electrical power. But even here, over the years the cooling pond can silt up, pipes can become clogged or can leak, heat exchange surfaces can become fouled, and more – without regular maintenance, eventually even the emergency cooling will cease to operate.
No matter what the cause, the more time we have to prepare for the eventual loss of the power grid, the better the odds of the reactors not melting down. We’ll get to the “what happens when flow stops and the core melts down” part in a few paragraphs – for now, let’s see what else is going on.
So now let’s see what happens if the apocalypse unfolds over, say, days to weeks – enough time to make some plans, but not enough time to set the reactors up for years of unattended operation. As to what might cause this sort of time scale…maybe it’s an inept alien conquest? In any event, I’m thinking of a case in which there’s time to shut down and cool down the reactor deliberately, to put it into the safest condition possible, but not to make the long-term arrangements I mentioned earlier. This is where reactor design makes a huge difference – older designs that rely on coolant pumps to keep the core cool are going to start having problems a few hours after they lose electrical power, which will likely be within a few days to a few weeks (depending on their fuel stores). Once they lose cooling the reactor will do its best to keep the core covered, but eventually – likely within several days – the fuel will heat up and melt (again, we’ll get to that shortly).
And finally there’s the apocalypse for those with short attention spans. Maybe it’s an asteroid that comes out of nowhere and gives us a wallop like what wiped out the dinosaurs, maybe a sudden nuclear war, an interfaith Rapture of biblical proportions…whatever it might be, it happens quickly. This is where the reactor safety systems are going to be important because we’ll be counting on them to scram the reactors, to make sure the coolant pumps are running, and to keep them operating for as long as there’s power on the grid and as long after that as the site emergency power keeps going. I honestly don’t know how long the electrical power grid will keep operating without any human intervention – different areas have greater or lesser degrees of automation, some are newer or older, some are better-maintained than others, some might just get unlucky. But when the first accident happens that affects the grid, the grid is stressed and the next hit happens a bit sooner; each subsequent problem stresses the system more and the brings the next problem even closer. In 2003 something as simple as a tree branch banging against electrical lines was enough to start a cascading series of electrical failures that left 50 million Americans and Canadians without power for a few days. That was a tree branch – picture the potential for cascading failures in the aftermath of an asteroid impact or a nuclear war.
The bottom line is that, even if every reactor scrams and every emergency cooling system comes on line, every emergency generator starts up as planned…at some point the power grid will grind to a halt, the emergency fuel will run out, and we’re back to the core heating up and melting down – so let’s see what happens then.
As I mentioned earlier, the fission of a uranium atom creates two radioactive fission products; when these decay they deposit energy in the reactor fuel, causing it to heat up. A reactor that was running at, say, 1000 MW of thermal energy (which is about 300-400 MW of electrical energy) will be producing several tens of MW of thermal energy from fission product decay; this is the heat that needs to be removed to keep the fuel from heating up. When the coolant pumps lose power and stop operating, all of that energy goes into heating up the reactor core. When the fuel temperature reaches about 2000° F the zirconium used to clad the fuel will begin to react with the water, breaking it down into hydrogen and oxygen. Much of the hydrogen reacts with the zirconium, releasing still more energy and some of the hydrogen will accumulate in the reactor. As we saw at Fukushima, if some of the hydrogen leaks out of the reactor system it can collect in the support building or inside the containment structure, if exposed to a stray spark it will explode. And as the temperature continuous to increase the zirconium cladding will begin to melt, along with the uranium fuel pellets contained within.
Luckily, as this happens the fuel should remain subcritical, keeping the nuclear chain reaction from rekindling. The melted fuel will collect at the bottom of the reactor vessel where it might lose enough heat to solidify, or where it might continue to melt its way through to come to a rest on the floor of the containment. Either way, it will be in a more compact form than in its original configuration, making it harder to release the heat within; as with, say, an ice cube melting on the kitchen counter the molten fuel will continue to spread out, making it easier to release the heat within. At the same time, the shortest-lived nuclides are decaying to longer-lived ones and the amount of heat production is going down. At some point the fuel will be radiating as much heat as is produced and temperature will stabilize and then will begin to drop as the rate of heat production continues to drop.
There’s more than just melting fuel of course – some of the radioactive fission products are gases or the vapors of volatile elements. Isotopes of xenon and krypton, both chemically non-reactive noble gases, are produced through fission and are released when the fuel melts, as are isotopes of volatile cesium and iodine. If the reactor plant’s pressure boundaries (the pipes and reactor vessel) are breached all of these can be released into the environment; they can also be released if a pressure relief valve opens and begins to vent steam and gas into the containment building (this happened at Three Mile Island), especially if the containment was damaged in whatever apocalypse set this whole sequence of events in motion.
As to the final part – how all of this will affect the environment – the most likely answer is “not as much as you might think.” This, I base on what we’ve seen around the Chernobyl site in the aftermath of the worst reactor accident we’ve yet seen. The ecosystem around Chernobyl was heavily contaminated and, in the immediate aftermath of the accident, there were some places (e.g. the “Red Forest”) that were badly affected by the radioactivity released. But over the decades the ecosystem has recovered – partly because of the decay of the radioactivity and partly because people no longer live in the area, allowing much of it to return to its original state of nature.
I know that Chernobyl was only one reactor, compared to the 400+ power reactors globally; the thing is, most of the world’s reactors have fairly robust containment structures that would do a good job of keeping the radioactivity from reaching the environment (or at least delaying its release until much of the initial radioactivity decays to longer-lived nuclides). And if you’re wondering about the effectiveness of a containment structure, it’s worth considering that the Fukushima nuclear reactor accident involved three operating reactors that partially melted down, yet released only a fraction of the radioactivity of the single reactor at Chernobyl (which did not have a containment). When I was in Japan shortly after the Fukushima accident, the radiation dose rates I measured were clearly elevated 20 km from the site and in the middle of the heaviest deposition – but it was still safe. Similarly, the meltdown at the Three Mile Island reactor plant released very little radioactivity to the environment, and the majority of that was deliberately released when hydrogen was being vented from the containment to prevent an explosion.
The other thing to remember is that the Earth has a surface area of close to 200 million square miles, about 60 million of which are land. Chernobyl contaminated about 80,000 square miles – it would take 750 Chernobyl-sized releases of radioactivity to contaminate the land on Earth to levels produced by Chernobyl. This suggests to me that, even if every reactor on Earth ended up melting down, life on Earth as a whole – whatever was left after the plague/zombies/aliens/whatever finish their dastardly deeds – will survive just fine.