13
$\begingroup$

I'm trying (failing) to write a story set within the exclusion zone around a nuclear power plant incident set in the near future. The 'incident' that leads to the exclusion zone being set up isn't the story itself, but I want the 'how' to be plausible.

Modern (generation 3 and onwards?) light-water reactors seem to have a whole host of safety features that would prevent a meltdown or accidental release of radioactive particles. Containment buildings, automatic control rod insertion, backup generators and coolant pumps all make a Chernobyl-like exclusion zone coming about seem like an unlikely situation.

A natural disaster could work, but the area I have in mind as the location of the plant is tectonically stable with no history of earthquakes or tidal waves. Plus just copying what happened at Fukushima seems like being cheap. So I'd prefer to avoid going that route.

I was thinking maybe some sort of 'cyber-warfare' attack on the control systems, but I'm not sure how plausible that would be either. A lot of the safety features seem to be independent systems that kick-in with a lack of power.

Any ideas?

$\endgroup$
  • 13
    $\begingroup$ Yay! Another question to get us tagged by the FBI, NSA, and the rest of the alphabet! +1 $\endgroup$ – Frostfyre Nov 4 '15 at 15:01
  • 2
    $\begingroup$ It's worth pointing out that the Chernobyl disaster only occurred because some of the safeties were offline at the time (due to human errors). I'm not enough of an expert in nuclear safety protocol to write an answer based on how badly a human would have to mess up in order to bypass the protocols enough to allow a full on meltdown, but it's a fair bet that someone, somewhere is idiotic enough to bypass something in order to get a job done quicker! $\endgroup$ – Joe Bloggs Nov 4 '15 at 15:07
  • $\begingroup$ Another point of Chernobyl was Western countries considered that particular reactor design unsafe. No Western country would have operated that type of reactor regardless of the safeties and state of repair. Chernobyl literally could not have happened in the West. $\endgroup$ – Jim2B Nov 4 '15 at 17:49
  • 3
    $\begingroup$ Unless it is designed and built by grossly incompetent people, modern reactors are heavily contained, and even a total melt down of the core will result in a molten mass buried under the containment dome. If you want to breach the dome and cause a catastrophe, hit the reactor complex with a tactical nuclear weapon (not much less will do). $\endgroup$ – Thucydides Nov 5 '15 at 0:59
  • 1
    $\begingroup$ @Jim2B The Russians considered the design unsafe as well. It was an early type of reactor where overheating accelerates the reaction instead of decelerating it. Which is why the safeties were designed to absolutely stop any attempt to run it above design power. Enter some people who want to make the reactor produce more power and do not know why the safeties refuse to allow it... RTFM. $\endgroup$ – Ville Niemi Nov 5 '15 at 17:20
13
$\begingroup$

Modern reactors are surprisingly stable. They're deemed virtually unstoppable, which is why the earthquake at Fukushima needed to literally shift the entire plant 100m along the ground before it started to fail. The ability to release fallout is considered so unbelievably utterly horribly bad, that it's nearly impossible to get a plant to do it without doing something grand. Politicians have trouble getting re-elected if nuclear power plants release fallout. They don't like that.

They'd certainly be hardened against cyber attacks, so if you want a cyber attack to be at fault, look at Stuxnet, which is the first known attack to clear an airgap. Even then, there's tremendous incentive to make them fail safe. For example, one medical grade reactor design I am aware of relies on bubbling air through the cooling water to lift the control rods. If anything goes wrong with the electronics, the air stops being circulated, and the control rods naturally settle.

You'd probably want to study a long term attack. While reactors are typically designed to be tamper resistant, nothing is perfectly tamper proof. With enough physical access over a long time, you may be able to disable all of the computerized safeties, jam all of the mechanical ones (many I presume are in high radiation areas, restricted access of course), and then provide the humans false data so that they don't think to hit the big red button when things go wrong. Even then, you probably want to try to time it around some giant event outside the plant to distract people. People at nuclear reactors are funny... its almost like the unconsciously want to protect the plant, and can smell a cooling motor misbehaving from 3 rooms away. Its almost like we hire them for personality traits like that.

Want more information than just vague "maybe" scenarios? Well, the information you ask for would qualify as a vulnerability in our national nuclear power capabilities, so I would expect it to be classified as TOP SECRET//RD, which requires a special "Q" clearance from the Department of Energy to access. Needless to say, if I had access to such a clearance, I would be choosing to not answer even hypothetical questions on a forum regarding my job. (Sometimes it's nice not knowing the real answers! I can hypothesise more that way!)

$\endgroup$
  • 3
    $\begingroup$ Fukushima's biggest weakness was that its backup power supply, required to manage decay heat after a shutdown, was also inundated with water, and as you imply the plant lost its grid connection due to the shifting. $\endgroup$ – KeithS Nov 4 '15 at 15:34
  • $\begingroup$ You don't really mean "literally shift the entire plant 100m", do you? $\endgroup$ – WhatRoughBeast Nov 4 '15 at 17:52
  • 1
    $\begingroup$ @WhatRoughBeast Thank you for the sanity check. I had remembered that it had moved as such, but you just made me re-check my facts and they were wrong. I believe the scene I had been remembering was a region which was indeed moved 100m due to the Tsunami, but it must not have been associated with the power plant. $\endgroup$ – Cort Ammon Nov 4 '15 at 20:21
  • 3
    $\begingroup$ Yeah, maybe I should just go for a vague "post-incident investigation ruled the cause to be sabotage; the details of which were kept secret for obvious reasons". $\endgroup$ – Nick Nov 4 '15 at 20:31
  • 2
    $\begingroup$ Fukushima was a dinosaur age-wise, first commissioned in 1971, and hardly classifies as a modern(ish) plant. $\endgroup$ – MauganRa Apr 9 '17 at 9:00
9
$\begingroup$

Just as one option, there are quite a few RBMK-type nuclear reactors (similar to Chernobyl's) in former Soviet/Warsaw Pact nations that are still operating. Theoretically, a similar sequence of events would do much the same thing there. One potential difficulty is that the final nail in Chernobyl's coffin wasn't under human control; the control rods to SCRAM the reactor for the test got stuck in exactly the wrong position at exactly the wrong time.

Basically, what you have to do is set the reactor up on a positive feedback loop, where an increase of power output causes a change in reactor conditions that drives a further increase in power. Exactly how you do this depends on the exact design of the reactor; most are designed with several layers of "passive safety", where in the event of something out of normal occurring, the design of the reactor itself will cause a negative feedback loop where the reaction mass drops to a subcritical state (no fission chain reactions). For instance, unlike the RBMK design, in many newer designs a loss of coolant renders the fuel subcritical because either the coolant is also the fuel and only goes critical within the specially-designed reactor core (such as in a molten salt reactor), or because the water coolant is also needed as a moderator, to slow the neutrons in order for chain reactions to take place; without the water, as in a loss-of-coolant accident, the neutrons are too fast and leave the core without causing enough chain reactions on their way out to keep the reaction critical.

All that said, once a nuclear plant is built and goes online, there's not much that can be done to upgrade any passive safeties developed after it was designed. Most of the plants in service today are 70s vintage, especially in the U.S. where Three Mile, Chernobyl and Fukushima happened at just the wrong times in the discussion about whether to build more nuclear projects here, and no new fission plants have gone online since '77. It's not talked about much, but despite the extremely conservative policies towards engineering a nuclear plant, things can go wrong.

$\endgroup$
  • $\begingroup$ The remaining RBMK's have been modified and are not allowed to use that type of fuel any more, which have knocked back the positive void coefficient that is a key component to the runaway reaction that happened on April 26, 1986. $\endgroup$ – MichaelK Feb 21 '18 at 8:31
6
$\begingroup$

The swiss plant at Beznau apparently is a fairly optimistic piece of work, and there also appears to be a ridiculously long list of fautls this reactor has.
If I remember correctly, people even drilled holes through the metal containment bubble to mount a fire extinguisher.
You will have to goolge the details, since i can only find german language articles at the moment. And i am pretty sure it is only one of a longer list of sad examples.

What I am trying to get at:
In theory, a nuclear power plant is very safe, with a lot of redundant safety mechanisms.
But in the real world it is one of those 25'000 moving parts, all provided by the lowest bidder things, combined with a lot of politics (or, to put it more bluntly: greed): there will always be weaknesses that should not be there. Provided you find them, you can play with them.
Or, for your story: just drop in a side note that they were there and greatly magnified problems that should otherwise not have been catastrophic.

What a nuclear power plant really, really needs is coolant. And something else it needs, surprisingly, is electricity. So, if you manage to deprive it of cooling water and sabotage the power lines so that they won't be fixed any time soon, i guess you might pretty soon end up with a situation where your overheated fuel starts melting its way out of confinement and into the environment, providing just the situation you were looking for.

$\endgroup$
3
$\begingroup$

Control rods are suspended, controlled and monitored very carefully, but any control system can be subverted if you falsify the inputs and/or override the outputs.

I cannot and will not give details on how to do this, but rest assured that there is no built in simulator that will just let you fake these. And if you tried it on my old submarine you'd have to kill every last one of us to manage it. And you'd have to bring your own power source, because I will personally destroy every piece of generation equipment before I let you do this. In all cases, a SCRAM is simply the removal of power (or other motive force) from the latching device and allowing it to drop. It is trivial to lose the fuses, smash the controls or axe the cables - all engine room personnel know where they are. The last option is to critically damage the unpressurized side of the rod drivers themselves. All I have to do is make sure there isn't enough wire left to do anything with.

In the specific case of slow-fission water moderated reactors, I can also blow all the reactor coolant overboard. It seems counterintuitive until you know that these use water to slow neutrons to the point where fuel will absorb them. The core will probably melt - but you will never achieve criticality, much less the kind of criticality accident this scenario would require.

Sources: My training as a former nuclear operator for the U.S. Navy is specifically designed first and foremost to understand and prevent this specific scenario at all costs.

$\endgroup$
3
$\begingroup$

Is the reactor itself important to your story? Is it important that it was an accident with the operation of the reactor itself? If not, go for spent nuclear fuel instead.

Spent nuclear fuel need to be stored before being moved to final deposition. This storage can be at the nuclear power plants, or at a central repository. In any matter, at one point, sooner or later, it has to be transported. While there can't be a nuclear explosion, there can still be terrorist attacks, or some other grand explosion that breaches the containment of the transport, destroys the canisters the fuel is handled in, and spreads the material.

However, to really create a big exclusion zone, you will want to have some way to spread nuclear material as ash from a fire - this was part of the reason you got so much nuclear material from Chernobyl to spread over such a large area. The initial explosion made a hole, and spread large chunks of radioactive stuff over the surrounding area, and then radioactive ash was spread into the atmosphere from the burning graphite.

$\endgroup$
0
$\begingroup$

Generation 4 reactors (there are currently some experimental Gen 4 reactors around) are designed so that failure of machinery and personnel to enact safety protocols cannot cause a containment vessel breach.

Meaning neither hacking, power failure, nor Homer Simpson could cause the reactor to spew radioactive fallout. It would take active physical intervention (e.g. explosives) to cause Gen 4 reactors to spew fallout.

$\endgroup$
  • 2
    $\begingroup$ And a lot of explosives. The containment dome is quite massive to contain a steam explosion if all the coolant suddenly boiled off and "accidental" things like aircraft crashing into it. Massive "bunker busters" could do the trick, and no one will notice you coming towards the reactor in a heavy bomber.... $\endgroup$ – Thucydides Nov 5 '15 at 1:03
  • $\begingroup$ or driving a couple of tanks overland to penetrate the containment vessel... $\endgroup$ – Jim2B Nov 5 '15 at 1:09
  • $\begingroup$ There are always lots of tanks blocking the drive through lanes, so no one will notice them driving towards the reactor site either...;-) $\endgroup$ – Thucydides Nov 5 '15 at 1:21
0
$\begingroup$

The plant on Three Mile Island, Pennsylvania, had an incident in '79 in which a small amount of radiation was released (still with a USD $1B cleanup), but what made the difference between what happened and a catastrophic meltdown was, basically, luck. And a meltdown would certainly have necessitated a sizeable exclusion zone. There are several still-operating plants built before TMI: the Yankee Clipper plants in Connecticut come to mind. (The TMI plant was only 13 months old at the time.)

Basically, a meltdown happens when the cooling stops (for any reason -- at TMI it was turned off), and results in the radioactive core melting through the containment and sinking until it hits groundwater, lots of which is turned into radioactive high-pressure steam, which could come up over a wide area, depending on geology. There's a discussion of this in the movie The China Syndrome, which was released the same month as the TMI incident.

There's a chapter on just what happened at TMI in Charles Perrow's excellent and very readable book Normal Accidents. I highly recommend reading the (whole) book for anyone trying to construct any believable fictional accident or disaster. He was on the commission that investigated the Challenger Space Shuttle disaster, and knows a thing or two about how disasters happen. And he's a sociologist, so he explains the scientific parts for an audience that doesn't know a megajoule from a kilowatt.

P.S. If you don't have a culprit, you could make the meltdown the work of a huge Atlanta real-estate developer, since it's the only major U.S. city more than 50 miles from any nuclear reactor. (More plausible than Lex Luthor making Arizona oceanfront, anyway.)

$\endgroup$
0
$\begingroup$

Look to the Fukushima disaster as an example of what can go wrong, that was a fairly modern reactor.

In that incident, the four reactors did not fail. They were shut down as a precaution against the effects of the earthquake.

What happened is - even after stopping the nuclear reaction, a great deal of residual heat is still present. Uranium, being a very heavy metal, will retain a great deal of heat, just as brass retains a lot more heat than a light metal like aluminum. Normally, the reactor's cooling pumps are kept running until that residual heat is gone, but that can take several days.

However... the tsunami that followed the earthquake flooded the backup diesel generators that were needed to power those coolant pumps, and the electrical grid that could have provided power from elsewhere was also down.

Without coolant pumps to take away the extreme heat still present in the uranium core, the reactors overheated, boiling off the cooling water that was present. Two of the reactors experienced internal explosions from accumulated hydrogen gas, and there was also a substantial leak of radioactive water. Fukushima could have resulted in a core meltdown, where the uranium melts and falls clear of the neutron absorbing rods that stop the nuclear reaction. If that happens, there's nothing to stop the nuclear reaction from not only restarting but spiraling out of control.

Fortunately, the situation was arrested before full meltdown, but it still made quite a mess.

So, if you want to make a disaster around a modern reactor, create some small fault that would result in a precautionary shutdown, and then fail the backup generators so that the coolant pumps can't run.

$\endgroup$
  • $\begingroup$ Heh, no... Fukushima were not at all " fairly modern [reactors]". They were early Generation II and they had not been modified to take in the lessons in the wake of the Three Mile Island accident. Also you are misdescribing the incident. The accident was not caused by existing heat, but heat caused by residual decay. Just like the embers of a campfire can burn you — not because the wood has not cooled, but because it is still undergoing combustion — the fuel of a reactor continues to produce more heat after it has been shut down as short-lived nuclear waste decays. $\endgroup$ – MichaelK Feb 21 '18 at 8:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.