Saturday, 25 July 2015

Myths of SF: Do Nuclear Reactors Explode?

Fission, Fusion, Reactors, and Bombs

   Fear of nuclear devices is deeply engrained in modern culture as a result of the Cold War; years of imminent nuclear annihilation has that effect.  SF often reflects this, with many works from the Cold War period or later revolving around nuclear war as the cause of an apocalypse.  Even though that has been replaced in new works by climate disruption, genetic mishaps, etc. that reflect the newest scientific advances the negative connotation of nuclear devices remains.

   The most common example of this is the use of Fission bombs to show a person or faction as being 'uncivilised' or to show how desperate the situation.  In the Dune 'Verse all the noble families posses Atomics the use of which is seen as unthinkable, while in the movie Oblivion the use of nuclear weapons against the invading aliens was used to indicate the desperation of Earth's defenders.  The second example, more prevalent in movies at least as far as my own experience goes, is that of a reactor exploding.  As well as tapping into people's fear of radiation and their knowledge of the destructive potential of nuclear weapons it is an easy way to add tension to a story.  Alien did this, as is Aliens a classic example with the damaged Atmosphere Processor, and is the B-movie Nuclear Hurricane, although in the latter example the reactor did not explode its use as a literary device is the same.

   For a really fanatical 'hard SF' fan that is a problem.  Nuclear reactors don't explode.  Or more precisely - a nuclear reactor will not undergo a nuclear detonation, producing the feared mushroom cloud.  They can still explode, but this is due to high pressure steam or chemical reactions, and while it may severely damage the reactor facilities and spread radiation, it will not level everything within a kilometres wide blast zone.

   It is also important to understand that in a modern nuclear 'physics package' it is actually quite hard to achieve nuclear detonation.  The explosive compression of the fissile core requires incredible precise timing to achieve the required densities.  While gun-type devices are less precise they are still a mechanism dedicated to achieving a fission explosion.  Thus it is highly improbable that any nuclear reactor is eve able to achieve the conditions for a fission detonation; the required conditions are too precise.

   Yet an immanent catastrophe is the perfect way to spice up an otherwise lagging plot, are to up the stakes just that little bit more, so what can the hard SF writer do?  Firstly, arrange the setting so that a relatively small explosion is catastrophic - "if the reactor goes down the plasma shields fail and the solar flare will kill us all!"  Specify a non-nuclear explosion - fusion reactors could be quite cooperative in this regard, as I will explain later - to avoid the critics, and way you go.  For more specialised situations there is a possibility of nuclear detonation, most revolving around spacecraft due to the inherent danger of a system that can suffer catastrophic failure.

   I'm going to look at the first and second options.  For the non-nuclear explosion a quick look at the Chernobyl and Fukushima disasters will outline the basic failure modes.  Then for more futuristic settings will be a look a Fusion, Antimatter, Black Holes, NSWR, NPP, and more.

Chernobyl after the disaster
Chernobyl & Fukushima

   The key to both Fukushima and Chernobyl hinges, to my (arguable limited) understanding on the fact that a fission reactor cannot be stopped instantly as, for example a car engine can.  A percentage of the power that the fuel outputs is not from the primary reaction but from the decay of short lived isotopes produced in the reactor.  This makes shutting down a fission reactor a tricky matter under the best conditions, as full cooling must be maintained uninterrupted through the process.  In Fukushima the failure of backup diesel genitors compromised the cooling system, and when the backup batteries run out it lead to a meltdown as the containment vessel overheated.  According to the wikipedia page

   "It is estimated that the hot zirconium fuel cladding-water reaction in Reactors 1-3 produced 800 kilograms (1,800 lb) to 1,000 kilograms (2,200 lb) of hydrogen gas each, which was vented out of the reactor pressure vessel and mixed with the ambient air. The gas eventually reached explosive concentration limits in Units 1 and 3. Either through piping connections between Units 3 and 4 or from the zirconium reaction in Unit 4 itself,[27] Unit 4 also filled with hydrogen. Explosions occurred in the upper secondary containment building in all three reactors.[28]"

   A similar situation occurred at Chernobyl.  Although it seems that the problem there was more due to the rapid boiling of the coolant water.  From the appropriate wikipedia page

  "Because of the positive void coefficient of the RBMK reactor at low reactor power levels, it was now primed to embark on a positive feedback loop, in which the formation of steam voids reduced the ability of the liquid water coolant to absorb neutrons, which in turn increased the reactor's power output. This caused yet more water to flash into steam, giving yet a further power increase."

   Basically the coolant flow dropped too low allowing steam to form.  As the steam does not absorb neutrons as well as the water the reaction rate in the core increased rapidly, finally reaching ten time the normal output.  This overpressure blew the containment vessel, venting all coolant and sending lumps of superheated graphite moderator into the air where they fought fire.  The secondary explosion was more powerful than the first and was probably a combination of chemical and steam. 

   The wikipedia pages provide more than enough information for any SF author to write a convincing plot centred around a failed nuclear reactor, and the citation links provide a huge mine of further information, so I won't go any further into the mechanics of a fission reactor failure.

A still of the atmosphere processor from
James Cameron's Aliens          Source
Fusion Reactor Failure

   One of the many advantages a fusion reactor would have over a fusion design would be its relative immunity to catastrophic failure.  The reacting fuel is a thin plasma that can be vented if problems arise, since most fusion fuels are non-radioactive.  Also, if the reaction is allowed to stop the fuel cools very rapidly, unlike solid fission fuel with its decay energy.  So it seems that with a good design catastrophic failure is unluckily in a fusion reactor.  There is, however, one possible medium through which it might occur. 

   Fusion reactors contain plasma through superconducting magnetics.  The superconductivity of such magnets is dependant on their being kept below a certain temperature, called the 'critical temperature'.  Above this point the conductors used in the magnetic become normal conductors, able to carry only a small fraction of the current that they can while superconductive.  If the cooling system was damaged it might be possible for the magnets to reach the critical temperature.  The resistivity of the coils would increase suddenly, heating them.  As the temperature rises so does the resistivity.  If the energy flowing through the coils is high enough it could be released in the form of an explosion as the coils are vaporised.  More energy would be added by the fusion plasma, although I have no idea how much that would be, given the extremely low densities.  If this is possible the effect would be most prominent in reactors with the strongest magnetic coils and high power outputs.

   Of course, any good reactor would be designed to prevent this from happening.  But incompetence, cost cutting, sabotage, and damage all offer an opportunity for any safety features to be circumvented.  The result will not be the nuclear blast of Aliens, but it could be more than enough to destroy a spacecraft or space station, two places where extremely powerful fusion reactors are likely to be found.  And of course boiling lithium or sodium coolant flying all over the place would add to the destruction, especially if there was large amounts of water present, or perhaps a fluorine atmosphere?

Warning sign by Anders Sandberg of
the Lifeboat Foundation

   This hardly needs explaining.  Atomic Rockets has a much better overview of the issues with storing antimatter than I could include here, so follow the link.

   Although not strictly speaking a 'reactor' antimatter might prove to be the only way to achieve certain things.  Interstellar flight, torchships via micro-fission or fusion sparked with minuscule amounts of antimatter.  However it has the fatal flaw of reacting with anything.  Which means no matter how good your containment is, damage through accident or design is a 'bad thing', which is why the containment cylinders on the starship Enterprise could be ejected.

   Superconductors also play into this scenario.  As a superconducting electromagnet does not loose all its power instantly when the power is cut off there might be a short delay between the failure and the magnetically levitated ball of anti hydrogen contacting the containment vessel and vaporising the ship.  It might be only seconds, but those seconds could mean the difference between the crew compartment automatically ejecting or getting atomised.

Robert Zubrin's NSWR from this paper
NSWR: Nuclear Salt Water Rocket

   For the people who think that the Orion Drive is impractical there is a concept known as the Nuclear Salt Water Rocket.  Innocent sounding name, but a rather terrifying mode of operation.  Water containing enriched uranium salts is pumped into the reaction chamber where it undergoes a continuous nuclear detonation.  Premature detonation is prevented by storing the fuel in a matrix of neutron absorbing material.  Once again I refer anyone interested in further details to Atomic Rockets.  The thing is that the NSWR offers such high performance that it might be used despite the obvious risks; military, interstellar probes, and massive commercial spacecraft all have obvious benefits.  They could even form the basis of a power system with the plasma from the exhaust guided through a MHD generator.  But should the neutron absorber be damaged or the 'nuke juice' accumulate to critical mass there will be a low yield fission explosion, perhaps powerful enough to cause detonation of the rest of the fuel.  Slightly safer than antimatter.

Credit Anders Sandberg
Black Holes

   Confusingly these favourites of SF are not, in fact, black.  Through some complicated physics I don't really understand black holes give off Hawking Radiation.  Not only that but they can evaporate.  The rate of evaporation is inversely related to the mass, as is the temperature of the radiation.  If you had a very small, as in atomic radius small, black hole it would give off quite a bit of energy.  If you could stabilise a micro-blask hole by forcing matter into it at the same rate as it lost mass it would be a 100% efficient mass to energy device.  Anyone with the tech to do this has a huge advantage in terms of starship propulsion as well as all the benefits of being a Kardashev level civilisation without having to build a Dyson Shell.  Obviously if the mass input was too low the black hole would 'explode', releasing far more energy than can be contained.  It is also perfectly predictable, if you know the mass of the hole then you know when that moment will come.  This, added to increasing output would be perfect for raising the stakes aboard a post-Singularity starship.  Note that as the amount of Hawking Radiation increases it becomes harder to get the black hole to accept matter due to the sheer energy output, exacerbating the problem.

NPP: Nuclear Pulsed Power

   In pulsed power reactor a tiny nuclear bomb is detonated and the resultant energy turned into electrical power.  A wide rang of techniques are used for both the bomb and the containment/energy capture.

   Most designs would be fairly safe, as under normal conditions the pulses do not put out enough energy to destroy the chamber even under worst case scenarios.  However, a system that had been modified to produce more power, run without spare parts, or one fuel it was not intended to use, could fail catastrophically.   If to the detonation produces to much plasma/debris the containment could be destroyed.  This is most likely in a overpowered magnetic containment design.  If the containment failed the impulsive shock of the detonation on the walls of the chamber could cause massive damage, although an 'explosion' as such is unlikely.

   The two main methods of energy capture are to harvest thermal energy, or to fuse the action of the plasma against magnetic of electric fields to directly generate power.  The former could fail in the method I have already described, but the latter has other modes.  If coolant flow was cut off but fuel detonations continued the coolant could boil, rupturing the system, and causing widespread devastation.  Also, and this applies to a magnetic design as well, the presence of material in the chamber - leaked coolant, gas, or a buildup of reaction products - could magnify the mechanical effects of the explosion, just as the atmosphere does for a nuclear warhead.

Other Systems

   Spacecraft need high performance more than any other application, so it is more than likely that technologies used in space will always be cutting edge, and thus posses more failure modes that tried and true technology.  There is also the slightly cold logic that an explosion in space will probably only harm the crew of one ship even if measured in megatons, while the same detonation on Earth could level a city.  Metastable helium, metallic hydrogen, and similar materials offer vastly increased performance in both spacecraft propulsion and in power generation, but also run the risk of catastrophic failure.  Even further into the haze of a speculative future there will be even more potent dangers.  Anyhow, that should be more than enough information to avoid the common misconceptions surrounding retain failures in SF, and to come up with a more realistic and original scenario.


  1. Maybe in the future fusion reactors will have such high density of plasma. That after magnets failure the will explode with great power?
    The Atmospheric Processor in Aliens is 1,5 kilometer high, and it detonates with only 50 megatons, Tsar Bomba had around it, well maybe it will be possible?

  2. An interesting proposal on how a fusion reactor could explode, though like with normal fission reactors they are not thermonuclear warheads going off. Though I would have to assume that the destruction of the fusion-powered spacecraft would be more of what would happen when a few sticks of dynamite are placed below a car: It's not obliterated or vaporized, but it is wrecked beyond logical repair.

    Though the idea that there being a second or few before the superconductor coils loose power and the antimatter reacted with normal matter is something to keep in mind. It does certainly give a moment of panic and pure horror if that were to occur and make a perfect recipe for dramatic tension that any science fiction story would need.

    As for Nuclear Salt-Water Rocket designs, I'm divided on the idea. Granted, failures do allow such spectacular explosions as seen in visual media, and the thrust could be used as supplementary power generation via MHD, there's just something about uranium being tossed out the back at such speeds that just urks me. Kinetic impacts aside.

    Black Hole reactors eh? And here I thought it was just a soft sci-fi trope that could never be plausible. Though it wouldn't hurt if there were some failsafe designs behind the singularity reactor. Still, more boom for the buck in both directions.

    Nuclear Pulsed Power? I didn't think it was that possible, though small yields would probably make the explosions manageable. Though I can't really imagine the mentality of any organization that would power their spacecraft with nuclear explosions. Propulsion via Orion-style drives, maybe, but as a reactor? Probably would have to sleep on it just to wrap my mind around the idea.

    Still, lovely ideas to play around with. Thanks for the inspiration Moran!

  3. Great post. What do you think about pulsed ntr, it could be used for propulsion and powergeneration much higher than a normal reactor. Also have you heard about lithium 6 nswr i have heard it is safer.