The Story

Same events as this question. Kit is ordered to let a task-force handle a possible sabotage of the stations fusion power plant. She ignores them and vents the core plasma out into space to keep the reactor from reducing the lower levels to radioactive slag, herself included.

This is entry 1.5 out of 3 set in the same place, a laboratory on the moon of a gas giant in the alien star system Ilus. If you'd like context as to the story, look up there.

The Question

Armed with a marginally better understanding of how fusion reactors work, as I can understand, you'd probably be better off using some kind of pulsed tokamak style reactor. The fusion cycle would burn D-He3 in an a-neutronic wave, momentarily increasing the pressure of the plasma torus, then letting the reaction expand against the containment field and the Lorenz force generate energy, which is used in the next cycle.

The reactor would be reminiscent of a stellarator, which has a preset geometry which the plasma torus rotates within, keeping it stable, and the plasma torus would stay at near-critical condition and sinusoid between fusion and generation, similar to general fusion's ram-and-spinning-ball-o-lead reactor, but hybridized with a stellarator.

I don't discuss it much in the book, because as atomic rockets puts it simply, if someone asks you how your drive works, you tell them it works just fine. A non-answer.

The question is wether a reactor working on the tokamak principles be able to blow itself up. There is one critical piece of information here, which is that it runs on the D-He3 cycle, and not the D-T neuronic cycle, which puts off deadly neutron radiation.

For one, the D-T cycle puts off a lot more energy at a given plasma pressure and temperature, but as deadly radiation that can only be used to make heat, a lot of heat. Around 50X more energy at a given ignition state. The reactor would not be shielded to handle a purely neutronic reaction, and if tritium is injected into the plasma and let fuse, it will irradiate and slag the entire engineering wing.

But would the plan of injecting way-more-volatile fuel into the core and let it fuse slag the place? I don't expect a show of sparks per-se, but of finding the entire engineering wing full of dying technicians, horribly burned and in pools of their dissolving skin, blood and uniforms... well if Kituki had listened to her superiors, at any rate. (Well... that got nasty.)

  • 1
    $\begingroup$ My answer to the other question largely applies to this one as well, except there's already hydrogen plumbed up to the reactor (so you just need to run in some oxygen for the first option), and more magnets to contribute to a BLEVE for the third option. $\endgroup$ Feb 10 at 16:58
  • $\begingroup$ Before I vote to close this question (and with the science-based tag, my vote might slam it shut), please explain how a more specific version of a previous question isn't yet a duplicate of the previous question? Nuclear reactors are intentionally designed to not explode like bombs but can explode like Cherynoble. It's not enough to say, "yeah, but this time it's a specific reactor!" What's the difference that you hope can be ansewred here that couldn't, ever, be answered there? $\endgroup$
    – JBH
    Feb 11 at 0:56
  • $\begingroup$ @JBH I am asking wether or not stuffing a tokamak reactor that is supposed to burn D-He3 with D-T will melt/destroy it. In retrospect I think the nuclear engineering or space travel stack might have been a better choice. $\endgroup$ Feb 11 at 3:22
  • $\begingroup$ @JBH I forgot to include an ignition energy curve in my question. D-T reactions are much, much easier to ignite than D-He3, thus the D-T ignition would release a lot more energy, or fuse fuel a lot faster than otherwise intended. I.E. an explosion. With my meager understanding of nuclear physics, if you were to look at, for example, Trenta, Helion Energy's dual FRC donut-slammer, they plan on building one for breeding He3 out of D-D, then powering 5 or 6 other reactors with the product. It's really hard to find He3 naturally, and mining it on the moon is pathetic. $\endgroup$ Feb 11 at 12:13
  • $\begingroup$ The thing is, D-D spews off neutrons while D-He3 doesn't, or at least way less. The breeder reactor would be extremely armored and make nuclear waste as a result, but the generators can be cheaper and simpler, with less shielding, to basically none at all. So if you feed it a fuel that does throw off neutrons and burn it, won't it irradiate the area? $\endgroup$ Feb 11 at 12:15

3 Answers 3


Not exactly

The most interesting thing with tokamak design is that it requires energy to contain plasma, and the output is (yet) smaller than containment spending plus heating spending. Thus, any breach of plasma will just shut the thing off, probably with minor to medium destruction in and around the plasma containment vessel. If your reactor would be able to produce say 1000x energy, then it'll "sort of" blow up when the entirety of plasma containment would discharge against the vessel, but since there will no longer be conditions for fusion to continue, it would blow up like a steam tank, not like a fission reactor, and any attempts to restart it would be futile but won't cause any extra damage to the environment.


D/T fusion is likely to be how future tokomaks already work

ITER in France, the next big tokomak, for example, is planned as a D/T reactor. We also see neutron radiation with existing reactors, so larger ones will have better shielding. But, you've still got options:

Capacitor banks

Sabotage these, and there's a big explosion. They'll have a bunch of safeguards

Blow up the reactor wall with conventional explosives

Sort of a no brainer, really - hot plasma/radioactive gas + beryllium dust (used in the reactor walls, and way worse than the radiation), and a bunch of superconducting magnets quenching simultaneously is going to make a serious mess of a place. Cleanup would be a few weeks to get the mess down to the point where you could start working on it.

More subtle - channel the plasma

Fusion reactors already have events where the plasma loses containment, or sort of runaway electron events, which induce massive charges in the reactor walls. It's one of the problems to solve on larger reactors. The plasma should, on spots without containment, do some serious damage - if you can focus it to one spot, you should be able to burn neatly through the reactor wall. I'm not enough of an engineer to know what happens when stupidly hot plasma reaches, say, the supercooled magnets, but I imagine explosions and shards of metal are going everywhere.

It might not kill anyone, but suddenly, this bit of the reactor that previously could be worked on by people in regular PPE suddenly requires fixing by people wearing radiation monitors and backpack respirators.

Kill the magnets

Similar to above, vent all the coolant for the supercooled magnets, and have them all quench. Containment gets lost, though it'll probably only trash the reactor walls. You might get some exploding magnets. Either way, this isn't a fusion reactor that's turning on for a while.

  • 1
    $\begingroup$ D-T fusion gives off an energetic neutron and a gamma. These will penetrate about a foot of low-Z material. If you have a fusion reactor that has run for years, the whole thing will probably be radioactive. $\endgroup$ Feb 10 at 10:33
  • $\begingroup$ Oh, yeah, I think the plan was to have giant, removable concrete blocks, that you take out and dump when it gets too radioactive. You could use boron carbide, which is pretty good at not being turned into radioisotopes by the neutrons. I'd argue that water might be the best call - if you dumped it far enough out at sea, you'd be unlikely to increase the radiation count of the ocean in thousands of years - it could be so dilute that would be nothing above background $\endgroup$
    – lupe
    Feb 10 at 15:42

Not an expert so tapping in something I read a while back. I hope it helps. https://phys.org/news/2009-08-honey-blew-tokamak.html https://archive.is/ZgHda (archive link)

Magnetic reconnection could be the Universe's favorite way to make things explode. It operates anywhere magnetic fields pervade space--which is to say almost everywhere. On the sun magnetic reconnection causes solar flares as powerful as a billion atomic bombs. In Earth's atmosphere, it fuels magnetic storms and auroras. In laboratories, it can cause big problems in fusion reactors. It's ubiquitous.

So it does have basis in our reality.

The basics are clear enough. Magnetic lines of force cross, cancel, reconnect and—Bang! Magnetic energy is unleashed in the form of heat and charged-particle kinetic energy.

Here's the mechanism ^

"For many years, researchers have looked to fusion as a clean and abundant source of energy for our planet," says Burch. "One approach, magnetic confinement fusion, has yielded very promising results with devices such as tokamaks. But there have been problems keeping the plasma (hot ionized gas) contained in the chamber."

"One of the main problems is magnetic reconnection," he continues. "A spectacular and even dangerous result of reconnection is known as the sawtooth crash. As the heat in the tokamak builds up, the electron temperature reaches a peak and then 'crashes' to a lower value, and some of the hot plasma escapes. This is caused by reconnection of the containment field."

A Fusion reactor could reproduce this phenomenon. Buuuuttt....

In light of this, you might suppose that tokamaks would be a good place to study reconnection. But no, says Burch. Reconnection in a tokamak happens in such a tiny volume, only a few millimeters wide, that it is very difficult to study. It is practically impossible to build sensors small enough to probe the reconnection zone.

So maybe something something Massive fusion plant, something something conditions make abnormal conditions that support a massive magnetic reconnection. I trust you can write something in.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .