Fusion reactor explosion.

[Sky Captain]

I finished watching Expanse third season and there in one episode shipboard fusion reactor exploded like a nuke when magnetic containment was lost. In real life it is said that fusion reactor like ITER can't explode at all.

However big difference is shipboard reactors in Expanse are crazily powerful capable of powering torch drives to do sustained thrust travel across the solar system in few weeks. That easily puts output into terawatt range even for small ships like Roccinante. If you have that much power in a small space and something critical fails it should make fairly big boom. It got me wondering are there any theoretical fusion reactor designs that could fail catastrophically if some kind of containment failure happens.

[Lord Revan]

I finished watching Expanse third season and there in one episode shipboard fusion reactor exploded like a nuke when magnetic containment was lost. In real life it is said that fusion reactor like ITER can't explode at all.

However big difference is shipboard reactors in Expanse are crazily powerful capable of powering torch drives to do sustained thrust travel across the solar system in few weeks. That easily puts output into terawatt range even for small ships like Roccinante. If you have that much power in a small space and something critical fails it should make fairly big boom. It got me wondering are there any theoretical fusion reactor designs that could fail catastrophically if some kind of containment failure happens.

short answer:none that I know of.

Longer answer:What is generally forgotten is not only how insanely difficult it is to get sustained controlled fusion reaction, but also why is it so difficult aka it's not control part that's hard but sustain. There's no theoretical fusion reactor design that doesn't essentially be struggling to stay on and still be practical to use.

That said in case of reactor failure you'd have a lot very hot hydrogen leaking from the reactor so you might still get a big boom just not an nuclear one.

[Jub]

That said in case of reactor failure you'd have a lot very hot hydrogen leaking from the reactor so you might still get a big boom just not an nuclear one.

If that's the risk you just need a blow out path for the hydrogen to follow and that's easy enough to do. On a spaceship that basically just needs to be a couple of one-way valves set to open if a pressure and temperature threshold are met and that close once they aren't and that's if the hydrogen even expands enough to warrant any safety measures at all.

[Zixinus]

A fusion reactor blowing up would be an implosion, as fusion reactors operate in a vacuum.

That said, a fusion reactor going out of control would be comparable to a regular nuclear reactor going out of control and there is the electrical surge. However, considering that the reactor would likely be behind heavy shielding, the overall reaction would most likely just turn most of the reactor (and directly connected systems) into slag. This would leave the starship dead in the void, as opposed to vaporizing the entire starship.

This of course is in the end just artistic license coming through that turns every nuclear reactor into a nuclear bomb, regardless of pretty much anything. Because even if the artists, directors and such care, there is the old but tried rule of "more explosions=better".

That said, if there is a lot of superdense energy storage or munitions, that too could catastrophically blow up.

[Sky Captain]

One thing to keep in mind is torchship reactors are extremely powerful for their size, like several terawatts or more in a space of small house. If that power goes out of control like for example reactor cranked to full power without that power being fed into engines something is going to fail rapidly. If several tons of metal flashes into plasma that is going to blow up a portion of a ship. Maybe engineering section blowing up and rest of the ship spinning out of control would be plausible result.

[Formless]

Given that fusion reactions require intense heat and pressure to keep going, and that many forms of combat damage would only serve to lessen the amount of pressure the reactor could create, I think the only way the trope could work with a fusion reactor is if they are using something like antimatter as a catalyst for fusion. In which case, its really when you lose containment for the antimatter that you are fucked, or maybe if you are using hydrogen and it finds something it can chemically react with like oxygen (which is as easy to prevent as simply keeping the hydrogen in a vacuum). However, the antimatter scenario will create a much bigger boom.

[Zixinus]

Another thing: there are actually only a few grams or so of plasma in the reactor. This is from the ITER site, so maybe that figure goes up but a few hundred grams of plasma in the reactor.

The question thus becomes: how much damage would say, 10 grams, of super-accelerated plasma do? Without being confined by electromagnetic fields, you have an electrically charged, expanding gas. So an explosion. I'm pretty sure it won't be pretty but the question is would that be enough to just slag its own reactor or vaporize up the ship? What would be the explosion be equivalent to in TNT?

As for intense heat and pressure, that's not quite accurate. The pressure is created by electromagnetism and the temperature created by accelerating particles inside the reactor. Since temperature is the average energy of any given amount of particles and since you accelerated all of them, you have high temperature. So it's not like temperature and pressure in the regular sense that you are using something to compress matter.

Again, the reactor is actually in a vacuum. Anything going INSIDE the reactor, even air (but particularly bad is iron), would grind the reaction to a halt.

One thing to keep in mind is torchship reactors are extremely powerful for their size, like several terawatts or more in a space of small house. If that power goes out of control like for example reactor cranked to full power without that power being fed into engines something is going to fail rapidly. If several tons of metal flashes into plasma that is going to blow up a portion of a ship. Maybe engineering section blowing up and rest of the ship spinning out of control would be plausible result.

Yes, but nuclear power doesn't work like a pressure boiler. You can't crank up or down the power quite like that (as far as I know). As far as I can figure out, a nuclear reactor's output is fixed. Fusion reactors in particular are relatively delicate devices that require to be constantly fed energy as much as they are required to give out. So the biggest thing a fusion reactor would be fueling is itself, or be fed by another fusion reactor. I would guess power banks would be used to deal with fluctuations and there must be some way to vent off excess energy.

[tezunegari]

You can't crank up or down the power quite like that (as far as I know). As far as I can figure out, a nuclear reactor's output is fixed.

IIRC Max output is fixed. You can regulate the output of a nuclear reactor. That's what the Control Rods are used for. Insert them into the reactor and the output goes down as neutrons are absorbed and thermal output decreases. Pull them out to increase the thermal output.

Perhaps the fusion reactors used in the Expanse are vastly different from the one being developed in RL.
Or it's just the common Rule-of-Cool with big explosions.

[Batman]

Tezunegari already dealt with fission reactors and I don't see why a fusion reactors output would be either. A minimum to keep the reaction going, sure, and a maximum everything beyond which would exceed the design limitations and likely result in reactor failure (though almost inevitably NOT in the often portrayed 'giant nuclear explosion' way) but everything in-between should be doable by varying how much fuel I feed into the reactor.

[Sea Skimmer]

This is a question of how much latent energy is stored inside the process at any given instant, and how that energy is stored. Nuclear fission reactors for example cannot blowup in sustained fission chain reactions, but depending on design they can blowup in multiple other ways, as demonstrated in multiple real life disasters while the amount and way they can explode radically varies depending on design details.

A fusion reactor like ITER totally could explode, but most of the energy stored inside the process is stored as the energy of the magnetic fields, like 40 gigajoules of energy level of magnetic field. If this exploded it would explode in the form of an electrical arc which would vaporize whatever it grounded into. The actual result would thus be very detail specific, but limited by that ~40 GJ stored energy. This also means the explosive results would be highly directional, whatever place that energy arcs explodes outward while the other side of the reactor would be largely unaffected.

I believe the stored plasma energy is only around 400 MJ for the ITER reactor, but I read that figure a long time ago and it may have changed since, that whole project is a nightmare of changing design points. But with any tokamak you need way more power charged up to maintain the containment field then you have in the plasma itself. It really wouldn't make sense otherwise given how weak magnetism is at doing anything!

But something else to consider, a reactor like ITER produces mostly neutron energy which meanss needs a massive radiation shield not just to protect human life and equipment, but simply to convert the neutrons into heat, and that shield would pretty easily mitigate the effects of an explosion. But different forms of fusion are possible given different fuel, and would allow for much less heavily shielded reactors. Still if you had a terrawatt class reactor the radiation shield would be something comical, like on par with 10m thick reinforced concrete sandwiched with some steel plates. If not greater, but radiation shielding and scaling effects get complicated due to spectral hardening. Even from a very high power radiation source a thin shield will stop most of the radiation effects, but as power rises the more penetrating part of the spectrum (that you can just ignore if the total power is low enough, all nuclear plants allow some of this) becomes more relevant and can be extreme hard to stop.

Anyway what that means is the radiation shield for a 1 TW plant is going to be massive, and probably directionally biased, so even if a very massive explosion takes place it would be directed away from the rest of the ship by the shield.

[Sky Captain]

Interesting info, Maybe some sort of futuristic tokamak design could have magnetic containment field so strong that total stored energy is far greater than in ITER leading to more spectacular failure modes if something goes wrong.

What if containment is not magnetic field, but some kind of sci fi force field allowing to compress fusion fuel like in a core of a star to ignite it. That certainly could explode catastrophically if containment is lost and all that plasma suddenly can expand freely.
But realistic designs that could be built in near future most likely would just destroy the reactor and surrounding equipment in case of failure while rest of the spacecraft are protected by massive radiation shield. If only one reactor then crew die slow death from lack of power not from explosion.

[Formless]

What if containment is not magnetic field, but some kind of sci fi force field allowing to compress fusion fuel like in a core of a star to ignite it. That certainly could explode catastrophically if containment is lost and all that plasma suddenly can expand freely.

The problem with that idea is that its trying to do things backwards. You are trying to make the reactors fail-deadly on purpose just to make it fit the SF trope. But real engineers NEVER try to make their machines less safe, they make them as explosion-proof as possible. And the best way to do that is to introduce passive safety mechanisms, like reactors with armored exteriors (armor doesn't simply cease to exist when the thing goes boom, which means it can contain or control the consequences of the explosion). That doesn't mean the things can't have a catastrophic failure mode, just that its idiotic to design things specifically to explode for dramatic effect unless exploding is their purpose-- that is, unless you are trying to make a bomb.

And there are interesting dramatic consequences for not using the SF trope of exploding spaceships as well. If the explosion of the reactor core doesn't necessarily atomize everything onboard, that opens up the possibility of the crew and equipment surviving, with all the dramatic possibilities that implies. Recovered wrecks, rescue operations, survival stories as a plot unto itself, and even boarding actions in a war stories where personnel and equipment can be captured-- suddenly the idea of space marines can actually make sense! It also justifies self-destruct mechanisms existing, because there are times where you do want the ship to explode, but can't rely on the reactor overloading to do the job for you.

[bilateralrope]

What about a reactor design that won't explode most of the time, but will explode if someone really wants it to ?

The episode I think Sky Captain is talking about is one where the reactor is intentionally made to explode because someone wanted to completely destroy the ship.

Salvaging wrecks left after a battle was something that happened on The Expanse. From the dialogue, it sounded like something that is often possible. So there must be regular instances where battle damage doesn't lead to large explosions.

[Zixinus]

What about a reactor design that won't explode most of the time, but will explode if someone really wants it to ?

The thing you have to remember that "reactor explodes" =/= "nuke". If you want to see what a fission reactor explosion looks like, look at Chernobly. The explosion itself was not that big of a deal compared to a bombing run during war-time but it still created catastrophic damage. This is ignoring the fact that most reactor domes are deliberately designed to contain an explosion if worst comes to worst. You can create an explosion in any reactor, no problem, but reactors are inherently designed to contain and handle reactions that gone out of control (to a point). It's like asking to make a fireplace that both prevents the possibility of wild-fires but also encourages on a whim.

If you wanted to blow the ship to unusable smithereens, because you're taking "death before surrender" literarily, it would be both simpler and easier to just plant bombs all over the ship. A ship whose reactors got turned to slag is out of the fight, there is no need to turn it into vapor other than the visual addage of "more explosions = bettter".

[Lord Revan]

What about a reactor design that won't explode most of the time, but will explode if someone really wants it to ?

The thing you have to remember that "reactor explodes" =/= "nuke". If you want to see what a fission reactor explosion looks like, look at Chernobly. The explosion itself was not that big of a deal compared to a bombing run during war-time but it still created catastrophic damage. This is ignoring the fact that most reactor domes are deliberately designed to contain an explosion if worst comes to worst. You can create an explosion in any reactor, no problem, but reactors are inherently designed to contain and handle reactions that gone out of control (to a point). It's like asking to make a fireplace that both prevents the possibility of wild-fires but also encourages on a whim.

If you wanted to blow the ship to unusable smithereens, because you're taking "death before surrender" literarily, it would be both simpler and easier to just plant bombs all over the ship. A ship whose reactors got turned to slag is out of the fight, there is no need to turn it into vapor other than the visual addage of "more explosions = bettter".

There's also the problem that if you got a reactor that is designed to explode intentionally, you'd have to design it so that it won't explode when you don't want it to.

to use the fireplace analogue you'd want a fireplace that prevents the possibility of wild-fire but also encourages them when you want that to happen but only when you want it to happen.

Basically you're getting this highly complex device to "re-invent the wheel" so to speak just to get some pretty looking explosions.

[Sky Captain]

What if containment is not magnetic field, but some kind of sci fi force field allowing to compress fusion fuel like in a core of a star to ignite it. That certainly could explode catastrophically if containment is lost and all that plasma suddenly can expand freely.

The problem with that idea is that its trying to do things backwards. You are trying to make the reactors fail-deadly on purpose just to make it fit the SF trope. But real engineers NEVER try to make their machines less safe, they make them as explosion-proof as possible. And the best way to do that is to introduce passive safety mechanisms, like reactors with armored exteriors (armor doesn't simply cease to exist when the thing goes boom, which means it can contain or control the consequences of the explosion). That doesn't mean the things can't have a catastrophic failure mode, just that its idiotic to design things specifically to explode for dramatic effect unless exploding is their purpose-- that is, unless you are trying to make a bomb.

And there are interesting dramatic consequences for not using the SF trope of exploding spaceships as well. If the explosion of the reactor core doesn't necessarily atomize everything onboard, that opens up the possibility of the crew and equipment surviving, with all the dramatic possibilities that implies. Recovered wrecks, rescue operations, survival stories as a plot unto itself, and even boarding actions in a war stories where personnel and equipment can be captured-- suddenly the idea of space marines can actually make sense! It also justifies self-destruct mechanisms existing, because there are times where you do want the ship to explode, but can't rely on the reactor overloading to do the job for you.

What if having a reactor with extreme performance has unavoidable side effect of it blowing up if damaged or sabotaged even when designed as safe as possible. It may be acceptable aboard a warship if performance benefits are big enough. While not very realistic for near future fusion reactors a catastrophic ship vaporising explosion certainly are possible with antimatter systems.

[Formless]

What if having a reactor with extreme performance has unavoidable side effect of it blowing up if damaged or sabotaged even when designed as safe as possible. It may be acceptable aboard a warship if performance benefits are big enough. While not very realistic for near future fusion reactors a catastrophic ship vaporising explosion certainly are possible with antimatter systems.

Here's a trick many people don't consider when talking about such dangerous fuels as antimatter or antimatter doped fusion fuels: you don't need to put the reactor or engines in the center of the ship like so many science fiction starship designers assume. You can put it behind the crew compartment in a separate module of the ship, and only mess with it via drones and automation or EVA suits when absolutely all other maintenance functions are offline somehow. This has the advantage that you don't need (but probably still want) as much armor surrounding the reactor or engine, whether that armor exists to stop radiation that results from fusion reactions or matter-antimatter annihilation or to prevent catastrophic explosions from destroying the rest of the superstructure. Think of the design of the spaceship from 2001: A Space Odyssey. Long and spindly, with a very long neck separating the main engine from the sphere containing the crew and mission equipment/machinery such as HAL's computer systems. The distance alone grants significant protection from radiation coming off of the engines, although that can be further enhanced with shielding which is not present in the movie. Its another fail-safe design that any science fiction universe could apply if the writers only realized its purpose, because even if the engine could blow up, it would simply sever the crew area from the engine. She would be dead in the water, but all she would be missing is her engines. The crew and computers would survive. Many realistic designs you will see out there for proposed Mars missions use a similarly spindly design philosophy even though we don't expect those spacecraft to be shot at. The engines and fuel for the same just doesn't need to be in close proximity to the astronauts.

In theory, this is why the Enterprise in Star Trek has a neck separating the engineering section and warp nacelles from the saucer section where everyone actually lives, and also the reason in theory that the Enterprise D specifically also has its saucer section separable from the engineering section. That's principally a safety precaution-- it gets everyone away from the engineering section just in case its about to blow. Although obviously in practice the two sections are far too close to one another, and you wouldn't actually want anyone working in the engineering section so you never have to evacuate it (which seems to be the only explicable reason they rarely ever use the saucer separation protocol other than it ups production costs for the show). But you see how this works: everything can be made fail-safe if you simply put enough distance between the dangerous parts of the ship and the parts where people actually live and work.

[Zixinus]

The other thing is that you don't need to put the anti-matter storage in the center of the ship, but far away and remotely. Sure, you'd armor it and such but if it is a warship you have to expect incoming fire. In which case, you'd plan for the anti-matter containment failing during combat and make it that most of the energy is directed away from the important, crew-held parts. A crippled ship with survivors and data is better than one that is vaporized. This assumes that the ship's primary defense is armor, rather than point-defenses or some sort of force-field shield.

[Sky Captain]

Having seperate engineering section connected to habitat module with long truss structure certainly would help if something blows up and also helps against radiation. However this design seems more suitable for civilian ships , Connecting truss would be a weak point, if it gets damaged structural integrity of whole ship is compromised. Armoring it would add a lot of mass. Long ship with heavy items at both ends would be difficult to maneuver . Warships would need to maneuver rapidly. That requires as much mass as possible concentrated in the middle to lessen moment of inertia.

[Formless]

Having seperate engineering section connected to habitat module with long truss structure certainly would help if something blows up and also helps against radiation. However this design seems more suitable for civilian ships , Connecting truss would be a weak point, if it gets damaged structural integrity of whole ship is compromised. Armoring it would add a lot of mass. Long ship with heavy items at both ends would be difficult to maneuver . Warships would need to maneuver rapidly. That requires as much mass as possible concentrated in the middle to lessen moment of inertia.

The connection doesn't need to be as thin as you see it in 2001, nor as exposed as it is in Star Trek. And the distance isn't necessarily that big, just big enough to let the Square-Cubed law protect you from most of the radiation that the engines and reactors put off during normal functioning. The center of mass will be somewhere in the middle of the connecting truss, but just because the truss serves a safety and structural purpose doesn't mean you can't put guns there as well. And armor. In fact, make some of the armor perpendicular to the direction of radiation coming off the engines and reactors while you are at it, and you can make the engines more powerful without frying the crew with x-rays. Then add heat radiators, because those need to be there on any realistic spaceship design anyway, and suddenly that connecting truss is actually serving a lot of functions beyond the obvious structural function. You design the ship around the need to separate crew from radiation-spewing machinery and far from making a less maneuverable bar-bell shaped object you instead end up with a logical, planned out attack ship that makes the most of its volume. In fact a lot of Hard SF designers assume a long cylindrical ship anyway to cut down the surface area exposed to weapons fire (especially lasers), but its not a problem because we aren't talking about a pressurized tin can here. Most of that space is (or would logically be) hard vacuum since no one need occupy it. With no atmosphere to absorb X-rays and Gamma rays, any explosion from the reactor or fusion/antimatter engines will mostly release a lot of radiation that will be absorbed by the aforementioned armor. As opposed to turning into a shockwave of superheated plasma that destroys the superstructure spectacularly (see every science fiction series since TNG). Point is, such designs are overall more, not less durable and combat effective by all criteria you present. Hell, probably the lightest section of the ship will be wherever the crew actually lives, probably but not necessarily the front section. These designs only sound like their weight should be distributed like a barbell, but that isn't necessarily the case.

[Zixinus]

I also have two schools of thoughts about how to store anti-matter.

The first has you distribute the storage for them that are on the outside the ship. So if one hit manages to get through shields, most of the explosion will happen outside the ship rather than inside. You can also take advantage of the fact that spaceships can choose to show whaterver side they want towards the enemy and simply put the vulnerable storage space on part of the ship that faces away from the enemy. You do want to do that with engines anyway, as you need to expose them anyway. Of course this potentially makes the ship vulnerable to flanking from opposite sides.

The other is to design the storage to be made in similar way as you handle munitions. You put them in a double-armored section (called citadel in a warship) within the ship and assume that any hit that manages to get far would have resulted in the warship's (and crew's) loss anyway.

[Sky Captain]

Having seperate engineering section connected to habitat module with long truss structure certainly would help if something blows up and also helps against radiation. However this design seems more suitable for civilian ships , Connecting truss would be a weak point, if it gets damaged structural integrity of whole ship is compromised. Armoring it would add a lot of mass. Long ship with heavy items at both ends would be difficult to maneuver . Warships would need to maneuver rapidly. That requires as much mass as possible concentrated in the middle to lessen moment of inertia.

The connection doesn't need to be as thin as you see it in 2001, nor as exposed as it is in Star Trek. And the distance isn't necessarily that big, just big enough to let the Square-Cubed law protect you from most of the radiation that the engines and reactors put off during normal functioning. The center of mass will be somewhere in the middle of the connecting truss, but just because the truss serves a safety and structural purpose doesn't mean you can't put guns there as well. And armor. In fact, make some of the armor perpendicular to the direction of radiation coming off the engines and reactors while you are at it, and you can make the engines more powerful without frying the crew with x-rays. Then add heat radiators, because those need to be there on any realistic spaceship design anyway, and suddenly that connecting truss is actually serving a lot of functions beyond the obvious structural function. You design the ship around the need to separate crew from radiation-spewing machinery and far from making a less maneuverable bar-bell shaped object you instead end up with a logical, planned out attack ship that makes the most of its volume. In fact a lot of Hard SF designers assume a long cylindrical ship anyway to cut down the surface area exposed to weapons fire (especially lasers), but its not a problem because we aren't talking about a pressurized tin can here. Most of that space is (or would logically be) hard vacuum since no one need occupy it. With no atmosphere to absorb X-rays and Gamma rays, any explosion from the reactor or fusion/antimatter engines will mostly release a lot of radiation that will be absorbed by the aforementioned armor. As opposed to turning into a shockwave of superheated plasma that destroys the superstructure spectacularly (see every science fiction series since TNG). Point is, such designs are overall more, not less durable and combat effective by all criteria you present. Hell, probably the lightest section of the ship will be wherever the crew actually lives, probably but not necessarily the front section. These designs only sound like their weight should be distributed like a barbell, but that isn't necessarily the case.

Those are good points. Maybe truss could also serve as a place for bolt on missile pods or extra fuel tanks if mission delta v demands it. Probably also could have docking hardpoints for combat drones or some other small craft.

[DKeith2011]

The best explanation for fusion reactor explosions I ever heard came from an old BattleTech sourcebook; Never let reality stand in the way of a cool explosion.

[GrosseAdmiralFox]

The best explanation for fusion reactor explosions I ever heard came from an old BattleTech sourcebook; Never let reality stand in the way of a cool explosion.

Actually, the source books have disprove the 'Stockpool' (sp?) thing. The worst a BT fusion reactor (the ones fitted onto aerospace fighters, battlemechs, protomechs, and vehicles to be specific, although a clan Battle Armor fusion reactor prototype was so unstable that once breached, there wasn't much left of the tester from what I've heard) you'll get a steam explosion, not a nuke. A nuke-like explosion that happened during the Jihad on a world was because of a combination of dilapidated facilities, unsafe expansions, and a series of events that led to the nuke-like explosion that occurred... although it also might be the Blakists actually nuking the place...

The only way you get usual SciFi explosions is having something similar to Honor-verse fusion reactors where the plasma is under such pressure that loss of containment means a lot of plasma being released.

[SpottedKitty]

The only way you get usual SciFi explosions is having something similar to Honor-verse fusion reactors where the plasma is under such pressure that loss of containment means a lot of plasma being released.

Isn't that a case of very different reactor designs? The ITER-style reactors mentioned upthread use magnetic containment on a low-density plasma. Honorverse reactors (the ones on large capital ships, anyway) use gravitic containment with a much higher energy density — that's partly why they go kaboom in such an impressive way when containment fails.