# Could we use nuclear bombs to generate electricity? [closed]

it seems the biggest problem with fusion power is confinement. on the other hand, we developed a fusion bomb in just a few years. i was wondering if it could be economically viable to use nuclear bombs as a power source.

specifically, i imagine 2 possible designes.

1. a giant air tank. if the tank is large enough, we could set off a bomb inside without exceeding the containment strength of the perimeter, then harvest power from the increased pressure in the tank. i imagine making a large air-tight tank might be cost-prohibitive, although a 19 million cubic foot tank is already in use.
2. perhaps we could build a giant piston engine, where the multi-ton pistons are driven up a cylinder by nuclear bombs, then energy is harvested from their slow decent back down.

is the main problem with these plans cleanup of the nuclear waste from the bombs? would the cost of the bombs themselves exceed the value of the electricity produced? or is the compression wave from the bomb simply too powerful to withstand at any distance small enough to harvest useful energy?

• I am under the impression that some of the nuclear material from decomissioned weapons is reused as fuel rods in nuclear power plants. – chaz327 Apr 21 '16 at 2:51
• I think another interesting question would be to ask how energetically efficient a nuclear weapon is? By that I really mean how much fuel you need to mine and refine per joule. A weapon is designed to release a large amount of energy very quickly, but it may not be very efficient. I don't know the answer to this. – tfb Apr 21 '16 at 9:55

## 5 Answers

This was actually considered under the Plowshare project The idea being to use an underground nuke to heat the surrounding rock and then run it like a geothermal resource. However, radiation problems.

• yes, radiation is the big problem. – koolwithk Apr 21 '16 at 18:21

Theoretically, yes. The problem is that there's no way to build a "small" thermonuclear warhead. Fusion isn't as simple as fission, the latter being as easy as smacking the right amount of Plutonium together. It is technically energetically favorable to fuse heavy isotopes of hydrogen into helium, but the conditions required to do so include giving the system in question an immense amount of energy to begin with; you have to get the hydrogen nuclei close enough that particles can undergo quantum tunneling between the nuclei. In practice, this means constructing a fission bomb of sufficient yield to act as a catalyst for fusion. This means that there's a minimum destructive force that comes with building a fusion bomb, one which is quite high. A chamber big enough to harvest energy from the induced pressure change without being destroyed would cost a fortune's fortune, to say nothing of the nukes you put in it. Currently, small scale fusion is being attempted (most successfully in a California lab named the National Ignition Facility) by concentrating dozens of extremely high energy lasers on a single chamber of fusionable material. As of 2016, they have yet to break even in terms of energy in minus energy out.

By popular demand (1 guy in the comments), If we consider a chamber that harvests energy using walls that are all pistons which presumably push a magnet through a coil or something (this isn't actually important, it just matters that the energy imparted to the walls is the captured energy), Then the energy on any one of the walls is the integral of the energy's intensity over the surface of those walls. For simplicity's sake, we'll take the container to be an expanding sphere, so it absorbs the entire energy for any radius of the sphere. Then, the energy is equal to the work done expanding the container: $$E=-W=-(-P\Delta V)\to\Delta V=\frac{E}{P}$$ So this spherical container, for the smallest of nukes would need to expand by about 16.52 cubic kilometers ($P=1$atm, $E=400$kTon of TNT, the first fusion weapon tested by the USSR in 1953). Which means the housing needs to be at least that big. I'm not a materials scientist, nor an engineer, so I have no idea how much this would cost, or what it would need to be built out of and how large it should be initially to withstand that stress, but you can see that the facility would have to be massive for the smallest of bombs.

Also, according to the WolframAlpha database, this volume is roughly three times the volume of the oil consumed by the entire world on a yearly basis.

• I suppose I should clarify, fission "bombs" can be made very small, because decay will happen, just not destructively for low density of fissile materials. – ocket8888 Apr 20 '16 at 18:31
• Lab in California - National Ignition Facility – Tweej Apr 21 '16 at 8:15
• @ocket8888 I think that very small fission devices are also very inefficient. They get to be very small by not burning much of the fuel, so they are certainly very dirty and also I suspect inefficient. – tfb Apr 21 '16 at 9:57
• Actually they're much more efficient than a regular destructive bomb, they work by performing slow, controlled fission in a water reservoir to make steam. Since the process isn't explosive, the fuel doesn't get blown apart, so it reacts more completely. – ocket8888 Apr 21 '16 at 18:04
• @ocket8888 Sorry: I meant 'fission device' to mean weapon, such as the W54: I'd not noticed your quotes around the word 'bomb'. – tfb Apr 21 '16 at 18:49
1. a giant air tank. if the tank is large enough, we could set off a bomb inside without exceeding the containment strength of the perimeter, then harvest power from the increased pressure in the tank. i imagine making a large air-tight tank might be cost-prohibitive, although a 19 million cubic foot tank is already in use.

2. perhaps we could build a giant piston engine, where the multi-ton pistons are driven up a cylinder by nuclear bombs, then energy is harvested from their slow decent back down.

Those are external and internal combustion engines powered by fusion instead of combustion. It won't work for the reason you correctly identified: containment. We have no way to contain a fusion bomb. But you're onto something.

Do it without the containment. This is hideously inefficient, but there's so much energy in a fusion reaction it's still better than a chemical reaction. This is the idea behind the Project Orion spacecraft.

Drop tiny nuclear bombs out the back of the craft. Have them explode a short distance away. It impacts a huge pusher plate on the rear of the spacecraft propelling it forward. Repeat. The extremely high mass-to-energy ratio of fission and fusion reactions, compared to chemical ones, makes this spacecraft design both very efficient on fuel, and able to produce very high thrust. The explosion can be shaped so more of its energy impacts the plate than a spherical one.

Obviously you don't want to do this on Earth, so it would have to be built in space. And it was planned using small fission bombs with less than 1kt yield for an acceleration curve that won't wreck the spacecraft or its crew. But with a massive enough spacecraft, a stout enough plate and buffers, fusion might be practical.

As you say, the problem is confinement (aka containment) - fusion is relatively easy; controlled, useful fusion turns out to be incredibly difficult and expensive.

So the easiest way to harness it, and the only technically and economically viable way (for decades from now, and possibly a century or more into the future), is to have a large fusion reactor in space, and then to directly or indirectly harness the electromagnetic waves generated by the fusion. As it happens, we're already experts at doing that harnessing via photovoltaics, hydro, biomass and onshore wind; and we're learning quickly in concentrating solar thermal and offshore wind. All of which are ways we convert the sun's fusion energy into electricity.

No... In a nuclear bomb, energy does not last long enough... For continuous power, simply not enough......

But it might be possible to power a satellite launch vehicle with a nuclear bomb... But hydrogen fuel cells have an excellent efficiency which (probably) makes it a better option...

All in all, I would say no...

Regards, Pradyoth Shandilya