Could a balloon use non-fissile isotope, like Pu-238, to heat air in the envelope to stay afloat for years?

let's put political and health related problems aside, and consider if it's technically possible at all. Say, as an unmanned, high-altitude scientific/weather station.

I know that:

  • Pu-238 has power output of 0.54 watts per gram
  • Its half-life is 87.7 years (power output halves after that time).
  • great most of the energy can be radiated out as heat, and transferred to surrounding air.

I don't know what buoyancy can the hot air balloons provide, and what kind of energy input they require to stay afloat. So - would it fly?

  • 1
    $\begingroup$ Well, what is the weight for a nuclear reactor for this reaction? $\endgroup$
    – kevin
    Apr 10, 2017 at 9:21
  • $\begingroup$ @kevin: 0.54 watts per gram, halved after 87.7 years ; as large as you like, within reason, Likely double the mass due to radiators and thermocouples, also substract 5% as electricity for the probe. $\endgroup$
    – SF.
    Apr 10, 2017 at 9:23
  • 4
    $\begingroup$ If this is not possible, there is still helium. $\endgroup$
    – mins
    Apr 10, 2017 at 16:45
  • $\begingroup$ well, I could attach it to a nuclear warhead and send it up in the sky for a few brief moments... $\endgroup$
    – dalearn
    Apr 10, 2017 at 17:04
  • $\begingroup$ Fortunately nuclear-powered bombers and cruise missiles projects were all canceled but unfortunately they are all technically possible. $\endgroup$ Oct 10, 2017 at 21:35

4 Answers 4


Most likely no.

RTG does not have enough power. The burner of the hot air balloon requires 2 to 4 MW of power (here) so you need at least 4 tonnes of Pu to produce that (2000,0000 / 500 = 4000 Kg), assuming 500 W per kg (0.5 W/mg). From the source, a moderate size balloon can lift few hundred kilograms at most.

The true nuclear reactors provide much more power, but they are heavy. Hyperion, for instance, provides 25 MW and weighs 50 tons, so performance per kg is even lower. It could power a cluster of say five balloons, but they would not be able to lift 10 tonnes each.

From the other side, nuclear reactors are probably not optimised for the minimal weight, assuming that just a heat output is required. They are built to match different requirements, so some specially built device may be capable of lifting the balloon.

Lockheed Martin is developing a very promising device.

  • 2
    $\begingroup$ I wonder if these 2 to 4 MW are the sustained value over the flight, or during the relatively short bursts when the burner operates. Because if you observe a flying hot air balloon, the burners are off most of the time. $\endgroup$
    – SF.
    Apr 10, 2017 at 10:04
  • 6
    $\begingroup$ "A typical flight lasting 45 minutes to 1 hour, and consume about 20 to 30 gallons of propane" src. LPG energy density 84,250 BTU/gal. src. 30 gal/hour is ~740KW, but that's including initial filling, not just sustained flight. $\endgroup$
    – SF.
    Apr 10, 2017 at 10:25
  • $\begingroup$ The amount of Pu required (750,000 / 500 = 1500 kg) would still be above that the balloon could easily lift, and the extra power may actually be required for maneuver. $\endgroup$
    – h22
    Apr 10, 2017 at 20:37
  • $\begingroup$ Without any parameters on what containment is needed, there is nothing you can say about if an actual nuclear reaction could power a balloon. A prompt critical reaction can easily produce vast amounts of power (and irradiate everything in it vicinty) that could easily lift (and incinerate) a baloon. However a feasibly balloon that also has a payload of cargo or living beings is likely not an option due to shielding taking up too much mass. $\endgroup$
    – Adwaenyth
    Jun 25, 2019 at 14:54
  • $\begingroup$ Reactor also contains parts like cooling system, moderator, even if we do not put any shielding on it. $\endgroup$
    – h22
    Jun 25, 2019 at 16:30

Probably not from decay heat as others have said, but ignoring the minor matter of the radiation and toxic byproducts it is clearly doable with a fission pile burning sufficiently enriched fuel.

The Soviet nuclear program has had at least one criticality accident with a PU core that ended up in a sort of equilibrium between its thermal expansion and reactivity, so say 10kg of fissile material and a load of heat spreaders to get the heat out into the air, bet you could do it with 50kg or so of machine (But I would not want to be the guy to have to go in and shut it down).

There was a proposal for a Venus probe using a nuclear ram jet that may be applicable, thing is once running there was no good way of stopping it because the decay heat would melt the core.

  • $\begingroup$ Do you have sources about the Soviet program & Venus probe ? I'm interested $\endgroup$
    – Skwiggs
    Jul 23, 2018 at 14:06
  • $\begingroup$ I agree, and I think that the weight could be quite a bit less than 50kg. Critical mass is as low as 10kg for some fuels. And graphite foam is extremely lightweight yet has extremely high thermal conductivity, so an effective heat spreader might only add a couple of kilos. Besides the fuel and the heat spreader I don't see that much other equipment is necessary, beyond a means to control the reaction rate, probably a simple neutron reflector. If unmanned, then no shielding necessary. If manned, one would simply need to have very long tethers between the balloon and the gondola. $\endgroup$
    – barneypitt
    Apr 1, 2019 at 13:55

The problem with a Pu-238 RTG is that is heavy and low-power. That is to say, the energy is released too slowly.

This is not a fundamental problem. Phosphorus-32 is (as the name indicates) almost 8 times lighter, and it has a half-life of just 14 days. This makes it about 15.000 times more powerful per gram as Pu-238 - 8kW/gram !

Typical burners are 3MW, but run at 20% duty cycle for an average power budget of 600kW. This means you only need 75 grams. You can't realistically have 3MW peak power, as this is not a controllable heat source. And using 375 grams for 3MW sustained would mean that you need to get rid of 2400 kW of power on average.

Practically speaking, this means that you need a heat buffer. This isn't a big issue, as this heat buffer can surround the P-32 and act as shielding - the absorbed radiation would directly be converted to heat. You'd want a salt melting near 373K, as that's the temperature you're aiming for. Such salts do exist, so that's not a big problem.

Technically, this looks feasible. But the practical problem is that your 3MW heat source is a major hassle. It has to be made in a nuclear reactor, and comes out hot - literally. Handling it is a major pain. You need to permanently cool it. That's just not going to be cheap.

  • $\begingroup$ Using a short-lived isotope, even if we ignore the whole million of technical problems, sort of defeats the whole point of being nuclear - at that point you're getting better endurance out of a low-density fossil fuel like methane. $\endgroup$
    – Therac
    Jun 28, 2019 at 1:48

It is unlikely that a balloon could use Pu-238 to heat enough air in the envelope to get airborne since Pu-238 is not powerful enough for that. What would come a little bit closer is some much more powerful radioisotope such as Cobalt-60 for example and that one will produce more than 13 KW/kg through its first years of decay.

  • $\begingroup$ Welcome to Av.SE. While this is interesting discussion, I don't see an answer to the stated question here. This being a Q&A site, we require "answer" posts to do that. Can you expand on these thoughts to give a conclusion, that a nuclear hot air ballon is or isn't hypothetically possible? $\endgroup$
    – Ralph J
    Jun 25, 2019 at 6:04

You must log in to answer this question.

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