I was thinking about ways of justifying practical balloon flight for a pre-industrial civilization, when I remembered that some microorganisms produce methane, and better yet, hydrogen.

This made me wonder, is it plausible to posit a hydrogen creating microbe that could produce enough lift gas to refuel a balloon, even in midflight?


A balloon craft in this scheme would keep a lightweight fuel tank attached to the envelope by a hose. The tank would be filled with microbes and their food (hydrocarbons, biomass, whatever it is that they like to eat). The microbes then produce hydrogen gas. To rise, you throw in more fuel. To descend, you release some hydrogen. Depending on the fuel source that the microbes eat, trading towns might also farm wells full of the microbes where smaller balloons could fill up.

Now on to the potential problems:

First, I don't know if any organism could realistically produce the required volume of hydrogen gas. I know that a hydrogen balloon lifts about 68 lbs per 1000 cubic feet. I also know that the efficiency of methanogensis (which is the best example I've got for hydrogenesis) can be quite high, and I've seen numbers ranging from 20% to 80%. But even with some kind of awesomely efficient microbe, from there I have no idea how to calculate the cubic feet of hydrogen that could be derived from a given fuel source.

Second, though less important, I also don't know if any microbe could physically generate gas fast enough to allow for reasonably controlled changes in altitude.

Third, even if all of that (literally) flies, I need to make sure there's a reason these microbes don't get loose and just eat up all the fuel outside the tank.

Finally, this question may end up being pretty important to the world I'm working on right now, so if it ends up being more complex than a straightforward "no" then I'll also offer a thank-you bounty to anyone that gives me an especially in-depth answer.

BOUNTY EDIT: Bounty goes to Dubukay. In general this question received many high-quality answers, so thanks everyone!

  • 1
    $\begingroup$ @AlexP: I love hot air balloons, but I've been having a heck of a time figuring out how to make one that can fly for long periods without advanced liquid fuels. The early hot air balloons I've read about typically used ground fires and descended within a few hours. I'm looking to establish traders for lightweight luxury goods, so I'd like them to be able to stay in the air longer. $\endgroup$
    – Random
    Feb 1, 2018 at 17:36
  • 2
    $\begingroup$ Use that fuel tank to hold purified nafta (also known as paraffin or kerosene) and burn it in a paraffin lamp. This is more or less how modern hot air balloonists do it. The ancients knew of petroleum; distillation was known since the late antiquity; just combine the two. $\endgroup$
    – AlexP
    Feb 1, 2018 at 17:39
  • 1
    $\begingroup$ Intuiting "no." Methanogens grow slowly (some clock in at a few hours doubling time, but most are the order of months.)The physiology (CH4 production) will reflect this. I think it is a simple mass balance. You need substrate, say ... acetate? and as you point out some fraction of that goes into their biomass. They don't grow 'in air' so you'd need a culture (=some H2O = weight.) Hydrogenesis... I don't think this helps. I suppose it depends on the buoyancy of CH4 but instinct says no. Not certain, so not answering, but ... I'll be surprised if you find a hard science way to accomplish this. $\endgroup$
    – SFWriter
    Feb 1, 2018 at 17:45
  • 5
    $\begingroup$ @AlexP While the idea of literally burning a free trade agreement sounds like it has some appeal to a great many people, and certainly gave me a laugh, I think you meant naphtha? $\endgroup$
    – bendl
    Feb 1, 2018 at 18:57
  • 1
    $\begingroup$ "better yet, hydrogen" - better? More buoyant maybe. But better? Oh the humanity! $\endgroup$
    – corsiKa
    Feb 1, 2018 at 21:19

5 Answers 5


Microorganism-powered aerial flight is plausible, but with caveats.

As you point out in your question, hydrogen is the best lifting gas we have. So this becomes a question of "What's the most efficient biological process that produces H$_2$?" The answer, of course, is algae.

Normally, algae get their energy from photosynthesis- taking in sunlight, water, and carbon dioxide to produce ATP, complex sugars, and oxygen. However, under the right conditions (mumble mumble sulfur-limitation heterocysts mumble) some algae will switch to a metabolic state of "anaerobic oxygenic photosynthesis". In this state, the oxygen produced by photosynthesis is used by the cell's own respiration, producing an anoxic environment, which in turn triggers the production of hydrogen gas.

What this means is that algae can produce H$_2$ gas almost directly from protons. Even better, we can collect it and are already well on our way to making it a cost-effective replacement for fossil fuels. Clean energy in our lifetime? Yes please.

However, that's not enough to answer the question, which asks about the rate of H$_2$ production. In 2001, a company built a 500 liter bioreactor that could produce an astounding 1 liter of hydrogen gas per hour. With that kind of potential, our balloon would need luck to even start inflating. However, that was 2001, and the first year the company started. At that time, they calculated a theoretical maximum of 20 grams of hydrogen per day- about 10 liters per hour. In 2004, a review came out that posited a maximum of 5.45 kg of H$_2$ per square meter per year. That's a rate of ~7 liters per hour- still a bit too slow. In 2011, we multiplied that rate by 5 times by creating biohybrid photosystems that use platinum nanoparticles. In 2013, we managed to do even better and increase our efficiency 4x by modifying the chlorophyll antennae, and that's since been pushed to 13x. So our current rate of H$_2$ production is about 450 liters per hour! Of course, this is an idealized maximum efficiency and we haven't yet managed it on a large scale.

So what does that mean for our balloon? In a world where people rely on balloons like this, I'm going to assume that they're operating pretty close to maximum efficiency, perhaps 400 liters per hour per square meter. Of course, there may be problems with this, but it's a decent estimate to start with. From skydrifters.com, we learn that an average hot air balloon weighs 800 pounds. Since we're traveling and trading, let's call it 500 kg total. Lifting 500 kg with hydrogen gas requires ~40 kg H$_2$. At normal air pressure and temperature, this occupies a volume of 450,000 liters. Thus, our balloon will take approximately 41 days to fill under its own power. That's going to be hard to pull off.

However, this calculation shows that algae can indeed produce enough gas to lift a balloon, and it'd certainly be an eco-friendly way to travel. It also allows for maneuverability in the air, and essentially permanent air travel. Once the balloon goes up, the algae can draw their CO$_2$ directly from the air and the balloon will be powered by light alone. Additionally, it's quite possible that towns and cities would start to farm fields of algae so that filling up at a city is quite easy and would entice ballooners to visit.

In the air, such a balloon would ascend normally as the algae produce hydrogen gas. Additionally, H$_2$ gas compresses quite nicely, and it might make sense for ballooners to keep a compressor on board to capture any excess H$_2$ produced for a quick burst if necessary. Descending is the easy part: the simplest would be venting the H$_2$ or compressing it for later. You could also spray the inside of the balloon with some kind of inorganic sulfur, which would shut down the hijacked photosynthesis pathways temporarily, or add oxygen, which would destroy some of the hydrogenase enzyme.

The mental image I have of this system is a very gross, quite large clear balloon. The outer membrane would be made of plastic wrap or some other impermeable lightweight clear solid, and there would be layers of algae directly inside. Any H$_2$ produced would fill the balloon and contribute to lift, displacing any denser air in the meantime. Since the microbes live best in an anoxic environment, this wouldn't even be a problem. Maintenence would essentially consist of replacing nutrients and removing dead cells from the inside, which would probably be done when on the ground but could be done in the air if one can hold their breath long enough.

  • 1
    $\begingroup$ This is great stuff! With the addition of the mentioned numbers I will definitely contribute a bounty as thanks. $\endgroup$
    – Random
    Feb 1, 2018 at 20:48
  • 2
    $\begingroup$ mumble mumble pre-industrial civilization mumble... $\endgroup$
    – M.Herzkamp
    Feb 2, 2018 at 12:28

So the first question. If the microbe doesn't have to exist on Earth already, then there is no reason for it to not be possible.
Just say that this super microbe does exist, and when combined with other methods like a catalyst of some kind (like nickle for instance), is able to produce enough gas to make it possible.

Another possibility is that the organism works slowly, but the gas builds up over time. So say to fill a new balloon up to pressure takes a couple weeks, but if you keep putting in resources then the microbe will just keep producing and keep the balloon topped off. This option would require a material that holds the gas really well.

Second, there are other possible ways to control changes in altitude other than venting. If you compress the lifting gas then it will be less buoyant, and so you will go down. you could put a balloon inside of the balloon, and pump the inner balloon up with air when you want to descend.

As to the third question, this is a bit trickier. As a wise man once said "Life, uh, finds a way."

There are several self limiting mechanisms that could be used.

  • The microbe has a natural microbe predator that keeps it from getting out of control in the wild. It could be another microbe, some kind of algae, etc.
  • There is a compound in raw fuel that hinders the microbe that isn't present in refined fuel or something is added to the refined fuel that gives the microbe a boost. This could also act as a catalyst.
  • The microbe has a genetically fixed colony size that is large enough to be used in a gas production system, but not large enough to cause problems in the wild.

I don't think so.

This starts to be a rocket problem. The more gas you need, the more microbes and nutrients you need for the microbes and that adds mass. To lift that mass, you need to produce more lift gas which requires more microbes and nutrients, etc. If you can't pass break even, you aren't going anywhere.

Also, microbes want to grow and reproduce. They evolve to be as efficient as possible into turning nutrients/energy into biomass with as little waste as possible. In this case, the lift gas is a waste product. While it might be possible to engineer a microbe to produce more gas than biomass, I don't think you'll find them in nature.

Furthermore, you would have to get that lift gas away from the microbes. Very few organisms can survive while immersed in their own waste products. At some point, the partial pressure of the gas in the air around the microbe is going to be high enough that the microbe cannot excrete the gas and will die.

The microbe will also have to have access to gasses that it can take in. I hope that the microbes are not oxygen breathers since oxygen and most biologically produced lift gasses do not mix well.

For this to work at all, I think that you would need a thin sheet of the microbe with air and nutrients on one side of the sheet and the lift gas collection on the other side. You could make a bag out of this sheet if you can make the sheet strong enough. Though, an engineered multi celled organism in the correct shape would work better than trying to hold microbes to a shape.

  • 4
    $\begingroup$ Quite right - it's a rocket problem. Simple electrolysis is easier, efficient, safe, already-developed, proven...and the weight of water is still too heavy to bother. $\endgroup$
    – user535733
    Feb 1, 2018 at 19:14
  • 3
    $\begingroup$ It's not quite a rocket problem, if only because you can do most of the envelope filling on the ground, and only need to carry enough nutrient to keep the envelope topped off, and you don't need to bring more microbes because the ones you start with can keep working through the new fuel as it's added. Rockets fall down if they run out of fuel. Balloons stay up until the existing gas bleeds out, which can take a long time under some conditions. $\endgroup$
    – AndyD273
    Feb 1, 2018 at 20:07
  • 1
    $\begingroup$ And oxygen isn't a problem because they can be anaerobic, just living in the petroleum sludge without needing or being bothered by oxygen or the lack thereof. The envelope would likely have some small amount of leakage, which is a way to keep the pressure manageable as a silver lining. $\endgroup$
    – AndyD273
    Feb 1, 2018 at 20:07
  • 2
    $\begingroup$ It is NOT the rocket problem. The rocket problem is about carrying propellant increasing the amount of propellant you need for the same performance. In this case the hydrogen for the lift is generated on the ground before the lift and you only need to carry enough to get the endurance you want. Which I think scales with desired endurance and square of the cubic root of the lift gas volume? Much better scaling than with rockets. $\endgroup$ Feb 2, 2018 at 8:22
  • 1
    $\begingroup$ Also, hydrogen is lighter than air: The microbes would be below the gas bag and gravity will remove produced hydrogen from them. It would look and work like hot air balloons do. $\endgroup$ Feb 2, 2018 at 8:26


Very roughly, hydrogen gas H2 is 2g per standard volume (22.4l), while air being 80% N2 (28g/mol) + 20% O2 (32g/mol) gives an average of 28.8 grams. So two grams of hydrogen, displacing 28.8 grams of air, generate 26.2 g of lift (probably a little less due to the balloon being somewhat compressed). Each gram of hydrogen gains us 13g of lift.

How do we get those 2g of hydrogen? We need some highly idrogenated feedstock, so a molecule with hydrogen bonded with the lightest possible elements and containing excess chemical energy.

Available light elements include:

  • lithium (lithium hydride "burns" to hydrogen by itself, just add water, no bacteria needed - LiH + H2O → LiOH + H2; we still need one water molecule and one LiH molecule for every available hydrogen molecule, which requires one oxygen (weight 16) and one lithiums (weight 7), for a total ratio of 2:23 or 8.7%)
  • beryllium (beryllium hydride, synthesised 1951. Let's not go there).
  • boron (hydrogen boranes. Doable, but a little too energetic).
  • carbon. Very promising: not only it binds hydrogen but binds to itself in stable compounds.
  • oxygen. This means water; not much energy there.
  • nitrogen. This means ammonia; but oxidising it results in nitric acid, not hydrogen gas. Dissociation problems: like water, but worse.
  • aluminum. Not unlike boranes for violence, and we need water.
  • fluorine. Same problems as ammonia but much, much worse.
  • sodium. Costly and unwieldy, and difficult to handle. Weight ratio exactly half of lithium (not coincidental: both lithium and sodium are group I), 4.34%. Going up in the periodic table will only make things worse.

The best option is saturated hydrocarbons. We need a metabolic pathway through which the bacteria dissociate CnH2n+2 hydrocarbon and oxidise the carbon, but not the hydrogen. There's energy enough available in hydrocarbons that we're not shortchanging the little critters:

  • CnH2n+2 + nO2 → nCO2 + (n+1)H2.

The hydrogen proportion in weight in CnH2n+2 is about 14%; so, one kilogram of feedstock will yield 0.14 kg of hydrogen, supplying a lift of 1.82 kg. Since we also get rid of one kg of feedstock that acted as ballast, the total lifting effect is 2.82kg.

Not much, really, and I think that's the best one can possibly do. But perhaps it is enough.

Gengineered algae

Another possibility is a pseudo-photosynthetic organism that harvests water from the atmosphere, photodissociates it and releases H2 and O2. But such an organism wouldn't have any advantage in doing so (it could do so once fully matured and stable, though), because the energy would entirely go in "waste" products, and the production rate would be even less than in the first case (the inbound sunlight must equal the chemical energy being stored in the dissociated gases, and for the hydroxygen mixture, that's a lot of energy, while sunlight is I think around 1.2 kW per square meter).

"a 100%-efficient electrolyser would consume 39.4 kilowatt-hours per kilogram (142 MJ/kg) of hydrogen" (Wikipedia), so we can expect about 30.4 grams of hydrogen per hour from every square meter of algal panel, or about 0.4 kg of lift per m2 per hour. Probably much less because green and blue algae don't absorb all of the energy of the whole solar spectrum. Considering the weight of an algal panel (which needs water and support), this probably means that this isn't a very promising road. Or is it?

Let's go down it anyway. The Hindenburg might have had about 9000 m2 available, producing 270,000 grams of H2 per hour in full sunlight. That is about three million liters per hour, or 3024 m3 of hydrogen per hour. The same Hindenburg required 200,000 m3 of hydrogen; that means that in one hour we can replace about 1.5% of its gas contents, in exchange for a weight of not less than 90 tons (ten kg per m2) or 180,000 lbs of its 511,000 lbs payload. In theory it's doable, but I think we're pushing things; the above values are all calculated from the most optimistic circumstances. A panel weight of 30 kg per square meter (and when you think glass - or hydrogen-proof yet thin and transparent plastics - and water, 30 kg are nearer than it seems) might well make the whole endeavour mathematically impossible.

Larger, flattish balloons might increase the convenience of it all, especially if we could build them with clear hydrogen-proof plastics and put the algae on the inside bottom surface. We still couldn't keep them uncovered (because we need to sequestrate and discard the oxygen they produce). But at that point there are big structural problems, and the fact that we don't really have a sunlight-clear plastic tough enough to withstand the stress, that will not leak hydrogen like a sieve. But this could perhaps be handwaved away ;-)

  • $\begingroup$ I like it. The correct way is to have the algae+fungus+bacteria(+virus) symbiote growing on the outside top of the balloon with a second membrane holding it in place. The top of the balloon is one way (hand wave) permeable to H2 and the generated gas diffuses in. Atmospheric water and O2 and CO2 and pilot Hydrocarbons or honey are dribbled over the outer membrane and diffuse into the symbiote layer. Make it generate so fast that venting is a simple option and added feedstock is fully and rapidly converted. $\endgroup$
    – KalleMP
    Feb 2, 2018 at 8:05

I think "yes" but with a couple changes in perspective:

Firstly think dirigible and not hot air balloon: Don't use your gas to control your altitude, it is too valuable to just vent it. Use impellers and control planes, like a blimp to FLY up and down.

The "throw in more fuel to go up" portion is unlikely... but that doesn't mean they can't generate all the lift necessary OVER TIME...

Which also solves the mass problem... keeping a colony of bacteria on hand that is just large enough to perpetually keep the dirigible topped off, and increase the volume for the sake of the occasional load of heavy cargo...


You must log in to answer this question.

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