Making the balloon

This Easter my son and I built a hydrogen reconnaissance balloon for our family Easter Challenge. The full story, and aerial photos are here. On this page, I’ll describe how we got to our eventual solution.

Hydrogen Production

So you want to make a balloon? You could just do what any sane person would do, and go out and buy a canister of helium. But helium is expensive: around 100 pounds for a canister with the quantity we would need. And there’s not much of a learning experience in just buying what you need. Much more fun would be to use hydrogen, and to produce it ourselves.

Now you’ve all heard hydrogen explodes, and you’ve seen the film of the Hindenberg, so surely this is going to be dangerous? Well, yes. But if you’re careful, you shouldn’t end up blowing yourself to pieces. If you look again at that film of the Hindenberg, you’ll see it doesn’t explode; instead it burns pretty rapidly. Hydrogen and oxygen, mixed will explode. But a large bag of hydrogen will burn quickly. So, we want to be careful to avoid oxygen mixing with the hydrogen inside our balloon. Of course a large fireball in your kitchen wouldn’t be good either, so use common sense. However, hydrogen is in some ways pretty benign to work with. It diffuses away very rapidly, so unless you’re working in an enclosed space, there’s no risk of a build up of hydrogen causing an explosion in your work area. In many ways this makes it safer than petrol to work with.

How then will we produce hydrogen? There are many ways, but the three I looked into were:

  • Electrolysis of water.
  • Zinc and Hydrochloric Acid.
  • Aluminium and Sodium Hydroxide.

Electrolysis is great if you want a hydrogen/oxygen mixture, suitable for powering rockets or egg cannons. But if you want rapid production of large quantities of hydrogen, it’s a bit slow, and it’s less efficient if you need to separate the electrodes so you can collect just the hydrogen.

HCl and Zinc is fast and simple, and the ingredients are easy to obtain. But we need quite a lot of hydrogen, and zinc chloride is somewhat toxic, which would leave us with a waste disposal problem.

Aluminium and Sodium Hydroxide (caustic soda) is also pretty fast, and the ingredients are easy to obtain. You can get pure caustic soda from pretty much any hardware store to use as drain cleaner. It’s nasty stuff, but this is because it is very corrosive rather than toxic. And the end result, aluminium hydroxide, is used in water puritication, so shouldn’t present a waste disposal problem.

Experiment 1: Hydrogen Production

If you mix half a bottle (350ml) of water and 50 grams or so of sodium hydroxide in a wine bottle, you get a fairly strong sodium hydroxide solution. This is quite nasty stuff. If you get it on your hands, it will start dissolving your skin. You get a strange soapy feeling, as it turns the fats in your skin into soap, but it doesn’t hurt because it dissolves the nerve endings too. So be careful, wear gloves, and keep a good water supply nearby to wash yourself thoroughly if you need to. Above all else, wear goggles. Getting sodium hydroxide in your eyes will likely blind you, and that would spoil the fun.

When you mix the sodium hydroxide, add it slowly and let it dissolve before adding more. If you add it too fast, it forms lumps at the bottom, and these are no use to you. As it dissolves, the solution gets warm. It’s going to get a whole lot hotter in a minute.

If you scrunch up a piece of kitchen foil and drop it in, you get vigorous bubbling as hydrogen is produced. Put a party balloon over the top, and collect the hydrogen. Tie it off tightly, and watch. The balloon deflates pretty rapidly. Hydrogen atoms are so small they dissolve through rubber balloons. If you’re quick though, you can fill a balloon and light it. Bang! Hopefully you’ve still got your eyebrows. It exploded partly because you didn’t flush out the air from the system, so you got an air/hydrogen explosion. And partly because it was under pressure in the balloon, so when the balloon burst the hydrogen was expelled and managed to mix with some external air as it exploded. We’re going to make a very large balloon. We really don’t want oxygen in there, and we don’t want it to be under pressure.

Oh, and the wine bottle got really hot around the neck. Mine got so hot, I got a large blister on my thumb. This reaction is strongly exothermic. We’re going to have to do something about that heat. And the foil was used up in a minute or two, so we’re going to have to sustain the reaction longer.

Experiment 2: Scaling up Hydrogen Production

The wine bottle isn’t going to produce enough hydrogen. It’s fine for small amounts, but if you try and ramp up production by adding a lot of aluminium, you get a reaction that is so vigorous you get sodium
hydroxide foam spraying out of the top of the bottle. With my second test balloon, this happened, and first I got a lot of sodium hydroxide inside the balloon, but then the pressure got great enough that the
hose popped out of the balloon and sprayed boiling sodium hydroxide solution around in a great arc. You did get the message about the goggles didn’t you? And probably best to keep a bottle of vinegar or some other weak acid around to neutralize any spills.

To reduce the chance of this happening, we need a bigger container. I used a demijohn that was intended for winemaking. The one I bought has a safety sticker on it saying “Do not expose to temperatures over 40C”. I never did believe these warning labels: we’re about to heat it a whole lot hotter than that. It’s not the absolute temperature that matters, but that we don’t have any sharp temperature gradients that will stress the glass.

In the demijohn, I mixed about 1.5 litres of water with 200 grams of sodium hydroxide. I put the demijohn in a bucket of water, and filled it so that the demijohn just about floats. This will help cool it once we add the aluminium.

To fill a balloon, I took a rubber bung (from the winemaking supplies), and drilled the hole out to half an inch. Then put some plastic piping (again from the winemaking supplies section of the hardware shop) through the hole and glued it with Araldite (two part epoxy glue). Araldite is pretty good at coping with extremes of temperature. Check the plastic pipe won’t melt either by pouring some boiling water over it – it should soften, but not melt. It’s going to get that hot when we’re making hydrogen.

Test balloon #2 was a black plastic big bag, with the top scrunched closed around a small piece of the plastic tubing. Bluetack will serve as a stopper.

For aluminium, I tried Coke cans, cut into pieces and blasted with a blowtorch to take off the plastic lining, then bashed into shapes narrow enough to fit through the top of the demijohn. These worked pretty well, but you’re going to need quite a lot of aluminium, and I resorted again to kitchen foil once the cans had dissolved.

Problem: the reaction generates so much heat that even with all the extra space in the demijohn, you still get a load of water vapour, steam, and some sodium hydroxide foam up the plastic piping into the balloon.

Experiment 3: The Condensor

The demijohn can produce enough hydrogen, but the gas is too hot and carries too much water vapour. What we need is a condensor and a trap to collect any foam that makes it up the pipes. I added a second container, made from a stainless steel coffee holder pot that had a nicely sealing plastic lid with a clamp-down top. The tube from the demijohn reaches just above the bottom of the pot, and is covered with about 3cm of water. The outflow tube comes out at the top of the pot, well above the water level. The hydrogen then bubbles through the water, cools, and water is condensed out. This entire pot is then immersed into a bowl of cold water, and this water is changed regularly as it heats up.

Finally, we can produce hydrogen at a fast enough rate without spewing caustic foam everywhere. During the filling of the balloon, about an extra 4cm of water condensed out into the condensor pot.


Making the Envelope

As we discovered with the party balloons, containing hydrogen isn’t easy. It will diffuse through latex, so we need a different approach. There’s a reason helium balloons you buy at the fairground are silvery. They’re made from silvered mylar, and this contains helium pretty well. Would it contain hydrogen? And where can we get enough of it?

Last year I found myself having to camp out rough, just below the snowline at night on Mt Fuji. We had some space blankets with us, and they were very useful for keeping warm. A space blanket is made from one large piece of mylar (220cm x 140cm according to the packaging), and it’s very light (about 50 grams). Could I make it into a balloon that was airtight enough to hold hydrogen?

The first step was it fold it in half, and apply a layer of double-sided sticky tape around the edge. Go slightly past half way, and overlap the tape on the corners. Try really hard to get no wrinkles, though you won’t succeed completely unless you’re better at it than me.


For the filler tube, cut a 2cm length of the plastic tubing, and make sure there’s tape applied both above and below it.

Finally, take the backing off the tape and stick the two sides of the balloon together.

Now, I didn’t completely avoid all wrinkles, and they might provide an escape route for the hydrogen. So, I also taped over the edges of the seams with a second layer of regular sticky tape. What seemed to work best was cutting a 30cm length of tape, and laying it sticky-side up on the carpet. Then stretch out the mylar and place it half-overlapping the tape. Finally fold the tape over the edge, sealing it. Probably best to overlap the corners slightly, as they’re a likely source of leaks.


Lift Calculations

Just how much hydrogen do we need? The density of air at standard temperature and pressure is about 1.2kg per cubic metre. The density of hydrogen is about 0.1kg per cubic metre. So each cubic metre of
hydrogen can lift one kilogram. There’s a reason airships are big. My wife’s Canon Ixus digital camera weighs about 200 grams. 100 metres of good quality kite string weighs about 50 grams. We need to
factor in the mass of the balloon envelope and the mass of any remote control for the camera. So, we’re looking at about 400 grams of payload, if we’re lucky, and perhaps more. And we need it to actually lift, or the wind will just blow the balloon horizontally, so we’d better add some extra margin. We’re looking at about 500 litres of hydrogen.

A space blanket, folded over is 140cm x 105cm. So is that enough?

140cm x 105cm x 2 gives a surface area of 29400cm2. A sphere containing 500 litres would contain 500000cm3. Volume of a sphere = 4/3 Pi * r3, so the 500 litre sphere would have a radius of 49cm. Area of a sphere is 4*Pi*r2. So a 49cm radius sphere has a surface area of 30200cm2. So if my space blanket envelope were a perfect sphere, it would be just about spot on for volume to lift the mass I need to lift. Trouble is it’s not a sphere, perfect or otherwise. It’s a pillow shape, which won’t hold so much.

Looks like I’m going to need two space blankets to lift the camera. That adds an extra 100g or so, but after the first 100 litres, any extra will lift payload. And with two lift bladders I don’t need to fill each so full, so leakage should be slower. I could tape both together into a single envelope – that would give more volume, but it would be nearly two metres long – hard to fill out of the wind in my father’s garage. It seems I’m learning about airships the hard way: very little lifting capacity, very large envelopes, and you really have to try very hard to minimize weight.

So, two space blankets it will be.

Camera Payload

The payload is a Canon Ixus 7 megapixel digital camera. It’s small, fairly light, and has reasonable optics. To protect it against bumps and bangs, I made a harness for it from polystyrene foam. This is held together with tying wire, so I can extract the camera to recharge it and extract the memory card.

To control the camera, I first switched off the auto shut off feature. Once it’s turned on, the camera will stay on until the battery runs flat. Then I bought a cheap four channel remote control from Ebay. It’s intended for radio control planes or helicopters I think, and is very light. The receiver can control two DC motors and two servo motors. I only used the servo outputs. One servo was set up to press the shutter release. The second was set up to allow the camera package to pan, as I can’t turn the balloon once it’s airborne. The servo only gives about 120 degrees of pan, but that’s still enough to be very useful. The power supply for the receiver is a Lithium Polymer cell, borrowed from my son’s radio control helicopter. This is pretty light, and has good capacity for the light duty it faces here.

The camera suspends from the plastic bar under the top servo, so it faces about 45 degrees down. Suspending the camera rather than firmly mounting it makes it less likely we’ll damage the shaft of the servo, which is a potential weak point.

All in all, the camera and remote control package weighs 300 grams. Can we create enough hydrogen to lift this?



Final Test

It’s the day belore the challenge, and we’ve not even inflated one envelope. Time for a test. I stocked up on foil plates (the foil is a bit thicker than kitchen foil). Lets see if this works.

The condensor did its job perfectly. Every so often I had to unstopper the demijohn and add more aluminium, but if I did it while the previous lot was still foaming, the strong outflow of hydrogen is more than enough to prevent any significant amount of air from entering. It took about an hour to fill the envelope about 3/4 full. This seems like it should be sufficient. The envelope is lifting strongly while it’s being filled, and needs holding down.

How much will it lift? My screwdriver weighs about 200 grams, and the envelope lifts it, though with not much to spare. The test is a success; two envelopes should be enough to lift the balloon including
the camera payload.


I realised it was going to take too long to fill two envelopes, so I rushed back to the shops and bought enough stuff to make a second hydrogen production rig. We’ll fill both envelopes simultaneously.

The Moment of Truth

We hadn’t had time for a complete test. The camera had never been fitted to a balloon. The second hydrogen production rig hadn’t been tested. And we didn’t even know the range of the remote control. But we were out of time; it was time to pack and head for the family get together at my father’s house. Either it would all work first time, or it wouldn’t work at all.

We got the hydrogen rigs set up, strung the balloons out, and started production. The new production rig wasn’t working – the condensor was leaking hydrogen. Well that was easily cured. I rested a couple of half bricks on top, and that sealed it. To make sure, we filled its bucket to immerse the condensor, and saw no bubbles were escaping. It looks good now.

Over about an hour, both envelopes inflated nicely. One of them still got a bit too much water vapour inside, as evidenced by occasional drops coming out alongside the filler tube, but not enough to be a big deal. Hopefully it won’t dissolve the adhesive in the tape sealing the balloon.

But would it fly? There was only one way to find out: strap the camera harness on.

Yessss!!! It flys well. It looks like the calculations were about right; two envelopes gives us lift to spare. Unfortunately one of the envelopes is dripping a little water right onto the camera. Not much we can do about that, except hope Canon did a good job. If I do this again, the filler tube will not be directly over the payload.

In the end, the whole rig worked flawlessly. It was fairly easy to point the camera in the right direction, just by observing which way it was pointing from the ground. With hindsight, we should have inflated the envelopes slightly more, to get a little more lift. This would have countered the light wind better. But just before dusk and early in the morning, when the winds were very still, it floated high in the sky, and I finally got all 100 metres of kite string played out.

The balloon exceeded our expectations. Over 48 hours, it showed no significant signs of losing hydrogen, and we made five flights getting pictures of my father’s house from all directions, plus pictures of most of the neighbouring houses. The two eight year old kids took almost all the pictures, and did a great job. They took about 800 pictures in total, of while about half were blurred, but the other half were usable. All in all, a fun weekend, and its not just the kids who have learned a lot.