I want to electrolyze water to get oxygen and hydrogen. I want to mix them in a regular balloon and ignite it. How much hydrogen would you need in a 2:1 ratio with air for it to be dangerous? Is this generally a safe experiment?

  • $\begingroup$ This great video shows what happens.youtube.com/watch?v=qOTgeeTB_kA Not the explosion that Uncle Al describes. $\endgroup$
    – user15427
    Commented Apr 7, 2015 at 18:30

4 Answers 4


First, let me say that I've enjoyed many times exploding soap bubbles of about one milliliter filled with hydrolysis gas. That is 1 cubic centimeter. That will give you a sound that rings in your ears in a decent sized living room. You may wish to use ear protection for the experiment.

50 ml will have an effect in a lecture hall that not only wakes up everyone, but also may make people complain.

Now while the explosive limits of hydrogen in air range from about 18 -- 60 % the flammable limits are from 4 -- 75 %, in oxygen the limit of flammability goes all the way from 4% to 95% read: for practical purposes, hydrogen in oxygen is always at least a flammable mixture.
For comparison, gasoline in air is flammable roughly between 1.5 - 7%.

However, you start with a stochiometric mixture which is ideal for explosion (after all, that's why you do it, right?), and it is in oxygen, not in air. So even if you "only" have a flammable mixture reaching an ignition source (e.g. electric switch), chances are quite high that as the ignition proceeds through the mixture, it can reach a zone where the gas/air mixture is explosive. Also, because hydrogen is so much lighter than air, it tends to accumulate under the ceiling, so even if there is enough air in the room that an ideal mixture would not be explosive, chances are that there is an explosive layer of gas. (Same with gasoline or, practially more relevant: solvent from glue for parquet floor tiles)

The safety relevant points are:

  • the limits of explosion and flammability are, even in air, extremely wide.
  • limits of flammability are important even if there happens to be no explosive mixture: hydrogen is a small and light molecule that diffuses fast. If the air you breathe happens to have a hydrogen concentration within the flammability limits, any ignition will burn deep into your throat on its way to the lungs.
  • $\begingroup$ I'm actually planning on making very little of this just to get a little sound. I've seen people ignite large party balloons with hydrogen before. Are you saying it could have been a big disaster? $\endgroup$ Commented Feb 15, 2014 at 21:14
  • $\begingroup$ It could, but there is one thing that usually makes the balloon experiment less dangerous (and IMHO somewhat misleading as to what a good explosion is): the hydrogen diffuses really fast through the rubber of the balloon, while the oxgen stays in. As it takes a while to fill the balloon, the resulting mixture is far from ideal. $\endgroup$ Commented Feb 15, 2014 at 21:16
  • $\begingroup$ So you're saying that there is very little hydrogen left, and thus making the explosion less intense? $\endgroup$ Commented Feb 15, 2014 at 21:18
  • $\begingroup$ Yes. I've also had school chemistry lecture hydrolysis gas explosion. After the soap bubble experiment I can tell from the sound that the mixture was not ideal. The trick with the soap bubbles is that this is really fast. We made them 2 - 3 per second with an autogenous welding machine for dental technicians. It uses a canule as tip, and we just put that into soap water and immediately lighted the bubbles. When not bubbling, we kept the welding torch burning - that is the safest way. $\endgroup$ Commented Feb 15, 2014 at 21:19
  • $\begingroup$ Aha, okay. I'm only intending to make a very small amount, anyway. Thank you for pointing this out. $\endgroup$ Commented Feb 15, 2014 at 21:20

Here is a neat experiment, but it unfortunately needs a welder to make the equipment, a metal sphere with a volume of about 100 mL. It should have a very small opening on top and a larger one on the bottom. It also needs feet to stand.

experiment setup

The metal container should be capable of some pressure, because the gas will explode inside and you don't want to hurt anyone. Additional you need bucket or bowl filled with water and hydrogen gas.

First, close the hole on the top of the container and place it in the bucket under water until it is completely full. Now you can fill the container with hydrogen gas until all water is gone from it. When you remove your container, make sure that the bottom hole stays on the bottom. As Hydrogen is much lighter than air, it will be trapped inside.

Place the container on a stable, nonburnable table/ bench. Have a lighter ready. Now open the top hole and ignite the hydrogen gas. Move away from the experiment. After a while you will hear a phenomenal boom (make sure you have your mouth open). Do not try this in a small room.

The trick behind this is in the ratios. As Hydrogen will burn on top, air will be sucked in at the bottom. As long as the mixture has not the right ratio, it will not explode but just burn at the top. Once enough oxygen has entered it will ignite the whole thing and do a phenomenal bang. The experiment is pretty safe.

The experiment is part of this lecture, which is in German unfortunately.

  • $\begingroup$ And this is with only 4 percent of oxygen. I wonder what would happen with a 1 to 2 ratio... $\endgroup$ Commented Mar 26, 2014 at 10:47
  • $\begingroup$ @Friendofkim According to this the 'concentration' of hydrogen in air should be about 59% when it explodes (you start at 100%). So Oxygen level should be about 8%. Nevertheless, if you have a 1:2 mix then cover your ears and keep your mouth open! $\endgroup$ Commented Mar 27, 2014 at 2:41
  • 1
    $\begingroup$ Ahh, I wrote the comment a bit too fast.. I was thinking about the burning percentage. $\endgroup$ Commented Mar 27, 2014 at 12:56
  • $\begingroup$ That experiment was performed at our school (yes, high school) multiple times using what looked like a simple old tin can (cylindric). Hydrogen was simply filled in directly from the bottle through the hose without any of the fancy water stuff. The teachers actually continued letting hydrogen flow through before turning it off and starting the explosion trigger. $\endgroup$
    – Jan
    Commented Feb 12, 2016 at 1:24

Oxygen and acetylene mixed in a ballon will produce a very large explosive bang when lit and a loud noise and gratifying plume of putrid black smoke. Hydrogen and oxygen is going to be many times more explosive, not sure how many times more though.

I was on the back of a flat bed truck with a parade display and a weather ballon on a string about 10 feet above the tray (possibly less was twenty years ago) was accidentally ignited (not by me!) and a few of us were knocked to the floor/truck tray and the closest person had open flesh burns all the way down his arm which was extended in the air towards the weather ballon (may have even been holding a party sized ballon with acetylene as he was lighting them while holding them which is very dangerous and I in no way endorse (he was a lunatic this guy and I had no authority over him)).

So bottom line start very small with these kinds of experiments and be aware things can and will go wrong so always take as much precaution as you can.


The most important point presented above is, in my opinion, the explosive range of hydrogen mixed in air. It is uncommonly and dangerously large!

Real world point is, for example, I would NOT recommend recharging your fuel cell car in your garage! Accumulation of odorless and colorless H2 in a room followed by ignition from a spark from flipping on the light switch is a potential disaster for you, the car and the entire house!

Relatedly, if you enter a room and smell natural gas, do not touch that light switch! Either leave immediately or turn off the power to the house and then open windows and such. Why? Because opening the window impacts temperature which could trigger a thermostat to turn on the heating/cooling system (possible spark or flame).


Not the answer you're looking for? Browse other questions tagged or ask your own question.