Explosive Mixtures

by Chris McMahon

I was reading Matthew Reilly’s Ice Station recently – a great book with lots of action – but one thing really had me gritting my teeth. At one point in the story the tanks containing the station’s refrigerant (to keep the walls frozen) are ruptured. In the book, this flammable gas hung around, making local fireballs every time someone had to fire a gun (and incinerating them as well). It was a convenient plot device so that the gunplay was replaced with hand-to-hand combat and scenes with little hand-held crossbows, but not realistic in my book – hence this post.

There were a couple of things wrong here. In Reilly’s scenario, the shooters who triggered the explosive gases were at the lower levels of the station. So what is a released refrigerant gas doing lurking down there? I have not looked up the specs of the gas he mentions, but refrigerants are notoriously volatile for a start – the whole reason they are used in refrigerant cycles in the first place. The released refrigerant would have immediately gone UP!

The second thing is – if the muzzle flare was capable of igniting the gas, it would have triggered a continuous explosion. The flame would have raced through the explosive mixture in all directions, creating a pressure wave from the rapidly expanding combustion gases. Basically at the first ignition, there would have been an explosion at the TOP of the facility that would have probably blown the roof off. Then the flammable gas would have been combusted – i.e. gone.

That’s if the gas actually ignited in the first place.

All flammable gases have an explosive range in air. Too little of the gas and it will not ignite – too MUCH of the gas and it will not ignite either (generally the lowest ignition energy is required in the middle of this range). These are called the Lower Explosive Limit and Upper Explosive Limit respectively. These are usually expressed as a percentage concentration of the gas in air. For example – methane (the major constituent of natural gas) – has a LEL of ~5% and a UEL of ~16%. A fairly narrow range. I also grit my teeth in all those Hollywood films where someone sets the gas on the stove and flicks a match behind them. If they do this immediately the methane concentration is unlikely to be 5%. If they wait too long no amount of flame will create an explosion at all. The one scenario I will credit is where an open flame is left burning with the gas filling the room gradually. In this case it will explode when it reaches around 5%.

Now consider the explosive range of a gas like Hydrogen – notorious for being dangerous to handle. This has an explosive range from 4 to 75% in air. And people want to have Hydrogen bowsers at the fuel station! Hydrogen also has a very high flame speed – which means a more intense pressure wave.

Acetylene (used in welding) is even more fun – with an explosive range of 2.5 to 82% in air and an autoignition temperature of only 305C (Hydrogen and Methane need temperatures approaching 600C).

Powders can also be explosive.

When dealing with powders, it’s the Minimum Ignition Energy that is critical. Generally if a powder (or vapour) has a MIE lower than 30 mJ (milli-Joules), then a person needs to be earthed to an industrial safety system to deal with it.

If you thought paracetamol could cure headache – it can cause them too. Here are some examples of the MIE of powders:

  • Paracetamol <10 mJ (this is produced as powder than pressed into tablets)
  • Magnesium 20 mJ
  • Sugar 30 mJ
  • Wheat Flour 50 mJ
  • Zinc 200 mJ
  • PVC 1500 mJ

Suger dust, magnesium and paracetamol powders can all be very explosive. Of course the MIE  of powders also depends on the particle size.


  1. Chris => putting the science back in Science Fiction.

    Okay, now I’m curious. Why build with ice if you need a refrigerant to keep it from melting? Or was it only the inside that was warm? And why have the inside so warm ithat insulation was insufficient to keep the walls up?

    1. Hey, you need to ask Matthew Reilly that one:) The station was dug into ice, which I guess is strong and insulates – if it stays solid – hence the refrigeration. The station was fictional, how close it relates to reality I’m not sure.

  2. The other complaints are at least somewhat valid, particularly the one about a continuous explosion, but your first point is in error.

    “Volatile” does not equal “lighter than air”. Many gases evaporate rapidly from their liquid form and diffuse quickly, but are heavier than air. Gases which are complex compounds, like refrigerants, tend to fall in this category. The effect is exploited in the electronics industry, where Freon is heated to solder-melting temperature in an open tank and circuit boards are dipped into it to melt all the joints at once.

    R-22, the most common refrigerant for fixed air conditioners, is enough heavier than air that it is an asphyxiation hazard. It can pool in low areas and displace the oxygen.

    It is somewhat improbable that anyone would use a flammable refrigerant unless it is firmly established that cash is short. The whole point of Freon is that the compounds are petroleum gases which have been “pre-burned” by incorporating fluorine or chlorine, which are stronger oxidants than oxygen. They aren’t flammable because oxygen can’t displace the strong oxidant. Flammable petroleum gases are cheaper because they require less processing to produce them.


    1. I was wondering about that – hence my comment that I had not checked the specs on the gas. I was aware of that – it’s down to molecular weight which correlates to density if you assume it’s an ideal gas. It seems weird to have a flammable refrigerant – he must have dug deep for that one.

      Damn you! Now I’ll have to check it next time.

      1. OK. I checked the book – he does not say what the refrigerant is other than its a chloroflurocarbon – if I’d remembered that I would have realised straight away it was heavier than air. I checked quickly for a few of them – Freon 11 has a MW of ~137 and Freon 113 a MW of ~187. Both would be much heavier than air.

      2. Which introduces another, more severe problem: a flammable CFC is wildly implausible, and for me, anyway, would result in a Book Meets Wall moment. Both chlorine and fluorine are much stronger oxidizers than oxygen is, so when oxygen and flame come along the CFC just shrugs and sits there.

        There are two types of practical refrigeration systems: absorption and compression-expansion. Absorption systems depend on the fact that there are substances whose mixtures are an endothermic reaction, that is, when you mix them they get cold. Ammonia/water is an example. Compression-expansion systems depend on absorbing or rejecting heat at phase changes — liquid to gas cools, gas to liquid heats (and requires energy input in the form of mechanical compression).

        Early experiments with compression-expansion used petroleum gases because they have phase changes at useful temperatures and pressures. Probably the very best refrigerant for human-useful temperatures is propane. The obvious dangers of that approach led to the development of Freon, the whole point of which is that it has the useful phase change temperatures and pressures without the flammability. R-19, which used to be used in cars (and still is, in lands where efficiency is more important than political correctness) is, for all practical purposes, propane with chlorine and fluorine added to keep it from exploding.


    1. Yes, but ammonia isn’t used in compression-expansion systems like your stereotype air conditioner.

      When ammonia mixes with water it undergoes an endothermic reaction, that is, it gets cold. It is then possible to distill the mixture, separating the ammonia from the water for another round. The interesting thing about that is that clever design can yield a system driven entirely by the energy from the heat source used for the distillation phase. That’s the so-called “Servel” refrigeration system (after the company that popularized it). Refrigerators that ran on the Servel cycle, powered by a gas flame, used to be common, and still exist for things like RVs where electrical power to run a compressor isn’t available.


Comments are closed.