Practical Pyrotechnics (Happy Birthday, Good Omens!)

The novel, Good Omens, was first published on 10th May 1990.

Today (10th May*) is the thirtieth anniversary of the release of the book Good Omens, which is an old favourite of mine, and one I’ve found science-based excuses to write about before. In honour of the day, I’m going to do it again—but this time I’m going to talk about fire.

Fire plays an important role in both the book and the acclaimed television adaptation. Of course, fire is rather easier to do in a novel, since reading words like “fire” and “flames” are generally quite safe. In TV land, however, it’s a bit trickier. In particular (spoiler alert), at the start of episode five, the bookshop owned by the angel Aziraphale is burning when Crowley arrives and walks in. Crowley, after all, is a demon. From Hell. Fire can’t hurt him.

Except, of course, he’s actually the lovely David Tennant, who is a very much not-fireproof human being. Which poses a few questions: did the film crew really set the bookshop set on fire? Did they really make David Tennant walk into a burning building? How is that done safely? And what did they actually burn?

It turns out that they did, in fact, burn down the bookshop set. According to The Nice and Accurate Good Omens TV Companion, director Douglas Mackinnon “wanted a real fire” and “there were thousands of books, tapestries and beautiful grandfather clocks inside the shop that were real.”

Actual books were harmed in the making of Good Omens (photo used with permission).

Which… argh. Actual books. In flames. I might be a bit traumatised. Give me a moment.

Anyway. The thing is, if you’ve ever set fire to paper you’ll know it’s not very controllable. You can’t just burn books and achieve consistent and, more importantly, safe, flames. The Good Omens TV Companion goes on to explain that the set was rigged with gas lines and flame bars. It doesn’t say what the fuel was, but the probable candidate is propane.

This is where we get to the chemistry. Propane is a hydrocarbon—a molecule made of hydrogen and carbon atoms—and the “prop” part of its name tells us that it contains three carbon atoms. The “ane” part tells us it’s an alkane, and from that, handily, we can work out its formula without having to do anything so mundane as look it up, because the formulas of alkanes follow a rule: CnH2n+2. In other words, take the number of carbons, multiply it by two, add two, and you get the number of hydrogen atoms. This gives us three carbons and eight hydrogens: C3H8.

Propane’s boiling point is -42 oC, meaning it’s a gas at room temperature. You may be familiar with propane canisters which slosh when moved, suggesting liquid, and that’s because the propane is under pressure. The only real difference between a gas and a liquid is the amount of space between the individual particles. In a liquid, the particles are mostly touching one another, while in a gas there are large spaces between them. If you take a gas and squash it into a small volume, so that the particles are forced to touch, it becomes a liquid.

Propane is stored in pressurised canisters (photo used with permission)

But once the propane is allowed to escape from the confines of a pressurised container, at room temperature, its molecules spread out once again, into a gas.

The expansion is BIG. Theoretically, at room temperature, one litre of propane liquid (with a density of 493 g/litre) will expand to occupy roughly 270 litres of space. But, of course, the space it’s expanding into also contains air, so the volume of flammable mixture—approximately 5% propane to 95% air—is actually much higher.

Gases burn faster than either liquids or gases. We know this, of course: it only takes a brief spark to light the gas burner on the cooker hob, for example, but you’d struggle to light a liquid fuel with the same spark (unless it was warmed, and therefore starting to vaporise). The reason is those big gaps between molecules: each molecule in a gas is free, none are “buried” in the middle of a volume of liquid (or solid), so they can all mingle freely with oxygen (needed for combustion) and they all “feel” the heat source and become excited more easily.

Propane is a hydrocarbon with three carbon atoms.

Apart from being a gas at room temperature, propane is also chemically very safe in that it’s non-toxic and non-carcinogenic. It’s also colourless and odourless—although small amounts of additives such as the eggy-smelling ethyl mercaptan (ethanethiol) are sometimes added as a safety precaution, to make leaks more noticeable.

Mechanically there are more hazards. There’s a significant temperature drop when a pressurised liquid expands into a gas. The simplest way to think about this is to think of temperature as the energy of all the particles in a substance divided by its volume. If the volume increases while the number of particles stays the same, the energy is spread out a lot more, so the temperature drops. Potentially, a sudden release of too much gas near a person could severely chill their skin, and even cause frostbite. Plus, of course, although propane isn’t toxic, if it displaces oxygen it could cause asphyxiation, and it’s heavier than air, so it tends to accumulate in the bottom part of a room—precisely where people are trying to do pesky things like breathe.

Yellow flames, and smoke, are a sign of incomplete combustion (photo used with permission).

Then there’s the issue of complete combustion. Generally, when hydrocarbons burn they produce carbon dioxide and water as products, neither of which are too much of a problem for nearby humans (up to a point). However, when there’s not enough oxygen—say, because the fire is inside a building—other products form, in particular carbon monoxide, which is very toxic, and carbon particles, which make a terrible, terrible mess.

I mentioned earlier that a flammable mixture is about 95% air to 5% propane, and this is why. In fact, it’s even more precise than that: for propane to burn cleanly it should be 4.2% propane to 95.8% air. In industry terminology, if there’s not enough propane it produces a “lean” burn, where flames lift from the burner and tend to go out. If there’s more propane (and thus not enough oxygen) it’s called a “rich” burn, which produces large, yellow flames, soot, and the dreaded carbon monoxide.

They did burn the bookshop. But it’s OKAY, it was restored again at the end! (Photo used with permission.)

You might, of course, want a certain amount of yellow flame and smoke, to achieve the right look, but the whole thing needs to be carefully controlled to make sure no one is in danger. It’s all manageable with the use of properly checked, monitored and maintained equipment, but you can imagine that a big effect like the bookshop fire needs a very experienced professional to oversee everything.

For Good Omens, that was Danny Hargreaves (of Real SFX), who’s worked on all kinds of projects from War of the Worlds to Doctor Who. As he says in the Good Omens TV Companion, “everything is under control [but] we took it right to [the] limit.” At one point, he says, he turned off gas lines sooner rather than later and, when director Douglas Mackinnon asked why, had to explain that the roof was about to catch fire.

So, yes, they burned the bookshop set. But it’s all right, everyone. It’s all right. Because (another spoiler) thanks to the powers of Adam Young, everything was restored again afterwards. Phew. All the books were saved. Shh.


*Funnily enough, everyone thought the anniversary was 1st of May. Including the whole Good Omens team. So they made a brilliant lockdown video** to mark the occasion and celebrate. And then it turned out it was actually the 10th. Just an ordinary cock-up, as Crowley would say.

**Which proves the bookshop, with all its books, was fully restored, doesn’t it? Told you.


If you’re studying from home, have you got your Pocket Chemist yet? Why not grab one? It’s a hugely useful tool, and by buying one you’ll be supporting this site – it’s win-win!

Want something non-sciency to distract you? Why not check out my fiction blog: the fiction phial. There are loads of short stories, and even (recently) a couple of poems. Enjoy!

Like the Chronicle Flask’s Facebook page for regular updates, or follow @chronicleflask on Twitter. Content is © Kat Day 2020. You may share or link to anything here, but you must reference this site if you do. If you enjoy reading my blog, please consider buying me a coffee through Ko-fi using the button below.
Buy Me a Coffee at ko-fi.com

What is Water? The Element that Became a Compound

November 2018 marks the 235th anniversary of the day when Antoine Lavoisier proved water to be a compound, rather than an element.

I’m a few days late at the time of writing, but November 12th 2018 was the 235th anniversary of an important discovery. It was the day, in 1783, that Antoine Lavoisier formally declared water to be a compound, not an element.

235 years seems like an awfully long time, probably so long ago that no one knew anything very much. Practically still eye of newt, tongue of bat and leeches for everyone, right? Well, not quite. In fact, there was some nifty science and engineering going on at the time. It was the year that Jean-François Pilâtre de Rozier and François Laurent made the first untethered hot air balloon flight, for example. And chemistry was moving on swiftly: lots of elements had been isolated, including oxygen (1771, by Carl Wilhelm Scheele) and hydrogen (officially by Henry Cavendish in 1766, although others had observed it before he did).

Cavendish had reported that hydrogen produced water when it reacted with oxygen (known then as inflammable air and dephlogisticated air, respectively), and others had carried out similar experiments. However, at the time most chemists favoured phlogiston theory (hence the names) and tried to interpret and explain their results accordingly. Phlogiston theory was the idea that anything which burned contained a fire-like element called phlogiston, which was then “lost” when the substance burned and became “dephlogisticated”.

Cavendish, in particular, explained the fact that inflammable air (hydrogen) left droplets of “dew” behind when it burned in “common air” (the stuff in the room) in terms of phlogiston, by suggesting that water was present in each of the two airs before ignition.

Antoine-Laurent Lavoisier proved that water was a compound. (Line engraving by Louis Jean Desire Delaistre, after a design by Julien Leopold Boilly.)

Lavoisier was very much against phlogiston theory. He carried out experiments in closed vessels with enormous precision, going to great lengths to prove that many substances actually became heavier when they burned and not, as phlogiston theory would have it, lighter. In fact, it’s Lavoisier we have to thank for the names “hydrogen” and “oxygen”. Hydrogen is Greek for “water-former”, whilst oxygen means “acid former”.

When, in June 1783, Lavoisier found out about Cavendish’s experiment he immediately reacted oxygen with hydrogen to produce “water in a very pure state” and prove that the mass of the water which formed was equal to the combined masses of the hydrogen and oxygen he started with.

He then went on to decompose water into oxygen and hydrogen by heating a mixture of water and iron filings. The oxygen that formed combined with the iron to form iron oxide, and he collected the hydrogen gas over mercury. Thanks to his careful measurements, Lavoisier was able to demonstrate that the increased mass of the iron filings plus the mass of the collected gas was, again, equal to the mass of the water he had started with.

Water is a compound of hydrogen and oxygen, with the formula H2O.

There were still arguments, of course (there always are), but phlogiston theory was essentially doomed. Water was a compound, made of two elements, and the process of combustion was nothing more mysterious than elements combining in different ways.

As an aside, Scottish chemist Elizabeth Fulhame deserves a mention at this point. Just a few years after Lavoisier she went on to demonstrate through experiment that many oxidation reactions occur only in the presence of water, but the water is regenerated at the end of the reaction. She is credited today as the chemist who invented the concept of catalysis. (Which is a pretty important concept in chemistry, and yet her name never seems to come up…)

Anyway, proving water’s composition becomes a lot simpler when you have a ready supply of electricity. The first scientist to formally demonstrate this was William Nicholson, in 1800. He discovered that when leads from a battery are placed in water, the water breaks up to form hydrogen and oxygen bubbles, which can be collected separately at the submerged ends of the wires. This is the process we now know as electrolysis.

You can easily carry out the electrolysis of water at home.

In fact, this is a really easy (and safe, I promise!) experiment to do yourself, at home. I did it myself, using an empty TicTac box, two drawing pins, a 9V battery and a bit of baking soda (sodium hydrogencarbonate) dissolved in water – you need this because water on its own is a poor conductor.

The drawing pins are pushed through the bottom of the plastic box, the box is filled with the solution, and then it’s balanced on the terminals of the battery. I’ve used some small test tubes here to collect the gases, but you’ll be able to see the bubbles without them.

Bubbles start to appear immediately. I left mine for about an hour and a half, at which point the test tube on the negative terminal (the cathode) was completely full of gas, which produced a very satisfying squeaky pop when I placed it over a flame.

The positive electrode (the anode) ended up completely covered in what I’m pretty sure is a precipitate of iron hydroxide (the drawing pins presumably being plated steel), which meant that very little oxygen was produced after the first couple of minutes. This is why in proper electrolysis experiments inert graphite or, even better, platinum, electrodes are used. If you do that, you’ll get a 1:2 ratio by volume of oxygen to hydrogen, thus proving water’s formula (H2O) as well.

So there we have it: water is a compound, and not an element. And if you’d like to amuse everyone around the Christmas dinner table, you can prove it with a 9V battery and some drawing pins. Just don’t nick the battery out of your little brother’s favourite toy, okay? (Or, if you do, don’t tell him it was my idea.)


Like the Chronicle Flask’s Facebook page for regular updates, or follow @chronicleflask on Twitter. Content is © Kat Day 2018. You may share or link to anything here, but you must reference this site if you do.

If you enjoy reading my blog, please consider buying me a coffee through Ko-fi using the button below.
Buy Me a Coffee at ko-fi.com