Gas density and why 30 metres is the line

Enriched air is heavier than air

Start with the fact that ruins the usual mental model.

The density of a gas is the ideal gas law rearranged: ρ = PM/RT, where M is the molar mass of the mixture. Pressure and temperature are the same for everybody at a given depth, so the only thing that separates one breathing gas from another is what the molecules weigh. Air comes out at 28.97 g/mol, which at 1 atmosphere and 15 °C is 1.225 g per litre — the same figure the cylinder tables use.

Now make it EAN32. You have replaced 11 per cent of the mixture, by volume, with oxygen. Oxygen’s molar mass is 32.0; the nitrogen and argon it displaced average 28.2. Oxygen is heavier than the gas it replaces, so the mixture is heavier: 29.39 g/mol, and 1.243 g per litre at the surface.

Enriched air is about one and a half per cent denser than air at the same depth. Not lighter. Denser.

That single fact reorganises the whole subject. Nitrox buys you decompression time — real, substantial, and covered in our guide to breathing gases — and it buys you nothing at all for the work of breathing. Whatever it is that makes air hard to breathe at 40 m (130 ft), a nitrox fill makes marginally worse. Only helium, at 4.0 g/mol against nitrogen’s 28.0, moves the line at all.

The limit nobody teaches

Density is not just an interesting property. It has a number attached, from work that most recreational divers have never heard of.

Gavin Anthony and Simon Mitchell, working on rebreather physiology, stratified a large body of work-loaded test dives by the density of the gas being breathed and looked at how often those dives failed because the subject’s end-tidal carbon dioxide went above 8.5 kPa. Near a respired density of 6 g per litre there is a sharp upward inflection in the risk, with most failures caused by dangerous CO₂ retention.

The work came out of rebreather diving, where a diver can be pushed to a high density with no bubbles to show for it, but the mechanism has nothing to do with the loop. It is the density of the gas in the airway, and an open-circuit diver at 40 m on air is breathing the same 6 g per litre as a rebreather diver on the wrong diluent. Anthony and Mitchell say so explicitly: the same analysis run on open-circuit trials produced a virtually identical result.

Their recommendation, in their own words, is an ideal maximum gas density of 5.2 g/L and an absolute maximum of 6.2 g/L. They convert those to air diving at 31 m and 39 m using reference densities at 0 °C; run the same arithmetic at the 15 °C convention used everywhere else in diving and the limits land at 32.9 m and 41.1 m. Either way, the answer is the same two numbers a recreational diver already knows: about thirty metres, and forty.

Read the nitrox rows carefully. EAN32 reaches the recommended limit at 32.3 m and the absolute one at 40.4, both marginally shallower than air. EAN36 is another two tenths of a metre shallower again. Trimix 21/35 — the first mix with helium in it — does not reach 5.2 g/L until 50.6 m, and 18/45 not until 59.2. Heliox is off the chart entirely, at 136.7 m.

The failure is carbon dioxide, and it is the bad kind

Nothing about breathing dense gas kills you directly. What it does is let carbon dioxide build up, and hypercapnia is the failure mode that turns a manageable situation into a fatal one, because the first thing it takes is judgement.

The sequence is familiar to anyone who has been slightly over-exerted at depth. You work — swimming into current, sorting out a problem, finning to hold position because you are badly weighted — and the extra CO₂ needs extra ventilation to clear. Ventilation is exactly the thing the dense gas has made expensive. So the CO₂ rises, which drives you to breathe harder, which costs more work, which makes more CO₂. Divers describe the endpoint as a feeling of not being able to get enough air from the regulator, and the instinctive response — breathing faster and shallower, moving more gas through dead space — is precisely wrong.

Two things make this worse than it sounds. Carbon dioxide is a potent narcotic in its own right, ranked around twenty times as narcotic as nitrogen, and it increases blood flow to the brain, which delivers everything else faster. And the standard recreational habits that raise CO₂ — skip-breathing to stretch a cylinder, over-exertion, a badly serviced regulator — are all things divers do on purpose.

The other reason thirty metres keeps coming up

Density is only half the story at that depth, because narcosis arrives at almost exactly the same place by a completely different route.

Every gas that dissolves in you has an anaesthetic effect at pressure, and the effect tracks lipid solubility closely enough that Hans Meyer spotted the correlation in 1899, with Charles Overton arriving at it independently two years later — the more soluble the gas is in oil, the less partial pressure it takes to impair you. Nitrogen is not special; it is simply what you happen to be breathing. Helium is barely narcotic at all, at 0.045 on a scale where nitrogen is 1. Neon is 0.3, hydrogen 0.6, argon 2.3.

Oxygen is 1.7 — more narcotic than nitrogen, which is the second reason enriched air does not clear your head.

Impairment is not usually noticeable below 30 m. Between 30 and 50 m the reported effects are delayed responses to what you see and hear, reasoning and immediate memory affected more than motor coordination, calculation errors, wrong choices, idea fixation, and a sense of well-being that makes all of the above harder to notice from the inside.

There is an arithmetic for this, and it is the one piece of technical gas planning worth borrowing at recreational depths. Equivalent narcotic depth is the depth multiplied by the fraction of the mix that is not helium, so trimix 18/45 at 60 m puts your head at about 28 m. Recreational divers have no helium and therefore no lever: on air and on nitrox alike, your equivalent narcotic depth is simply your depth.

Three independent limits on the same depth scale, and the striking thing is which one is missing. Oxygen — the constraint recreational divers are actually taught to compute — does not reach a partial pressure of 1.4 bar on air until 56.5 m, sixteen metres past the recreational limit. It is not what stops you. Narcosis and density are, and they arrive within a few metres of each other.

That is the honest answer to why the limit is 40 m: it sits between the recommended density limit at 33 m and the absolute one at 41, at the depth where measurable narcosis has been present for ten metres already. It is not an arbitrary line in a manual. It is where two unrelated pieces of physiology happen to converge.

What to actually do about the martini rule

The martini rule — one drink per ten metres below twenty — is folklore. It is a teaching device that gives new divers something familiar to hold on to, and it is wrong in the way that matters: narcosis varies enormously between individuals and between dives for the same individual, and the things that make it worse are cold, stress, fatigue, hard work and carbon dioxide retention rather than depth alone. Two divers side by side at 35 m are not having the same experience, and neither of them is a reliable witness to their own.

You cannot build a tolerance to it either, though you can learn to work through it: practised, rehearsed tasks survive narcosis far better than novel ones, which is the real argument for doing a deep dive the same way every time. The effects also lag the depth by a minute or two while the blood catches up, and at least one study found they persist for half an hour after the dive — so the diver who feels fine on the boat is not evidence of much.

The two halves of this guide compound each other, which is the part worth carrying away. Dense gas raises carbon dioxide; carbon dioxide is itself strongly narcotic and increases the blood flow carrying every other narcotic gas to your brain. A diver working hard at 40 m on air is not experiencing narcosis and breathing difficulty as two separate problems. They are one problem with two names.

The practical consequences are unglamorous.

Work less. Density limits your capacity to work, so plan deep dives that do not require any. Currents, long swims and heavy cameras are the things that turn a survivable gas into a problem.

Fix the weighting. A diver two kilos heavy is finning continuously to stay off the bottom, which is work, which is carbon dioxide. The arithmetic is in buoyancy and weighting.

Do not skip-breathe. It raises CO₂ deliberately, at the depth where you can least afford it, to save gas you should have planned for instead — see how long a tank of air actually lasts.

Service the regulator. A demand valve that breathes acceptably at 10 m and stiffly at 40 is a regulator that needs attention, not a fact of life.

And past 40 m, use helium. Not because it is glamorous, but because it is the only variable in ρ = PM/RT that a diver controls.

Where the density bites

These are the sites in our data where recreational divers routinely go to the depth this guide is about: 35 to 45 metres, mostly on air, mostly with a working ascent still to come.