Canister stoves in the cold: pressure, blends and what actually works

A gas canister is not a fuel tank, it is a small pressure vessel holding a liquid in equilibrium with its own vapour. The stove burns the vapour. While the liquid's vapour pressure sits above the air outside, gas flows. When it does not, the stove goes out and opening the valve does nothing. Everything else follows from that.

Why a cold canister stops feeding the stove#

Vapour pressure is a property of the liquid and its temperature, nothing else. Warm the liquid and molecules leave the surface faster until the vapour above pushes harder; cool it and the reverse happens. The Antoine equation describes that curve for each pure hydrocarbon, and the numbers it produces explain every cold weather canister failure people report.

What matters to a stove is gauge pressure: absolute vapour pressure minus atmospheric pressure. That is what pushes gas through the jet, and when it reaches zero the flow stops.

Read the n-butane column. At 0 degrees C it has almost nothing left to give, and by minus 10 it sits below atmospheric pressure, so it will not push out of the can at all. That is why a cheap butane cartridge is a summer item. Isobutane buys about 11 degrees over n-butane, and propane is in a different class, holding usable pressure into the minus thirties.

These are pure components. A blend behaves like none of them, which is where the interesting part starts.

What is in the blend, and how it changes as you burn#

Take a nominal 230 g canister at 25 percent propane by mass: 57.5 g of propane and 172.5 g of isobutane. Pressure in a liquid mixture follows mole fractions rather than mass fractions, so convert. Propane has a molar mass near 44.1 grams per mole, isobutane near 58.1.

• Propane: 57.5 divided by 44.1 is 1.30 moles.
• Isobutane: 172.5 divided by 58.1 is 2.97 moles.
• Total 4.27 moles, so propane is 30.5 percent of the liquid by mole.

Treat the mixture as ideal, which is close enough for two similar hydrocarbons, and apply Raoult's law: each component contributes its own vapour pressure times its mole fraction in the liquid. At 0 degrees C that is 0.305 times 4.79 bar for propane, or 1.46 bar, plus 0.695 times 1.57 bar for isobutane, or 1.09 bar. Total 2.55 bar absolute, about 1.5 bar gauge. Comfortable.

Now the part almost nobody accounts for. The vapour leaving the can is not the mixture it left behind. Propane contributes 1.46 of the 2.55 bar, so the vapour is 57 percent propane by mole, which is roughly half propane by mass. You are burning gas that is 50 percent propane out of a canister that is 25 percent propane. That is fractionation, and the liquid gets poorer in propane from the first minute of the first burn.

Run it forward. Burn the first third, 77 g, at that vapour composition, and you remove roughly 38 g of propane and 38 g of isobutane. What remains is about 19.5 g of propane and 134.5 g of isobutane: 12.7 percent propane by mass, 16 percent by mole. Redo Raoult at 0 degrees C and the total is 0.77 plus 1.32, or 2.09 bar absolute, about 1.1 bar gauge. Pressure has fallen by more than a quarter with no change in air temperature.

Keep going and the propane runs out before the isobutane does. By the last third you are running what is effectively a pure isobutane cartridge: 0.55 bar gauge at 0 degrees C, 0.30 bar at minus 5.

This is the honest reason a half-used canister performs badly in winter while the same canister was fine in September. It is not the valve, not the stove and not the cold alone. It is that the fuel left inside is no longer the fuel you bought.

The canister cools itself while you cook#

The second effect is simpler and just as consequential. Evaporating a liquid takes energy, and that energy comes out of the liquid and the metal around it. During a sustained burn the canister therefore falls below the temperature of the air around it. You can feel it: the can goes cold and often wet with condensation, and in freezing conditions it frosts.

The size of the drop depends on burn rate, how much liquid remains as thermal mass, and what the can stands on. A nearly empty can on frozen ground with the burner wide open cools fastest, because little liquid is left to buffer the heat loss and nothing conducts warmth back in. That is why output which was fine at ignition fades over a five minute boil.

Combine the two effects and the failure mode is obvious. Air at plus 3 degrees C, a canister two thirds empty and therefore isobutane rich, plus several degrees of evaporative cooling during the burn, puts the liquid near or below zero with a fuel that has no margin left. The stove sputters and dies halfway through the pot of snow.

Weighing: knowing your remaining gas to the gram#

Shaking a canister tells you nothing useful. Weighing is exact and takes ten seconds.

Cartridges made to EN 417 carry a marked net content, and most carry a stamped tare, the mass of the empty can and valve, usually on the base or shoulder. Remaining gas is gross weight minus tare.

1. Get a scale. A digital kitchen scale reading to 1 g is enough. Weigh at home, not at the trailhead.
2. Find the tare. If it is stamped, use it. If not, burn one canister of that size to nothing and weigh the empty. Tare varies by size and maker but is consistent within a product line, typically 90 to 160 g for the common 100 g and 230 g sizes.
3. Subtract. Gross minus tare is the gas you own.
4. Convert to cooking. At the 18 g per litre figure derived in stove fuel planning for a stove with a lid and a windscreen, 90 g of remaining gas is five litres of boiling water. Feed the number into the stove fuel calculator rather than estimating.
5. Log it. Write the gross weight on the canister in marker before and after every trip, and keep part-used cans labelled between seasons, as gear storage covers. Two trips of this and you know your own consumption rate better than any published figure.
6. Discount the residue. Treat the last 10 to 15 g as unavailable in cold conditions. It is there, but it is isobutane rich and will not deliver at a useful rate.

The winter consequence: carry canisters that are full or clearly labelled, and burn part-used cans on warm trips. A part-used canister is a summer canister.

The cold weather decision ladder#

Take the cheapest step that covers your conditions, and read the temperature as the coldest you will actually cook in, which is usually breakfast rather than dinner.

Two of these need explanation. Standing the canister in liquid water works because water above 0 degrees C is a large, self-limiting reservoir: freezing a kilogram releases 334 kilojoules, so it takes a long time before the bath stops holding the canister near zero. It cannot overheat the can, which is what makes it safe.

Inverting the canister only works on a remote stove built with a preheat tube. Inverted, liquid runs down the hose under gravity and vaporises in a loop heated by the burner, so vapour pressure stops being the delivery mechanism and fractionation stops mattering. Light the stove upright, let the loop get hot, then rotate the canister. Inverting an upright stove sends raw liquid to the jet and produces a flare.

For the wider choice between stove architectures, see camping stove types compared, and for pots and windscreens see camp cookware and kitchen kit.

Remedies that are dangerous#

A heat reflector under an upright stove. Reflectors sold to bounce heat back at the pot also bounce it at the canister directly beneath the burner. This is the commonest way people overheat a cartridge, because the intent is benign and the effect is invisible until it is not.

A windscreen wrapped fully around an upright stove. Same mechanism: the screen makes an oven with the cartridge inside it. Use a partial screen that shields the pot and leaves the cartridge in open air, or a remote stove where a full screen is safe.

Insulating a canister that sits under a burner. A sleeve on a remote canister on the ground is fine. The same sleeve on an upright stove traps burner heat against the can.

Hot water baths and stove exhaust. Cool water is the point. Hot water is a heating device with no upper limit that you control.

The safe list is short: body heat before use, a foam pad underneath, a bath of cool water, a windscreen that shields only the pot, an inverted remote canister, or liquid fuel.

Common mistakes#

Taking three part-used canisters on a winter trip. Each is butane rich and each carries its own steel. Take one full can.

Sleeping with the canister and then cooking outside for twenty minutes. The warm start lasts a few minutes before evaporative cooling erases it. Warming helps most when paired with something that keeps heat in the can, such as a water bath.

Assuming a "four season" blend fixes everything. A higher propane fraction raises the starting pressure, but fractionation still strips it out first. The last third of a high propane canister is still isobutane.

Cooking in the tent to stay warm. A cold canister is an inconvenience. Carbon monoxide in a closed shelter is not, and a stove indoors also floods the tent with vapour, as tent condensation explains.

Frequently asked questions#

At what temperature do gas canisters stop working?#

There is no single number, because it depends on the blend and on how full the can is. A fresh canister at 25 percent propane still has usable pressure at minus 10 degrees C. The same canister two thirds empty is effectively pure isobutane, marginal at 0 and useless by minus 10. Judge by fill level as much as by air temperature.

Why does my stove get weaker halfway through a canister?#

Propane boils off preferentially, so the vapour leaving the can is richer in propane than the liquid inside it. On the arithmetic here, a 25 percent propane canister delivers roughly 50 percent propane vapour at first, which strips the propane out early. By the final third you are running on isobutane alone, at a fraction of the original pressure.

Is it safe to warm a gas canister?#

Body heat, a jacket pocket, a sleeping bag or a bath of cool liquid water are all safe, because none can raise the canister above a few degrees C. Flames, stove exhaust, hot water, car heaters and reflectors under an upright burner are not, because none has an upper limit you control. Cartridges carry a marked maximum exposure temperature for a reason.

Does putting the canister in water really work?#

Yes, and it is the most underrated trick in winter camping. Liquid water cannot be colder than 0 degrees C, and freezing it releases 334 kilojoules per kilogram, so a pan of water holds the canister near zero while the stove draws heat out of it. It also stops the evaporative cooling spiral that kills output mid boil.

How do I know how much gas is left in a canister?#

Weigh it. Cartridges carry a marked net content and usually a stamped tare weight, so gross weight minus tare is the gas remaining, to the gram, on any kitchen scale. Divide by about 18 g per litre of boiling water, for a stove with a lid and a windscreen, to turn that into meals.

Is an inverted canister stove better than liquid fuel for winter?#

It is lighter and simpler and removes the vapour pressure problem completely, so for most winter trips it is enough. Pumped liquid fuel wins when you are melting large volumes of snow for a group, running below about minus 15 degrees C for days, or working where cartridges cannot be bought. The trade is 200 to 400 g and a maintenance habit.

Does a canister work better at altitude or worse?#

Slightly better, in terms of feed. Atmospheric pressure falls with altitude, so the gauge pressure driving gas through the jet rises a little. What gets worse is cooking: water boils at a lower temperature, so food takes longer and you burn more fuel. Altitude and cold usually arrive together, and the cold is the part that stops the stove.