How much stove fuel to carry, worked from the energy up

Fuel planning is one multiplication and one division, done in the right order. Work out how many litres of water you will heat, multiply by the energy that takes, divide by the fraction of your stove's output that actually reaches the pot, then divide by the energy in a gram of fuel. Almost everyone gets the first number roughly right and the third badly wrong.

The three numbers the whole calculation rests on#

The energy in the water. Water has a specific heat capacity of about 4.186 kilojoules per kilogram per kelvin. Raising a kilogram (one litre) from 10 degrees C to 100 degrees C is a rise of 90 kelvin, so 4.186 times 90 is 376.7 kilojoules. Round it to 377. If your water starts at 2 degrees C in a mountain stream, the rise is 98 kelvin and the number becomes 410. If it starts at 20, it is 335.

The energy in the fuel. Published net calorific values, measured by bomb calorimetry to methods such as ASTM D4809, put canister gas near 45.8 megajoules per kilogram, white gas near 43.5, and denatured alcohol near 26.8. Conveniently, megajoules per kilogram and kilojoules per gram are the same number, so a gram of gas carries about 45.8 kilojoules.

The fraction that arrives. This is the one people skip. A stove's rated output tells you how fast it burns fuel, not how much of that heat enters the water. Under a water boiling sequence of the kind ISO 19867-1 defines, delivered efficiency for camping stoves ranges from around a fifth to around three fifths depending almost entirely on lid, wind and pot geometry.

Put together: fuel in grams equals 377 divided by efficiency, divided by the calorific value in kilojoules per gram.

Why efficiency matters more than fuel choice#

Read down the gas column. The difference between the worst row and the best is 24 g per litre. The difference between gas and white gas in any row is 1 to 2 g. Fuel type is a rounding error next to whether you put a lid on the pot.

That has a blunt consequence for packing. A windscreen weighing 25 g saves about 6 g of fuel for every litre you boil, so it pays for itself before the end of day two and keeps paying. A lid weighing 15 g saves about 13 g per litre and pays for itself in a single day. Pot geometry belongs in the same conversation, and is covered in camp cookware and kitchen kit.

A worked example: two people, four days#

The scenario. Two people, a four day spring trail trip at moderate altitude, camping in trees, an upright canister stove with a lid and a partial windscreen, water drawn from streams at roughly 10 degrees C. Four breakfasts and four evening meals.

The menu, per person per day:

• Morning: 300 ml of boiling water for porridge, 200 ml for coffee. That is 0.5 litres.
• Evening: 400 ml for a dehydrated meal, 200 ml for a hot drink. That is 0.6 litres.

So 1.1 litres per person per day, and 2.2 litres for two.

Now the arithmetic, step by step.

1. Total water. 2.2 litres a day times four days is 8.8 litres.
2. Fuel for boiling. At 45 percent delivered efficiency, gas costs 18.3 g per litre. 8.8 times 18.3 is 161 g.
3. Fuel for simmering. Two of the four evenings involve five minutes of simmering. At 1.8 g a minute that is 10 minutes times 1.8, or 18 g.
4. Subtotal. 161 plus 18 is 179 g.
5. Margin. Add 15 percent for a cold morning, a spilled pot and a slow relight: 179 times 1.15 is 206 g.
6. Residue. Add 12 g for the fraction of a canister that will not deliver at a useful rate: 218 g.

One 230 g canister covers this, with about 12 g spare, which is two thirds of a litre of boiling water. That is a genuinely thin margin, and it is why people who "always take a 230 for a weekend" run out on a four day trip. Two 100 g canisters, at 200 g of gas and 24 g of residue, would leave you 42 g short.

Now change one variable at a time and watch what happens.

• Forget the windscreen and pitch somewhere exposed. Efficiency falls to 22 percent, gas per litre rises to 37 g, and the boiling fuel becomes 329 g. With margin and residue, 390 g. You now need two canisters.
• Swap to an integrated heat exchanger pot. Efficiency rises to 62 percent, gas per litre falls to 13 g, boiling fuel is 117 g, and the total with simmer, margin and residue is 167 g. One canister with 63 g spare.
• Do the same trip on snow. Melting doubles the cost per litre to about 37 g, and the total lands near 390 g, the same as losing your windscreen.

The headline from this example: forgetting a 25 g windscreen costs you as much fuel as camping on snow. Efficiency is not a refinement, it is the plan.

Feed your own menu into the stove fuel calculator rather than reworking this by hand each trip.

Grams per person per day, by trip style#

These rows are the same arithmetic applied to five common patterns. They are models, not measurements, and the efficiency column is the assumption that drives everything.

The spread between the top and bottom rows is nearly seven to one. Any single rule of thumb quoted in grams per person per day, without stating the efficiency and the water volume behind it, is describing one of these rows and pretending it describes all of them.

Melting snow, and what simmering really costs#

Snow is expensive in a way that surprises people the first time. Ice at 0 degrees C needs 334 kilojoules per kilogram just to become water at 0 degrees C, before any heating starts. Add the 418.6 kilojoules to take that water from 0 to 100, and a litre from snow costs about 753 kilojoules against 377 from a 10 degree stream. Exactly double, near enough.

If the snow starts at minus 10 rather than 0, add roughly 21 kilojoules per kilogram to warm the ice itself, which is under 3 percent and can be ignored next to the other terms.

Volume is the second surprise. Fresh snow has a low density, so a full pot of loose snow yields only a fraction of a pot of water, and you will refill the pot several times per litre. Always start with a little liquid water in the bottom, because a dry pot packed with snow scorches and the snow insulates rather than melts.

Simmering is the quieter cost. At about 1.8 g of gas a minute, ten minutes of simmering costs 18 g, which is roughly one litre of boiling water at 45 percent efficiency. Twenty minutes of simmering a stew for four evenings is 144 g, more than half a large canister, and it appears in nobody's plan. If your menu involves real cooking rather than rehydration, budget the simmer minutes explicitly.

Cold weather adds a third effect that is not about energy at all: a canister that has lost pressure delivers gas more slowly, so boil times stretch and the stove spends longer losing heat to the air. The pressure side of that is worked through in canister stoves in the cold.

Calibrate your own number over three trips#

Published efficiency figures are a starting point. Your figure depends on your pot, your windscreen habits and the places you camp, and you can measure it with a kitchen scale and a notebook.

1. Weigh the canister before the trip. Digital scale, 1 g resolution, gross weight. Write it on the can in marker.
2. Count what you heat. Each time you boil, note the volume, using the graduations on your pot or a marked bottle. Note simmer minutes separately.
3. Weigh it after the trip. Gross weight again. The difference is the gas you burned.
4. Repeat for three trips, ideally with a mix of weather, and total both columns.
5. Divide. Total grams burned, minus 1.8 g for each simmer minute, divided by total litres boiled, gives your grams per litre.
6. Back out your efficiency. It is 377 divided by (your grams per litre times 45.8). For example, 250 g over 11 litres is 22.7 g per litre, and 377 divided by 1,040 is 36 percent.

Six trips of this and you will know your own consumption better than any table, including this one. Anyone who wants the tare weight side of the same protocol, for knowing what is left in a part-used can, will find it in canister stoves in the cold.

Common mistakes#

Planning in boils rather than litres. A "boil" is not a unit. A mug is 300 ml, a pot for two is 1.5 litres. Manufacturer claims of boils per canister rarely state the volume, the starting temperature or the wind.

Using a manufacturer burn time as a fuel figure. Burn time at full output tells you how fast the valve passes gas. It says nothing about whether that heat entered the pot.

Forgetting the water is cold. Planning from 20 degrees C when you are drawing from snowmelt at 2 degrees understates the energy by about 22 percent, because the temperature rise goes from 80 kelvin to 98.

Ignoring the residue. The last 10 to 15 g of a canister exists but will not feed a stove at a useful rate in the cold. Deduct it before deciding a canister is enough.

Carrying no margin. A 15 percent margin on a four day trip for two is about 27 g. It costs nothing and it covers the evening the pot goes over. Fuel is consumable weight in any case, so it sits outside base weight and falls every day you walk.

Assuming a bigger stove uses less fuel. Output and efficiency are separate. A higher output burner reaches boiling faster but often loses more heat around the pot sides while doing it. Wide pots and lids beat powerful burners, as camping stove types compared sets out.

Frequently asked questions#

How much gas does one person need per day?#

Between about 11 g and 73 g, depending on the trip. A summer trip boiling 0.8 litres a day on an integrated pot lands near 11 g; a three-season trail trip with a lid and windscreen near 20 g; melting all your water from snow near 73 g. Any figure quoted without stating litres and efficiency is one of those cases dressed up as a rule.

How many litres will a 230 g canister boil?#

At 45 percent delivered efficiency and 18.3 g per litre, about 12 litres, or 11 once you deduct the residue that will not feed the stove. With an integrated heat exchanger pot the same canister does about 16 litres. With no lid in wind it does about 6. The canister is fixed; your setup decides the answer.

Does melting snow really double the fuel?#

Close to it. The phase change from ice to water at 0 degrees C costs 334 kilojoules per kilogram on its own, and heating that water to boiling costs a further 419, so a litre from snow needs about 753 kilojoules against 377 from a 10 degree stream. Warming the ice from minus 10 to 0 adds only about 3 percent more.

How much fuel does simmering use?#

On the order of 1.8 g of gas per minute at a genuine low simmer. Ten minutes costs about 18 g, roughly one litre of boiling water. Four evenings of twenty minute cooking is about 144 g, which is why menus built around real cooking need a fuel plan that counts simmer minutes rather than only counting boils.

Is white gas or canister gas more fuel efficient?#

Neither, in any way that matters. White gas carries about 43.5 megajoules per kilogram and canister gas about 45.8, a difference of roughly 5 percent, and the stoves reach similar delivered efficiencies with the same lid and windscreen. Choose between them on cold weather behaviour, metering and fuel availability, not on grams per litre.

How much margin should I add to a fuel calculation?#

About 15 percent, plus a fixed deduction of 10 to 15 g for canister residue. The margin covers cold water, a windy pitch, a spilled pot and one extra hot drink on a bad evening. On a four day trip for two that is roughly 27 g of margin, which is not worth arguing about and has saved a lot of cold dinners.

How do I work out how much gas is left in a used canister?#

Weigh it. The net content is marked on the can under EN 417 and most cartridges also carry a stamped tare weight, so gross weight minus tare is the gas remaining, to the gram. Divide by your own grams per litre figure, or by 18 g as a default for a stove with a lid and a windscreen, to convert it into meals.