The Real Reason Why Canister Stoves Are More Fuel Efficient Than Alcohol

Stoves have long been a popular topic in backpacking circles. Alcohol stoves have seen a great deal of popularity in the past decade but recently integrated canister stoves (i.e. Jetboil) have been increasing in prominence. Alcohol stoves won converts from canister stoves by reducing the weight of the components (stove, fuel container) while offering decent fuel economy and a simple operating experience. Since then canister stoves have been on a weight diet and more importantly, integrated canisters systems now offer a less tip-prone cooking experience with remarkable fuel economy while retaining fast cooking speeds. Accordingly, alcohol stoves have been pushed into smaller niches like short trips where the weight of a canister is highly disproportionate to the fuel needed.

The Energy Density Theory
The generally accepted wisdom about canister stoves is they use less fuel because their fuel is more energy dense. These claims are all over backpacking websites. Example:

A quick look at the chemistry seems to support this. Canisters use a variety of fuels but all of them offer very close to 46 megajoules per kilogram. Conversely, the commonly used alcohols (methanol, ethanol) lag behind at 20 and 29 megajoules per kilogram. Big difference? No, because this advantage is entirely offset by the canister weight, which is mandatory because these fuels would be gases without huge pressure present. A 4oz fuel canister isn’t much more than 50% fuel, while a 16oz canister maxes at 70% fuel by weight. So when we look at energy density in terms of energy provided from the combined weight of the fuel + container, canister systems drop from 46 megajoules per kilogram to 25-32 depending on the canister size (4/8/16oz = 25/29/32 MJ/kg). This is the best case scenario as it assumes your trip requires burning the entire canister, yet results are still similar to ethanol, which drops by about 1 megajoule per kilogram when you include the weight of a 1oz pop bottle. Bottom line: Packing ethanol in pop bottles provides nearly the same energy (28 vs 29 MJ/kg) as packing 8oz fuel canisters. Even on long trips, there’s no substantial energy density advantage to a canister vs ethanol, and only a small advantage over methanol. Thus based on energy density alone a canister stove can’t offset the extra weight of the stove, nor come out substantially ahead on the fuel load.

Jetboil: An incredible 70% efficient at heat transfer
Jetboil: Transfers energy at 70% efficiency

The Energy Transfer Theory
Based on the above, it seems that we should be able to do the same amount of cooking on a canister as a similarly heavy bottle of alcohol. This is far from true. Typically you can boil about 50% more water with a similarly heavy canister.

Why? It’s due to increased efficiency at energy transfer, not energy density. Using the BackpackingLight article on integrated canister stoves and applying some additional math (see bottom of this post), we see that canister stoves can be 70% efficient at transferring energy from fuel to water in unrealistic indoor conditions. It’s amazing stuff. Conversely, alcohol stoves are lucky to hit 50% and most hover around 40% (personal data) in ideal conditions and much less in the outdoors (~30%). The best alcohol system I’ve ever tested (1.3L wide pot + caldera cone + Starlyte stove) was 53% efficient indoors for a pint. Canister stoves – and particularly integrated canister stoves – have several efficiency advantages over alcohol. The ones that come to mind are (1) pressured fuel is better mixed with air, resulting in a more complete burn, (2) integration between canister and pot (i.e. Jetboil) reduces heat losses and (3) insulated pots reduce heat losses (again, see Jetboil). This large difference in efficiency is why integrated canister stoves can theoretically more than offset the additional weight of the integrated stove on long trips where fuel demand closely coincides with the fixed amounts of fuel commercially available. Actually achieving this is tougher than most realize, as any scenario where the canister wins requires (1) starting with a close to full canister and (2) using the vast majority of the fuel present. Taking a half full canister will always be heavier.

This is interesting partly because several of these efficiency advantages aren’t intrinsic to canisters. Someone could easily design an elaborate integrated stove/pot system for alcohol with fins everywhere to increase the efficiency. Doing so would increase the system weight, but decrease the fuel used. You will never get quite as efficient or heavy as a canister system (unless you used a pressurized canister, which you could) but the weight-efficiency relationships are similar. For a short trip, a lighter but less efficient system wins, while for longer trips weight invested in efficiency pays off. Some alcohol manufacturers have attempted to add complexity to improve efficiency but nothing exists that’s nearly as elaborate as an integrated canister stoves. Perhaps this is for good reason, as an alcohol system that occupies the same place on the weight-fuel efficiency relationship isn’t that appealing. If similar efficiency was achieved, the weight would be similar so the alcohol system would be quieter, but slower and unable to simmer. Not a clear winner.

My 750mL solo setup. 124g. Everything packs in the pot. Fuel can be stored in the stove.
My 750mL solo setup. 124g. Everything packs in the pot. Fuel can be stored in the stove.
50% efficient indoors. 8 min pint boils from 4 degrees C.
50% efficient indoors. 8 min pint boils from 4 degrees C.

The core appeal of alcohol is the potential to strip away many of these features and occupy an extreme position on the weight-efficiency spectrum (very light but low efficiency) that isn’t possible with a canister because of the mandatory container mass. Thus unless something big changes, integrated canisters will continue to make sense on long trips, while a few ounces can be saved on short-medium trips with alcohol. I actually do think something big is going to change, but it’s a work in progress and it’ll be spring before I really know if it works.

With that said, there’s going to continue to be a contingent that prefers alcohol because it’s much cheaper to operate than canister stoves and it doesn’t entail the waste of spent canisters. Include me in this group. I can buy a summers worth of methanol for the price of an 8oz canister. I also have little need for speed or simmering and I like how easy it is to monitor my fuel situation. Thus I use an alcohol system for all 3 season trips, knowing that is lighter on most trips, and for longer trips I accept the initial weight penalty for the low cost, low waste and easy fuel monitoring of my alcohol system.

The super geeks out there might be interested in calculating the energy efficiency of their own cooking system. It’s a simple thing to do if you have a decently accurate way of determining your fuel usage (either by weight or volume):

Geek Section – Determining Efficiency
1) Determine Energy Absorbed by the Water
A gram of water requires 4.184 joules to rise one degree Celsius. So the total energy gained is 4.184 x water weight (g) x degrees gained.

    Ex: Heating a pint (473ml = 473g) from 40F (4C) to a boil (100C) is 4.184 x 473 x 96 joules or 190,000 joules or .19 megajoules.

2) Calculate Theoretical Fuel Needed
Methanol provides 20 megajoules per 1000g. Ethanol is 29. Canister fuel is 46. Divide the energy gained (step 1) by energy density (i.e. 46) and multiply by 1000 to convert to grams.

    Ex: .19 megajoules of heating theoretically requires 9.5 grams of methanol (.19 / 20 * 1000)

3) Compare to Actual Fuel Usage
Weigh stove before and after to find actual weight of fuel used or measure the volume used and adjust for density to find the weight. Your efficiency is the theoretical heat needed divided by the actual fuel used.

    Ex. 9.5 g / 20g actual fuel used = 47.5% efficient

My typical efficiency in ideal conditions (inside) with a cone and Starlyte is 50%. The best I’ve been able to achieve is with a wide bottom pot (Evernew 1.3L) + cone + Starlyte which resulted in 55%. Typical alcohol stoves are around 40-45% in ideal conditions, and obviously lower in the wild.


  1. This particular cone isn’t custom but it’s no longer available. As I recall, it’s a ULC cone for the Evernew ECA278 pot. Instead of using stakes to raise up the pot over the stove, I use the silicone wrist band ($1 eBay) with the shorter Starlyte stove. This is more stable, adds adjustability and it gives me more coverage up the sides of the pot for better efficiency vs the 12-10. The Starlyte is much better than the 12-10 stove (easier to snuff, save fuel in stove, lights easy in winter, unspillable, similarly efficient). TrailDesigns would probably make you this cone if you ask. If you’ve ordered before, they’ll sell you just a cone without all the extras for less.

    With that said, I think the ultimate set up for a solo hiker is the untreated wide Evernew 0.9L pot (ECA252) with a sidewinder cone and an unrestricted Starlyte stove. With the Starlyte you can ditch the stakes normally needed with the Sidewinder and just rest the pot directly on the cone. The gap will be right. Extremely simple, stable, light and efficient. My wife and I use the 1.3L Evernew wide pot in this manner and it’s as good as it gets.

  2. It has long been established canisters are more efficient than alcohol stoves for multi-day trips.

    The only reason why I use a canister is because a remote stove is efficient down to about -20 Celsius Alcohol stoves will work at low temperature, but it needs to be primed; and the amount of fuel required to generate the heat implicates that is not worthwhile for a solo hiker and is better in team expeditions or sled-dog journeys where melting large volume of water is necessary and fuels are difficult to find.

    Nevertheless, I still like the cat-can stove because of its simplicity, fewer parts and ease of finding fuel. It kind of sucks to walk into a remote village and being unable to find canisters or white-gas. At least I can always burn high-proof vodka, anti-freeze or paint-thinners.

  3. This is awesome stove geekery. I never knew such thought and calculation went into stoves.
    To me canister stoves are like ink jet printers. No matter how well they work, they are expensive and wasteful.

    One question… Does more power improve efficiency because the pot is losing heat for a shorter amount of time the faster you heat it? Or does more power just create more waste heat that never makes it to the pot?

    1. I did a pretty in depth post at the link below discussing power and efficiency as two good metrics of stove performance, complete with some neat graphs:

      But to answer your question, stoves of any type (alcohol, canister, liquid fuel) are almost always more efficient at lower power.

      I think the main reason for this is that heat transfer efficiency declines as power increases. A stove at 1/4 throttle is producing less hot gas (CO2) so that hot gas can spend more time in contact with the pot before newly created gas pushes it out of there. Plus the actual flame does less spilling out the sides of the pot. Presumably the consequences of cranking up the throttle are more severe for smaller diameter pots, and less severe for pots with better heat transfer designs such as a JetBoil.

      The other plausible sounding reason is that at high throttle stoves have difficulty mixing in enough oxygen to burn cleanly, so efficiency declines From talking with the engineers at Trail Designs I’m told this isn’t normally a big factor, but it can be if you’re using a fuel that’s difficult to burn (i.e. propanol in an alcohol stove)

      Opting for less throttle does improve efficiency, but there is an optimum throttle position which is typically above minimum throttle. If you took 3 days to heat water to a boil, the total heat lost from the hot pot to the atmosphere would be substantial. So at some point this negative consequence of cooking slower outweighs the transfer efficiency gains. It’s kind of a mute point though because no one seriously wants to wait an hour for dinner to save a gram of fuel. From years of stove testing, the optimal efficiency occurs at a surprisingly slow rate of heating. A 30 minute boil is routinely more efficient than anything shorter with an alcohol stove. JetBoils have an insulated pot so you could theoretically cook really slowly (i.e. 1 hour boil) and still probably get pretty good efficiency, but it wouldn’t be very advantageous since the stove is already transferring energy and close to maximum efficiency if you’re doing 10-15 minute boils.

      This all ignores the effects of wind, which can change things quite a bit. Wind increases the heat lost from the pot to atmosphere (favoring faster cooking) and also increases heat transfer loses, so the optimal throttle position is probably higher.

  4. Dan do you know what material is used to insulate the jet boil. For winter I wonder if it would be worth making one for my reactor pot. I don’t use the reactor stove but still really like the heat exchanger

    1. The JetBoil sleeve is neoprene, about 5mm thick. A lot of people use aluminized bubble wrap (available at home depot) to create pot insulating sleeves but these wouldn’t be good if you want to use it while it’s on the stove.

  5. Im thinking that for winter I would like to see if it becomes more effiecent melting with neoprene on the pot

  6. Trip length will also come into play with canister stove fuel efficiency. Your calculations assume you start with a full canister and finish with it empty. This is often not the case. Even if you vent the canister before the trip to avoid carrying unnecessary fuel, the fuel weight / canister weight ratio drops and so does your efficiency.

    1. Yup. Good points. I was trying to charitable to canister stoves here, but the difficulty of matching canister capacity with fuel needed is a real con. Ideally you want to have just a little extra left as buffer, but actually achieving this means collecting a whole lot of 90% empty canisters that no one wants to finish off.

      It’s too bad the outdoor industry hasn’t gotten around to re-fillable canisters. If you could buy an awesome titanium canister and then fill it with the amount of fuel you want that might tip the balance quite a bit.

Leave a Reply

Your email address will not be published. Required fields are marked *