Alcohol fuel efficiency for backpacking stoves – ethanol vs methanol
Posted: May 12th, 2013 by: h2
I’ve been working on the problem of alcohol stove efficiency in my spare time, as a hobby, to relax from more pressing regular work matters, and I finally realized, the information on the internet is not very accurate, so I’m going to work out here the basic chemistry of alcohol fuel efficiency. My starting point was the very fine article on the Ion Stove by Sgt. Rock, which set me on my quest to get the actual data as closely as I could. If I make any errors in my calculations, do please post a comment of correction and I’ll update this posting.
While I might appear to get lost in the details, let me remind you of just why alcohol stoves are so lovely: they are silent. You can use any fuel, from toxic methanol to good quality ethanol in them, and they will still work, though of course you’ll need to carry about 20% more weight in fuel if you use a primarily methanol based fuel, as you will see as we do the calculations below. See appendix 1 for fuel types and how to get them. See appendix 2 for a table of fuel use / water starting temperature for ethanol and methanol.
The key point is that there is no such thing as ‘alcohol’, there are types of alcohol, and when you learn how to use your stove, and read how people get certain efficiencies, you need to be aware of significant differences in efficiencies, for example, ethanol, at 21.1 kjoules per ml, contains a whopping 33% more energy per ml than methanol, at 15.8 kjoules per ml. And, turning it around, methanol contains only 75% of the energy of ethanol. A blend like SLX probably contains around 18-19 kjoules / ml, though it could be less.
The basics – Joules to heat 1 pound of water
To make things easier, we are going to work from 2 cups of water, 16 oz, which weighs one pound. As always, we need to be clear on the units, we’ll be using 1 pound of water because 2 cups conveniently weighs 1 pound. wikipedia tells us that:
The British thermal unit (symbol Btu or sometimes BTU) is a traditional unit of energy equal to about 1055 joules, 1.054kj. It is the amount of energy needed to heat one pound of water by one degree Fahrenheit. In scientific contexts the Btu has largely been replaced by the SI unit of energy, the joule.
Appendix 4 shows some more conversion information for energy units.
For our purposes this is convenient. So 1 kj, actually 1.054 kj, heats one pound of water 1 degree F. 1 kj is about the energy released by one wooden match burning (pdf has energy content of many things, conversion units etc). However, it’s not quite that simple, since 2 cups of water weighs actually about 16.7 ounces by weight, ie, a bit more than a pound. Since 2 cups of water actually weighs 16.7 ounces, or 1.04375 pounds, this will throw off the following numbers for 2 cups slightly, but not a whole lot. However, all we need to do is multiply the total by 1.04375 to get the amount of fuel needed for the volume 2 cups, which is easy, so that’s what we will do below.
Energy content of ethanol
Sgt Rock lists the energy content per pound of ethanol as 12,550 btus per pound ie, 12,550btusx1.054=13,227 kj, and the energy content of methanol as 10,200 btus per pound (10,750 kjoules). First let’s check this. This source gives the energy content of ethanol as 23.4 MJ per kg. This comes from the IEA. Keep in mind that you can basically never find or buy 100% pure ethanol, all ethanols you will buy contain about 5% water, so that confused the matter. This pdf from ipst.gatech.edu for example lists it at 11,500 btus per pound. So clearly the first thing to do is find out what the actual energy content of 95% pure ethanol is. The question is made more difficult by the mix of units, ie, joules/btus per gallon (a volume measurement) vs joules/btus per pound, by weight, that is. So during this process we will need to find standard units to get things consisten.
The “heat content” of a gallon of ethanol is 76,000 British Thermal Units (BTU). But 1 BTU = 1054 joules, so the energy density (by volume) of ethanol is 21.2 kilojoules per cubic centimeter (aka, 1 ml).
Since the density of ethanol is 0.789 grams per cubic centimeter, the energy density of ethanol is 26.8 megajoules per kilogram. src
Ok, that clears it up. 11,500btu x 1.054=12,120 kjoules per pound. At 21.2 kj per cubic cm, we get the energy density of 1 ml. This is quite close to the number I arrived at using a different method before I did this research, and a bit lower than sgt rock’s numbers. So 1 ml of ethanol will heat at 100% efficiency 1 pound of water 21.2 degrees F, to make these numbers meaningful.
Or, using agmrc.org convenient charts, we see that: 1 gallon = 75,670 Btu – LHV; 1 liter = 21.1 megajoules – LHV, and there we have it.
Now we’re getting somewhere. If we start at 70 F, we need to heat the water 142 degrees F, so we will need: 142/21.1 = 6.72 ml of ethanol, let’s call it 6.7 ml of fuel at 100% efficiency to boil 2 cups of water using ethanol.
Keeping in mind the problem that water weighs a bit more than 1 pound for 2 cups, we can use a touch less than 2 cups to make up for this problem. x = 16×16 / 16.7 will yield the fluid ounce measurement, or: 15.3 fluid ounces. Then we can calculate the amount required to boil each ounce fluid. Or we can change the specific gravity to account for this issue. or we can just use basic algebra: if we need 6.7 ml ethanol to boil 15.3 fluid ounce, then x/6.7 = 16/15.3, x = 7 ml ethanol to boil the full 2 cups.
Now for methanol energy densities and boiling
Note that LHV (lower net heating value) is the number we will use if you see that since it corresponds to what you would expect to see in the real world. HHV is the highest possible value, but wouldn’t be useful for real world measurements.
1 liter methanol gives: 1 liter = 15.8 megajoules – LHV. 1 gallon = 56,560 Btu -LHV. Since 1 liter is 1000 cubic cm, that means 15,800,000 joules / 1000 ml gives us 15.8 kj per ml for methanol.
The density of methanol is about 0.8, a touch heavier than ethanol, that is.
Using the same formula as above, we need 142/15.8 = 8.98 ml methanol at 100% efficiency to boil 2 cups of water starting at 70F.
So that’s where sgt rock got the 9ml at 100% efficiency, but that’s only for methanol, as we see above, the numbers for ethanol are much lower, 6.7 ml required at 100% efficiency.
For the 2 cups weight/volume issue, if we need 9 ml ethanol to boil 15.3 fluid ounce, then x/9 = 16/15.3, x = 9.4 ml methanol to boil the full 2 cups.
Moving on to the actual boils
So now we have the information, at 100% efficiency, which of course is not possible to achieve, we need 9 ml of methanol and 6.7 ml of ethanol to boil 2 cups of water.
The best efficiency I’ve seen (confirmed by sgt rock as using SLX, 50% ethanol/50% methanol roughly) is 12 ml to boil 2 cups. You’d need a bit more with a pure methanol fuel like heet, which you can find at autoparts stores, SLX should give similar results though the formulation can change. You can find SLX at your hardware store, big food stores sometimes, and REI. My own tests give a best result of about 14 to 15 ml SLX to boil 2 cups water. 15 ml mostly ethanol, like Klean Strip Green, from slx as well, works fine to boil 2 cups with my penny stove modified for narrow pots, and a tall wind screen. A sgt rock type ION stove/stove screen is more efficient, and slower, that’s what got the 12 ml boil of a wide pot using SLX.
SLX, assuming it actually is what it says, is what I will test on because it’s easy to find and most people can obtain it. It will require at 100% efficiency about the difference between ethanol and methanol, ie, 8.2 ml of SLX.
I am going to use 55% as the efficiency because that’s what I can achieve easily in my testing, in real world conditions. This results in a full rolling boil and keeps boiling for up to 60 seconds or so, give or take 20 seconds. The numbers I list for methanol are pure methanol, something like SLX would have 20-40% ethanol, which you can work out via testing to show the true boil times.
- Normal water – ethanol: Start temp water: 70F: 142/21.1 = 6.7 x 1.04375 = 7 ml at 100%, assuming 55% efficiency, 7/x = .55, or x = 7/.55 = 12.7 ml to boil normal water.
- Normal water – methanol: Start temp water: 70F: 142/15.8 = 9 ml x 1.04375 = 9.4 ml at 100%, assuming 55% efficiency, 9.4/x = .55, or x = 9.4/.55 = 17.9 ml to boil normal water.
- Normal water – SLX 50/50% ethanol/methanol: Start temp water: 70F: 142/18.45 = 7.7 ml x 1.04375 = 8 ml at 100%, assuming 55% efficiency, 8/x = .55, or x = 8/.55 = 14.5 ml to boil normal water.
I have been unable to reproduce sgt rock’s extremely efficient boils in my own tests, though other people claim to have gotten to around 13 ml boils. I have gotten to around 14ml, and 15ml consistently, so that’s what I will use for my numbers. His numbers give about 60% efficiency, but again, I have not been able to reproduce that, even with a pseudo Ion stove setup I made for testing this, so for now I consider 13 ml as theoretically possible if everything is perfect, but 15ml actually achievable with standard setups. A cone type wind/heat screen should boost efficiency I believe by 1 or 2 ml, but they are awkward to pack away and make, so I haven’t done much testing with those, but i believe a very low output stove like the Ion coupled with a full height screeen may do the trick.
See appendix 2 for a table of fuel use / water starting temperature for ethanol and methanol.
Conclusion: yes, you too with very little work can achieve 1/2 ounce by volume 2 cup boils
Making stoves that can do this 1/2 oz for 2 cups boil is not hard, here is one how to make an ION stove showing how to do just that, and at a very low cost, using just regular SLX denatured alcohol for fuel. If you use methanol based fuels like HEET, expect to use 2 ml or so more to boil your water per meal. That’s not a huge amount, but it does add up after a while.
The real goal, of course, is to be able to bring one week’s fuel in an 8 oz bottle, and, providing you use SLX or mainly ethanol fuels, you can do that fairly safely, with an ounce leeway.
Keep in mind also, the amount of fuel required is a function of the starting temperature of the water. All super efficient results you read online are a based on tap water and inside tests, in almost all cases.
Major point: remember the specific gravity of the fuel when comparing canister fuel weights to alcohol
I see all the time people comparing the grams a canister stove consumes to the ounces measured, by volume, that an alcohol stove uses. This is a sneaky way to try to make the canister stove look better while nobody notices. Gas in canisters is always sold by the gram, so there’s no problem with units there.
You can read more about specific gravity/density/weight here. I believe I will need to slightly modify the water/fuel calculations re weight, but for now the information is fairly close.
Let’s take a 7 day trip at 30ml ethanol consumption a day, which conviently, is 1 ounce. So we would use in that 7 day trip, cooking two meals a day, 7 ounces of alcohol. But here’s the thing, remember the specific gravity we mentioned above? Now take that 7 ounces and multiply by .79. So that’s 5.53 ounces of fuel by weight, not volume. Add about a 1 ounce fuel container, and you are 6.5 ounces, then add about 1.5 ounce for the stove/pot stand/wind screen (the heat shield weighs about 3/4 ounce), which is reasonable to expect, maybe a bit less, and you’re at 8 ounces total, not including reflectix cozy to let the food cook after you boil the water, which weighs about 1 oz. The pot, say, a 900 ml titanium pot, weighs about 3.5 ounces, so the total weight, everything in the cookset except the cups/spoons, which would be the same for any cookset you use, weighs 12 to 13 oz total, for 7 days.
Just for comparison, an average canister stove with 100 grams of gas, which is the size you’d need to bring for a 7 day trip, well actually it wouldn’t quite make it if the average of about 10 grams per boil happens, then you’d need two, or an 8 ounce, 250 gram one, with the heavier cannister. I believe a 100 gram fuel cannister weighs about 3 ounces for the metal container, an 8oz one weighs about 5 ounces I believe, and, unlike alcohol, you can’t measure out the amount you need for the trip. Suddenly the canister stoves don’t look so great anymore re the weight, do they? So a 100 gram canister at start of trip weighs about 7 ounces, two of them, 14 oz, an 8 ounce weighs about 12 or 13 ounces, plus 1.5 to 3 ounces for the stove. Remind me again why people think using non re-usable canisters is a good idea? Unstable, prone to wind, hard to make real wind screens for due to danger of overheating canister itself if you use the direct attachment method which most people use. Even with the greater efficiency of the canister gas fuel, the weights simply do not balance out. And if you use the even heavier jetboil system, which is more efficient, but that efficiency is lost due to the greater weight, it’s even worse, particularly on shorter trips.
So now you know why long distance hikers use alcohol stoves. Plus, and really, this is the major point, as noted, they are silent, totally silent, delightfully, beautifully silent.
Real world examples, cooler/ cold water brrr
I don’t know about you, but whenever I dip into streams to get my water, the water is cold. So the first lesson is, let the water warm up if the water is colder than air temperature, ie, let it sit for a while before cooking. This could save some fuel.
Here’s some examples:
- Cool water – ethanol: Start temp water: 50F: 162/21.1 = 7.67 x 1.04375 = 8 ml at 100%, assuming 55% efficiency, 8/x = .55, or x = 8/.55 = 14.5 ml to boil cool water.
- Cool water – methanol: Start temp water: 50F: 162/15.8 = 10.25 x 1.04375 = 10.7 ml at 100%, assuming 55% efficiency, 10.7/x = .55, or x = 10.7/.55 = 19.5 ml to boil cool water.
- Cool water – SLX – about 17 ml
- Cold water – ethanol: Start temp water: 35F: 177/21.1 = 8.38 x 1.04375 = 8.75 ml at 100%, so: 8.75/x = .55, or x = 8.75/.55 = 15.9 ml to boil cold water.
- Cold water – methanol: Start temp water: 35F: 177/15.8 = 11.2 x 1.04375 = 11.7 ml at 100%, so: 11.7/x = .55, or x = 11.7/.55 = 21.3 ml to boil cold water.
- Cold water – SLX – about 18.6 ml
See appendix 2 for a table of fuel use / water starting temperature for ethanol and methanol.
We still have not yet covered altitude (see appendix 3 for chart of altitude/temperature of boil). But as you can see, over a week trip, going to a place, say, in the cold time of year, where the water is 35 degrees F, you will use a substantial amount more fuel than in a more temperate area/season.
Appendix 1: Sources of alcohol fuels
Remember: METHANOL IS POISON!!!!!!. Ethanol is not that toxic, and certainly doesn’t cause blindness and other health issues when drunk in moderation. All denatured alcohols contain toxins to make them poisonous, but some contain much more than others, like SLX, which is mostly methanol.
- SLX – standard, probably about 50% methanol based on my tests/performance numbers. Available at most hardware stores (Ace hardware is a great resource), REI also sells it. Decent as fuel except for lower heat content. Burns with a blue flame and fumes and fluid are toxic, don’t breath in or burn indoors. Mostly methanol, which is poisonous.
- Klean Strip Green – Also made by SLX, this is about 90% or so ethanol, you can order it from Ace, OSH has it, Home Depot (but avoid that store if you can, it’s an awful place), True Value probably has it too. $7.50 a quart, not bad as fuel, but leaves soot and burns with a slightly yellow flame. Also harder to optimize stoves for in terms of air flow.
- Crown denatured regular, Lowes has it, it’s about the same as SLX I believe.
- Crown Next (Lowes) – warning, contains ISOPROPANOL, which can cause soot.
- E-NRG – DO NOT USE THIS, I tried it and it burns horribly, but I’m listing it here just in case you think it might be a good idea. All it is good for is making nice yellow flames in the fireplaces it’s designed for, and for adding massive amounts of soot to your pot each burn. Skip it..
- Update: do not buy this, the gasoline component (2-5% of total) makes this burn far too hot for alcohol stove setups. It might be good for a compressed injection type alcohol stove however. Ignite E98 Ethanol HyperFuels That’s right, you read it right, E98. I’m ordering this to test it, it costs the same after shipping as Kleanstrip green, about 7.50 a quart. I’ll report on how this works, I think it should be very good if it is actually almost all ethanol, though I assume they mean 98% of the fuel is 95% ethanol, not lab grade at higher concentrations. With shipping, about $30 a gallon. They also sell esssentially 100% methanol as well. WARNING: e98 contains 2% gasoline. I will test this and report back, it may work with dilution by SLX, we’ll see.
- HEET – any autoparts store. Remember, this is methanol, they are under no obligation to not put serious poisons in this, it’s sold as gas line antifreeze and if they think some other chemical would help that primary role, there is no way you will know that. Methanol is a serious poison, don’t take it lightly, university labs do not allow you to even open a bottle of it unless you do it under a ventilation hood, just because some backpackers treat toxins so cavalierly doesn’t mean you should. Methanol burns well at higher altitudes however, so there are times it’s good.
- Everclear 95%, 190 proof – not legal to sell in the following states: There are several states that don’t sell “full strength” everclear. These are Ohio, California, Minnesota, Pennsylvania, Michigan, New York, Virginia, West Virginia, Washington, Hawaii and Florida. Don’t waste your time with the 150 proof stuff, it won’t burn well. Expensive, and if the racing fuel above works, I see little reason to use everclear, at $15 to 20 a fifth it’s very expensive fuel.
Appendix 2: table of fuel required to boil 2 cups water methanol/ethanol
Here is a basic table that shows the boiling times for the standard temperatures you’d expect to see backpacking in non winter conditions. All amounts assume boiling 2 cups of water, 1 pound, at sea level (212F), and a 45% stove efficiency. Keep in mind that the volume 2 cups of water weighs 1.04375 pounds, we we’re multipling the outcome by that number before then calculating the actual fuel used.
The formula is this: amount of fuel required to boil 2 cups at 100% efficiency/actual efficiency of stove. like so:
9/x = .55, or x = 9/.55 = 16.4 ml (x is the actual fuel used)
|Water Temp||Ethanol||Methanol||50% m/40% e|
|70 F||12.7 ml||17 ml||14.5 ml|
|50 F||14.5 ml||19.5 ml||17 ml|
|35 F||15.9 ml||21.3 ml||18.6 ml|
50% m/50% e approximates something like SLX or Crown assuming their msds is reasonably close to reality.
55% efficiency is not difficult to achieve, and does not require any bulky and hard to pack cone device, and the stoves are not hard to make. All numbers are aproximate and based on non pure fuels I have access to. Actual efficiencies for ethanol require testing with straight ethanol. HEET is adequate for testing for straight methanol. However, these numbers are close to what you can expect to get with standard denatures alcohols or HEET.
Ethanol, at 21.1 kjoules per ml, contains 33% more energy per ml than methanol, at 15.8 kjoules per ml. And, turning it around, methanol contains only 75% of the energy of ethanol. So anyone who tells you that it doesn’t matter what type you use simply has no idea what they are talking about. SLX will contain roughly 18.3 kjoules per ml.
However, because pure ethanol has a lot more energy, it’s actually harder to make a stove that will burn it efficiently, but you can tune your design to optimize it by allowing for a lot of air flow, and probably some other features.
Appendix 3: Chart of boiling temperatures at different altitudes
Unlike the above complicated formulas and calculations, the altitude formula is very simple: at 0 feet altitude, water boils at 212 F.
Every 250 feet in altitude removes about 0.5 F from the max heat of the water, so .
Keep in mind, if you use a real lid, you can pile a rock on your lid and make the pressure in the pot higher than it would be without a real lid, thus probably saving yourself far more in weight of fuel used than you’d save using an aluminum foil lid or whatever zany idea ultralight backpackers come up with. Just give me a lid I can put rocks on to weight it down and I’m happy.
You can use this chart of boiling points/altitude (shortened to just show the normal altitudes we’d expect to hit on our hikes, click the link if you want to see the lower or higher numbers). Short story, 1000 feet, water boils at 210F, 2000 – 208F, 3000 – 206F, and so on. That’s why it’s hard to get water hot when you get higher up.
|Altitude compared to Sea Level||Boiling Point|
Appendix 4: Units of energy conversion
agmrc.org shows us a variety of useful measurements, which I will quote here just so you know:
1 joule = 0.0009478 Btu 1 Btu = 1,055.05585 joules 1 kilojoule = one thousand joules 1 kilojoule = 0.947817 Btus 1 Btu = 1.055 kilojoules 1 watt hour = 3,600 joules 1 kilowatt hour = 3.6 megajoules 1 megajoule = one million joules 1 terajoule = one million megajoules 1 watt = 1 joule per second 1 watt = 3.412 Btu per hour 1 Btu per hour = 0.293 watts 1 kilowatt (kW) = one thousand watts 1 kW = 3,412 Btu per hour 1 kW = 1.341 horsepower 1 horsepower = .746 kilowatts 1 megawatt (MW) = one million watts 1 gigawatt (GW) = one billion watts 1 terawatt (TW) = one trillion watts 1 GW = one million kWs
Appendix 5: Standard chemical calculation
Q = m x c x T (where Q is the heat energy in Joules, m is the mass in grams, c is the specific heat of water = 4.184, and T is the change in temperature). By determining the temperature change for the given mass of water that is being boiled, we can calculate the heat released by the fuel. This is much more useful than saying 15 mL of ethanol boils 2 cups of water, etc., etc., because that discussion doesn’t include the initial and final temperature of the water. If one wants to talk about the efficiency of certain fuels, the most useful unit is kJ / mol. By knowing the appropriate chemical equation, you could probably look up this number in a table somewhere and given the # moles, you could determine the theoretical heat yield. But since not all that energy is transferred into the water, it’s also useful to talk about the actual yield, or percent yield. Or just grams of fuel consumed / kJ heat energy transferred into the water. This is where the efficiency of stove designs comes in. src
T is in Kelvins. Use this fahrenheit to kelvin table to find out that say, 70 f to 212f is 294.26 k to 373.15, or a change of 78.89 k