Stove Testing and Humidity

So in the conditions when you need warmth the most, in the cold and wet, is when alcohol stoves perform worst. I assume esbit and canisters suffer similar issues, but I wonder if there's a "poor weather" winner. Probably canisters since you can just up the BTUs but sacrifice efficiency.

Dan is certainly an expert and knowledgeable about his stove designs. With respect to low barometric pressure and humidity, this may not affect all alcohol stoves the same (or at all). Humidity and temperature do not have a significant impact on our FLAT CAT Stove designs. We have tested these stoves to +12,000 feet, low temperatures, low & high humidity as well as in the rain & snow. Interestingly enough, I live close to Camarillo, CA and often experience the climate changes due to the coastal influences. During this time of year, we often experience daytime temperature well over 100 degrees and humidity in the 10% to 20% range. Then in the same week, the temperatures will be down in the 60 F with high humidity (for California). Through all of this, our stove performance remains pretty stable. So, as I said earlier, Dan’s comments are correct for his stove designs but may not apply to all alcohol stoves. Best regards – Jon

Sometimes Extreme cold has an affect on stoves.

Dan, it’s interesting you decided to post this. I’ve also been noticing an inexplicable change in boil times/fuel amounts lately, for the worse, I had wondered if it was the warmer weather coming around (sf bay area) but haven’t been able to pinpoint it to anything concrete yet. Starting fuel temp, and as you speculate, pressure changes/humidity, should account for a meaningful change, measurable.

With these stoves, there’s so many variables, particularly in screens, shape, height relative to bottom of pot, top of pot, distance from pot, pot width, fuel composition, cone, non cone, height of stove/pot distance, differences based on starting fuel amount, for each type of design, etc.

Here’s an example of a good text file log of stove tests from penny stove creator.

I think the thing that I find most intriguing about playing around with these little stoves is the extreme variability of results based on what are apparently small tweaks to designs, it’s getting clear to me that the the stove/pot/screen/stand however those are configured are one thing, and if you change one part there, you then have another thing, which will have different behaviors which are hard to predict.

I do wish that everyone would use exactly the same calibrated measurements and water amounts so that comparisons would be easy but only a few people out there do so it’s hard to really accurately gather/compare data.

I, sad to say, also recently gave up and started mainlining the fuel, ie, bought a gallon instead of the quarts I’d been getting…

Alcohol will absorb moisture from the air over time. The more water you have in your fuel the lower the stoves BTU output will be. Air pressure would affect the air fuel mixture which can also have a BTU effect. Wood will also absorbs moisture from the air so it is also affected by humidity.

Caister fuel is free of water inside the canister. White gas will not mix with water. So for white gas and canister fuels it is air pressure,wind, and temperature that have the bigest effect on performance. Solid betrolium based fuels are probably similar.

Basically air presure can effect the air fuel mixture which affects heat output. Temperature affects the fuel evaporation rate and for canisters, the pressure in the can. Cold temperatures may reduce fuel flow and heat ouput. Water when mixed with fuel will also reduce heat output.

When I use alcohol, the fuel is either in the fuel bottle or in the stove so water/humidity is not a factor. When assembled and light my cooking system, my stove and pot are contained inside the windscreen (I use a heat shield). The internal temperatures rise quickly so again, humidity is not a factor. No additional water in the fuel, no change in performance.
Does humidity impact the burn rate? I haven’t seen this to be the case. In cold temperatures, there is very little water in the air. Where I have seen a fair amount of water is condensation on the bottom of the pot. I believe that a lot of this is due to the byproduct of H2O during the burning process. Does it affect performance? I haven’t seen this to be that case.
With altitude, the amount of oxygen will decrease however; the ratio of nitrogen to oxygen remains the same. Does altitude impact the oxygen/fuel ratio and thus impact performance? I haven’t seen it in my systems. Our stoves are designed to deliver about 375 watts and our windscreens are designed to be “oxygen starved”. If the heat output (burn rate) were higher, it may choke the system. I have tested this system multiple, multiple times in Sequoia NP and Yosemite NP from 5,000 to 12,000 and have kept track of fuel usage, water volume, water temperature, barometric pressure and boil times. I can say that I see little variation in stove efficiency.
In my experience, cold temperatures (water/hardware) will change the fuel consumption and time to boil, but I have found that it has little to no effect on fuel efficiency. Fundamentally, there is some basic and predictable heat transfer dynamics that are going on that are very predictable and repeatable. I have found that the biggest impact by far in due to the ambient wind. That being said, the windscreen design is probably THE most important component in your cooking system.
As I said before, making general conclusions about alcohol stoves does not mean that all alcohol stoves behave that way. Dan has made comments on his stove designs and I believe that he is correct (about his stoves). I track a lot of stove data and have not seen humidity or barometric pressure have a significant (measurable) impact on our perfromances. My 2 cents – Jon

Hi Harald, thank you for your insight into the world of testing and creating.

Stove design has bee the most interesting hobby that I have ever had. Sgt. Rock over at Whiteblaze.net set the testing standards for me and many others.

My design goals were only two. Create a stove that will boil 2 cups of water with 1/2 ounce of denatured alcohol and the stove was to have an integrated pot support. I was impressed with SuperCat stove and it's the inspiration for my design goals.

When I begin testing a stove design there are 5 things that are to be constant:

1. My 5.5" diameter aluminum pot with a 2.5 cup capacity.

2. Starting water temperature at 70 degrees

3. Room temperature at 70 degrees

4. Small 1 ounce capacity graduated medicine cup for fuel measuremnt. Always using 1/2 ounce of Sunnyside Denatured alcohol.

5. Room air was to be calm, no breeze, which meant having to wait till the furnace or AC turned off.

The stove was tested by itself, no windscreen.

If I made a change to a stove design it was only 1 change at a time. For instance, I would not increase distance from stove to pot and number of jet holes and then test.
If the 1st test looked promising, I would do at least 2 more to make sure of the results. Always waiting for the stove and the surface it rested on to be cooled down befor continuing.

Having a single lever kitchen faucet is nice for easy access to 70 degree water.

Once you have a stove that repeatedly give the same performance under the optimum conditions as stated ubove, then you are ready to take it out into the garage and do some testing in high humidity, low barometric pressures and extreme temperatures. Then once you are satisfied, you post it onto a forum to show and tell to see how others like it. Once it leaves the garage, we need to do all we can to preserve the heat going to the pot in a controlled way. That means to protect it from the elements, wind and rain.

I can testify that I know for a surety that low barometric pressure and high humidy have an affect on the way things burn. There will always be non beleivers out there. Ask your nearest Forest Service Agency about low pressures and humidity and the affect it has on fire.

There are stove makers out there that are "perfect" I have not attained that status, never will. All I can do is try my best.

If your alcohol also has water, it has to be evaporated which takes heat which reduces efficiency of the stove

Jerry, that's correct. There are some stove manufactures that advocate adding water to alcohol. That is a no-no. Never add water. Keep your stove user friendly.

Low barometric pressure and humidity have an affect on all alcohol stoves. When testing a new design keep in mind the stove will give different results when taking it outside in the real world.

This might help some of us to have a better understanding why humid air has an affect on stoves:

"Understanding air density and its effects"

By Jack Williams, USATODAY.com
Humidity and air density

Most people who haven't studied physics or chemistry find it hard to believe that humid air is lighter, or less dense, than dry air. How can the air become lighter if we add water vapor to it?

Scientists have known this for a long time. The first was Isaac Newton, who stated that humid air is less dense than dry air in 1717 in his book, Optics. But, other scientists didn't generally understand this until later in that century.

To see why humid air is less dense than dry air, we need to turn to one of the laws of nature the Italian physicist Amadeo Avogadro discovered in the early 1800s. In simple terms, he found that a fixed volume of gas, say one cubic meter, at the same temperature and pressure, would always have the same number of molecules no matter what gas is in the container. Most beginning chemistry books explain how this works.

Imagine a cubic foot of perfectly dry air. It contains about 78% nitrogen molecules, which each have a molecular weight of 28 (2 atoms with atomic weight 14) . Another 21% of the air is oxygen, with each molecule having a molecular weight of 32 (2 stoms with atomic weight 16). The final one percent is a mixture of other gases, which we won't worry about.

Molecules are free to move in and out of our cubic foot of air. What Avogadro discovered leads us to conclude that if we added water vapor molecules to our cubic foot of air, some of the nitrogen and oxygen molecules would leave — remember, the total number of molecules in our cubic foot of air stays the same.

The water molecules, which replace nitrogen or oxygen, have a molecular weight of 18. (One oxygen atom with atomic weight of 16, and two hudrogen atoms each with atomic weight of 1). This is lighter than both nitrogen and oxygen. In other words, replacing nitrogen and oxygen with water vapor decreases the weight of the air in the cubic foot; that is, it's density decreases.

Wait a minute, you might say, "I know water's heavier than air." True, liquid water is heavier, or more dense, than air. But, the water that makes the air humid isn't liquid. It's water vapor, which is a gas that is lighter than nitrogen or oxygen. (Related: Understanding water in the atmosphere).

Compared to the differences made by temperature and air pressure, humidity has a small effect on the air's density. But, humid air is lighter than dry air at the same temperature and pressure.

---

In conclusion we can understand that water vapor in the air acts like all other gasses.. it provides partial pressure proportional to its content in the air. in other words, the moist air will contain less oxygen at the same pressure (the sum of partial pressures being the same, the more water in the air, the less oxygen for combustion).

Less oxygen for combustion occurs in California where Flatcatgear is located. Take with a grain of salt what they say about their stoves. Listen to what is said about humid air. It applies to all stoves.

Thanks for digging into it Dan, re the claims of all stove makers, I currently will believe only the claims I can validate personally in the real world, with reasonably well calibrated measurements. I’m not saying they are true or false, only that I only believe the ones I can see. This is now particularly the case after I am seeing such a variation in the same setup using hte same fuel based on some other factor, possibly weather/pressure/humidity.

However, any sealed heat/wind screen is going to give a few ml boost in efficiency I believe, that’s what my tests showed me, though during those tests, of course, I also learned that you can’t use a high output stove in those cones or you end up with burned cone. Aluminum is weird once it becomes a metal ash, heh. I believe the flatcat designs use a non conical method of trapping all the heat, a nice design by the way, I’ve dabbled with it, but, again, the heat has to be low and the screen high or you end up with roasted aluminum.

I do however agree, it’s great fun to look for that difficult to reach 15ml to boil 2 cups consistently. That quest certainly provides me with very welcome relief from work. I think I’m going to go back to 2 cups because that’s what everyone out there is actually testing so that’s a fine standard, 500ml is so close it’s not really that big a deal. Ounces however I’m not using because they are too big as units to be accurate, ml are easier to calibrate, a longish thinnish cover for an irrigation syringe I found proved to be the ideal measuring device because it’s thin enough to give a good space per ml in the height, so it’s easy to be very accurate.

Anything that uses measuring spoons or anything like that I do not trust at all re results because they are subject to the rounding / surface tension thing plus they aren’t accurate to begin with, so I now calibrate all the devices with actual irrigation syringes, 10ml, is a good size because the ml are very large looking and it’s easy to get it right. I had these lying around from my sawyer squeeze mod testing, though I now use a 20ml for backflushing.

I see you’re using the sunnyside fuel, which is one that I can’t find around here anymore, unless Lowes has it, but I don’t have a local lowes, but last I checked it’s mostly ethanol. I believe that has been rebranded as crown something or other, and I don’t know if it’s the same fuel anymore.

That means you are getting the slight boost from ethanol content that is high, ~90%, I believe, though again, there is only one way to actually know, and that’s to send it to lab, the msds is totally worthless in my opinion to actually know what is in something, particularly re percentages, a fact I discovered while doing some simple arithmetic on some and discovering that the percentages listed were mathematically impossible. Well, arithmetically is all that is required.

Just to add an incidental observation, I believe that Kleen strip green is not what it claims, because it burns closer in appearance to the e-nrg I tested and discovered was in fact isopropynol or whatever denatured, which leads to yellow flames, on purpose (ie, the goal is to get yellow flames in fireplaces, which then burn full as they go upwards, but if they are interupted by a pot bottom they create vast amounts of soot), the same, or similar, yellow flames I see with kleen strip green, which also does not seem to burn as hot as slx, though I’m not certain, but it certainly does have yellow flames. So I’m starting to suspect that the msds for kleen strip green is a lie, they certainly seem like reasonably convincing works of pseudo science fiction from my observations so I don’t think they can be trusted at beyond crudely ambiguous suggestions as to possible content re the can you bought that day.

Another observation is that the specific gravity by weight I get for slx is less than either ethanol or methanol, which is odd. I’ve measured that several times to confirm it, but it is consistent.

That’s I think one area that makes comparing results very difficult, since there is a roughly 10% energy density difference between methanol and ethanol, and those numbers themselves are for the pure substance, whereas real ethanol has 5% water content as well. So it’s hard to get the real numbers consistent.

For those interested, you can buy real racing methanol, 99%, and real racing ethanol, 98%, here. The prices are quite good, far cheaper than getting everclear, but the shipping is high. But they have bulk too, the 5 gallon container is quite reasonable with shipping per gallon.

I came across this on an old bpl thread on fuels, but this resource seems to have been forgotten so I thought I’d bring it back to light.

That is fuel you can make solid technical and empirical observations with because it’s formulation is designed for the actual energy content of the liquid, not to be a vaguely labelled stripping agent sold to clean stuff up or off.

What’s interesting is that if sunnyside for example is the roughly 90% ethanol it was claimed to be, and if slx is the 47/47% ethanol/methanol plus chemicals and water, by my rusty math sunnyside or other mostly ethanol fuels yield a roughly 5% greater heat content over meth/eth mixtures. Not a huge deal, about 1ml I think difference in efficiency, but it is a difference.

Realizing I just didn’t get the differences, I researched a bit and came up with this page on ethanol / methanol energy content. If there are any math or logic errors, feel free to post corrections, I want that page to be as correct as it can be.

I realize I have probably made some mistakes in the math, but the overall data seems fairly consistent. Most of the stuff I’ve read in the backpacking world wasn’t very accurate, but some things are hard to really pin down, like the actual energy content of standard ethanol, at 95% ethanol/5% water, but I’m assuming the data out there is referring to standard ethanol, what they use in fuel that is, and that’s it reasonably correct.

By this, the results surprised me, ethanol (21.1 kjoules per ml) contains 33% more energy than methanol (15.8 kjoules per ml), and methanol, 75% of the energy of ethanol, inversely.

In my tests, this now explains some of the anomolies I’ve been seeing, I thought the energy difference between methanol and ethanol was significantly less. I would guess from my testing that slx is somewhere around 60% methanol, 40% ethanol.

I’m ordering some of the racing e98 ethanol, it costs 30 a gallon including shipping, which is the same price as klean strip green by the quart, at $7.50.

This explains why slx was failing to boil at comparable efficiencies to klean strip green or other mostly ethanol fuels, like the old sunnyside, now crown next.

I included all the information I could remember/think of re where to get alcohol etc, the brands, etc, along with some sample data tables.

This is not a stove comparison page per se, but more just trying to document with actual references and science/chemistry the raw data for the fuels, then some samples based on easily achievable efficiencies, that do not require cones or other non flat screens.

It looks to me like sgt rock is using mostly clean ethanol, I tried making one of his ion type stoves and just don’t see how it could achieve those efficiencies with methanol or slx type fuels, the ion I made burns very slowly but isn’t all that much more efficient than my four jet penny.

I’m wondering about other factors re the fuel types and weather, but that’s going too deep into variables so I’m going to leave it alone.

Thanks Dan for noting you are using sunnyside, I’m finding small details like that are almost always present when excellent efficiencies are reported. I still haven’t tested on cones, mainly because I don’t particularly want to carry a cone around backpacking.

Hopefully this page covers most of the technical details that are generally ignored or skipped or missed, I’ll add to it as I have time / energy.