Selecting a Canister Stove for Cold Weather BackpackingPart I: Stove and Fuel Fundamentals

[Ryan Jordan]

Companion forum thread to:

Selecting a Canister Stove for Cold Weather Backpacking
Part I: Stove and Fuel Fundamentals

[kevin davidson]

This is the kind of article series that should epitimize BPL journalism. First the theory, then get down to application ( w/ promises of modding tips). I look forward to part 2.

Good show, Roger.

[Michael Martin]

Bravo, Roger!

I also can’t wait for part two.

Cheers,

-Mike

[Rick Dreher]

I’ve had canisters in my car trunk in summertime without its exploding into a ball of flame, and it surely gets into the 130-140 range in there, but prudence prevails anyway. Perhaps the important point is that a system heating the canister to feel warm to the touch could well continue to increase the temperature further, to the danger point. Intervention might be warranted before that can happen.

I agree that it’s impossible to compare dry metal surface temperatures to water temperatures. We’ve all picked up items far warmer than 105 without burning our hands (if briefly!). I can’t handle a hot tub over 105, but have sauna’d much hotter than that.

Could one “over-shield” the cartridge for cold conditions? Probably.

[paul johnson]

Rick, Does your trunk really get that hot? I wonder. The interior of a car gets much warmer than a trunk, typically. It’s the sunlight that penetrates the glass in the interior (passenger compartment) that causes the 140deg temps. The trunk is generally cooler. Check it out sometime after your car has been standing still in the sun. You will notice that it is much cooler in the trunk than in the passenger compartment. A simple test which many reading this may have done is to transport their old LP records ( not that they played them in a car) in the trunk of a car. Had either been stored in the passenger compartment, in the sunlight, they would probably be ruined. But this doesn’t happen in the trunk – to my knowledge. I did however, as a teen, transport a stack of LPs on the floor of the backseat. Left the car in the sun for an afternoon and the LPs were ruined/warped. Subsequent to that, I transported several times LPs in the trunk of my car and left them there throughout an afternoon in the sun with no damage to them.

Please don’t misunderstand me. I’m NOT saying that a trunk can’t get hotter than the OAT. I’m only saying that it generally doesn’t reach the same temp as the passenger compartment.

My point of this is I don’t think that the trunk test is a valid test if we’re trying to get to 140deg canister temp. Maybe we’re getting to 115 or 125 at most – frankly, I don’t know.

[cushing hamlen]

You know…I hate to say this, but the question here could be tested. (I hereby issue the required “do not try this at home…it could be dangerous” warning, and that by saying what I am about to, I in no way encourage anyone to actually do this….it is just an intellectual exercise. What we are REALLY interested in is what pressure can one of these cylinders support without rupturing. As the cylinder heats, the liquid gas inside will be shifted to a higer vapor pressure (that pressure being dependent on temperature). This suggests the following experiment: Take a cannister, run it until it is almost empty (say 1/2 oz of fuel or so), attach a thermocouple to it, run the wire a sufficient distance away (preferably behind a tree), attach to an appropriate meter, light the stove, arrange a windscreen around it, return back to the safety of said tree (making sure nobody else is or can go anywhere near the thing….then track the temperature and see if/when the thing lets loose. Again…take this as a thought only – not a direct suggestion….undoubtedly with more thought a safer method of doing this experiment will come to light…..

[paul johnson]

“safer method” – Nah…, testing is often done to destruction/failure. Please report back your findings. I would suggest multiple samples, so that any manufacturing “defects” in one particular canister can be somewhat statistically eliminated. Go for it! Can’t wait to see your test results.

You know, if a max. safe canister temp is determined, then it might be an easy matter for someone to manufacture a small light weight thermo-chromatic “Go No-Go” temperature indicator that can be stuck/afixed to the canister via an adhesive backing. When the temperature indicator changes from black or dark-green to red, the canister must be shielded or the burning halted. Not sure of the safety margin we would want before maximum safe temp is reached?

[cushing hamlen]

Paul, Oh I agree – testing to failure is a necessity….by “safer” I was referring more to safety of life-and-limb.. :-)

[Dondo .]

Thanks for the feedback, Neil. I may tighten up the windscreen and ditch the reflector, at least for cooler weather. In cold weather, say below the mid-twenties, I carry the effective, but much heavier, Coleman Xtreme. I’ve been paying attention to Bill Fornshell’s threads and am curious to see what Roger Caffin has come up with. Which hose-feed stove do you use and how do you support the inverted canister? Thanks.

Dondo

[paul johnson]

Understood. Your suggestion is always a good idea, especially if it is your life and limb on the line!!

[Drowned Lemming]

I use a Windpro, because it uses ‘standard’ cannisters. Regardless of what Coleman may say, Xtreme cannisters are not readily available in UK/Northern Europe, while the ‘standard’ versions are available in remote areas. As cannisters cannot be taken by air, it makes a lot more sense to use what is locally available!

The inverted cannister is held up by three pegs and a thick rubber band – total weight less than 25g.

[Rick Dreher]

Hi Paul,

Hard telling actual temps reached in there, but based on how hot items are when I retrieve them, I’ve got no doubt that under the “right” conditions, my trunk gets very, very hot (e.g., on a 110-degree day in an open asphalt parking lot). Given sufficient time the temps will tend to equalize throughout.

To follow your point, it may well be that one of my previous rides could get hotter in back than my present one because they were hatchbacks (even more glass!).

(Come to think of it, since more than half of passenger vehicles sold are “light trucks,” trunks aren’t terribly common anyway–everything is technically the passenger compartment.)

One of many lessons I’ve learned the hard way is get the wine home quickly in the summer. It’s remarkable how much liquid will force its way past the cork. Doesn’t smell as good as one might think, either.

To bring my tangent back to planet earth, I’ll wager that some percentage of fuel canisters inadvertantly get very hot during routine handling and transportation without spectacular failure (e.g., freight car parked on a siding for a few days in Timbuktu). I further speculate that it’s more likely that one would melt the melty bits on a stove well before they’d achieve the temps necessary to burst the canister.

[All standard disclaimers about this being an uninformed opinion in no way reflecting the opinions of the staff and management of BPL.com should be applied here, and should be consumed along with a grain of Dean and Deluca sea salt.]

FWIW the spectacular stove failures I’ve been privy to have all been safety valve releases on Primus, Svea and Optimus self-pressurizing white gas models. Wahoo!

[paul johnson]

Rick, You got me doin’ the “Ropa-Dopa” – Good points. BTW, never owned a hatchback. I’ll keep your advice about rear storage area temps in mind if I ever get one.

[Bill Fornshell]

English translation with Computer Translator:

ikaros

Yukio-san’s Stove Site

[Anonymous]

Hi guys,

Any possibility of an ETA on part 2? I do look forward to it!

Brian

[Carol Crooker]

Tonight!

[Donald Horst]

My pan and stove are like the picture from Donald B, but my shield is simply one flat sheet of double foil that sits on the ground. As Neil says, over heating is not a problem for me. I never let the can get more than slightly warm to the touch. After all, the stoves work fine at room temperature.

I ALWAYS use the shield because it takes
almost no effort, shortens the time to heat my water, and saves a lot of fuel. It is almost like a Jetboil, but a lot lighter. :-)

Re water temperatures, I am a hot springs fan and usually carry a thermometer. For those of you familiar with Iva Bell/Fish Creek near Mammoth CA, the water at the top of the upper meadow comes out into a little pool that stays about 110. I have seen someone dunk in that pool for 10 or 20 seconds, but it is not for me. The immediately adjacent pool is usually kept at 105 – 107. At 105, I can be comfortable for 5 – 10 minutes at a time. At 107, it feels very hot getting in, and I cool off for a couple of minutes after about 4 minutes in the water.

Note that time to scald is nowhere close to linear with temp. At 120, it takes several minutes to do real tissue damage. At 130, skin damage is almost instant. That is why some people now recommend 120 for water heaters.

[Roger Caffin]

> 1. Here’s a common comment one reads in this regard (i.e. “how warm is too warm”): if you can’t keep your finger on the the canister, it is too warm. Now, I don’t really know how valid that “test” really is. Certainly, I would agree with it as stated, but does that mean that if one can keep their finger on a can, that it is ‘ok’?

The standard physiology rule of thumb is that a normal person can keep their hand on something at 40 C (104 F) but over that there is an ‘ouch’ response. This temperature IS SAFE for a canister.

OK: at some stage in the future I will be writing an article on canister safety and maximum temperatures. I do have some info from manufacturers on the safety testing procedures, and I can confirm that they all routinely test at temperatures higher than 40 C.

But I’ll leave the details of the article until later. Sorry!

[paul johnson]

Dr. C,

Many thanks for the reply. Good info on this canister temp issue. Enjoyed the articles on Stove Theory. Can’t wait for more from you on any topic you’d like to write on (enjoy your insightful, common sense Forum Posts on various topics). Tech articles, like yours, L/UL philosophy and technique, and gear reviews also (don’t like wasting my money on gear that will disappoint) are the reasons I was first attracted to BPL.

[cushing hamlen]

Dr. C. – thank you also for your thoughtful and informative pieces. Question: given your statement that you have information from manufacturers on cannister testing, is there any need for me to go ahead and instrument up a (nearly empty) cannister and test to failure? I could do this, but: I have in-hand the vapor-pressure curves for both propane and isobutane…so if you have a value for maximum pressure, we can get a pretty good ballpark figure for temperature on a non-empty cannister…..

[Brian James]

Regarding canister failure testing, the “failure-point” temperature would actually vary based on fuel level and fuel composition. Testing an “almost-empty” canister would yield misleading results: you would need to test a full, high-propane canister that is of the weakest design and construction.

In order to know the pressure exerted by our fuel in a closed system like a gas canister, we need to consider the pre-existing system pressure as well as the available volume.

The article correctly points out the boiling temperature of the related gases, but that’s most likely at SATP. (That’s a standard reference conditions of temperature and pressure when discussing chemical properties of a substance.) The thing is that our gases are not *at* SATP: there under significantly more pressure, giving them differing boiling points and different properties in general.

Container failure will of course occur when the force exerted by the contents’ attempting to boil exceeds the canister’s burst pressure. This effect will be greater when the canister has a) less empty space, and b) more propane in it.

Thus, we would need to conduct burst temperature tests on a *full* canister (with a high propane content) in order to establish a true safety-margin burst temperature for canister stoves. Testing an almost-empty canister or one with less propane (or even a stronger canister design) would give misleadingly high burst-temperature data.

Further, different brands (and even manufacturing lots) of canisters will have different burst strengths and will be filled with (slightly) varying mixtures and volumes.

I work at a manufacturing facility for high-pressure expansion joints, and there is more variation than one might think: All burst pressures are kept well above the spec, (usually 200% of the spec value or more) but their burst pressures can still vary by 100%+ with no change in procedure. Metal forming is a delicate art and tiny, acceptable changes in tolerances will increase or decrease strength exponentially.

Hmm… sounds like enough variables to be the subject of a BPL article series — or maybe Part 4 of 3 in this series? A safer way of testing would involve connecting empty canisters to high-pressure inert gases and pressurizing them to failure point. Once a failure *pressure* was established, a Chemistry/Chemical Engineering major (not me) could calculate the temperature at which a 30% Propane + Butane mix would reach that pressure, given defined quantities of gas and volume.

Brian

[paul johnson]

Go for it, my man – and…report back the results, please. There are many who would like to know of a conservative safe temperature.

[Roger Caffin]

> Regarding canister failure testing, the “failure-point” temperature would actually vary based on fuel level and fuel composition. Testing an “almost-empty” canister would yield misleading results: you would need to test a full, high-propane canister that is of the weakest design and construction.

RNC: Um … Not sure that this is 100% correct.
Any liquid, including propane and butane, will have a standard pressure/temperature relationship. It does NOT matter how much liquid is left in the tank, as long as there is some. However, butane and propane will have different values for this relationship. So a pure Propane liquid would be the more severe test, but only by a small margin.

> In order to know the pressure exerted by our fuel in a closed system like a gas canister, we need to consider the pre-existing system pressure as well as the available volume.

RNC: You can calulate the internal pressure from the temperature, using the vapour pressure relationship.

> Container failure will of course occur when the force exerted by the contents’ attempting to boil exceeds the canister’s burst pressure. This effect will be greater when the canister has a) less empty space, and b) more propane in it.

RNC: I will have to disagree here about the ‘less empty space’ bit.

> Further, different brands (and even manufacturing lots) of canisters will have different burst strengths and will be filled with (slightly) varying mixtures and volumes.

RNC: Fortunately, the Safety Standards do set minimum required limits.

[Roger Caffin]

> Take a cannister, run it until it is almost empty (say 1/2 oz of fuel or so), attach a thermocouple to it, run the wire a sufficient distance away (preferably behind a tree), attach to an appropriate meter, light the stove, arrange a windscreen around it, return back to the safety of said tree (making sure nobody else is or can go anywhere near the thing….then track the temperature and see if/when the thing lets loose.

RNC: patience, patience … :-)
Hum – I wonder if Echelon will pick this up?

At this stage I will just repeat what I said earlier: if you can hold the canister, it’s safe (at least as far as over-heating goes).

[paul johnson]

Dr. C,

agreed. if you can hold it, it is safe. no disputes there. this simple qualitiative test is fine, but you’ve got the techno-geeks among us (well, at least me) really curious. we(?)/i want hard, quantitative numbers on approx. burst temp.

I’d love to see these numbers and a “go/no-go” stick-on temp indicator for use with the canister.

What’s your opinion on the pros & cons of using a DrJ pie-plate under the burner in the cold? Sure, it prevents canister over heating, but it also doesn’t allow the canister to be warmed by the burned gases.

[Author]

Posts