Companion forum thread to:

Exploding Gas Canisters: The Hazard of Overheating

I have not tried this yet to see how well it works. I bought a very flexible silicone hot-pad. It was square and I cut it round to fit inside my pot. I cut a hole in the center slightly larger than the lip of the fuel canister. When operating, the silicone should protect the cannister from over-heating. I have not had the nerve, yet to see if a radius cut should be made from outside to the center hole of the silicone. I may do this after a few trials and hope that the silicone remains rigid enough to stay in place. If it does, I can remove it when the stove is off and use it as intended, as a hot-pad. The thing cost about $8.00 at Target. A less flexible one, also available there, surely would survive the radius cut but it is more difficult to grasp hot cups and pots with.

Too bad you didn't have a video camera rolling... I gotta say (not that I endorse it in anyway or suggest you actually do it), if there had been a little lit candle setting on top of your "chamber" you would have had an amazing effect... ;)

Anyhow, very cool Roger, very cool...

I wonder what you with technical knowledge think about the safety of running the new jetboil and reactor stoves at full tilt, or in any conditions that might stress them. Are they safer in some respects or what?

Evan

Hi Roger,

Thankyou for a well written informative article, as an amateur stove designer I found the information in the article very usefull. It goes along with the episode of Myth Busters that was on TV in Australia last week about the myths of exploding cigarette lighters in hot cars and around Arc welding.

Tony

Excellent article and many thanks to Roger Caffin for his thorough studies, research, and testing on this matter. The findings actually support my own suspicions about these canisters: that the bulging bottom may cause failure of the crimp before obtaining a new (temporary) higher volume intended by the design. The canisters are intended to contain the explosive gas in such a situation and some of the manufacturers listed this requirement in their own testing descriptions.

Yes, I am very satisfied that the canister reached 98 degress Celsuis (perhaps near 400psia but with all of the author's caveats about interpreting the results) before failing and I recognize that all of this information and testing implies that all of the available brands are safe under even fairly extreme situations. However, I would acutally prefer that the bottom evert at a *lower* pressure point if it would help contain the fuel. While the popping bottom is still a risk to the user (I certainly don't want one of these things shooting at me), it is a lesser risk than combining high-velocity canister/stove/pot/scalding water with a BLEVE to finish the user off. If the bottom attachment cannot survive the violence of the bottom evert, then the evert serves no real purpose. If the fuel escapes the container, there is no point to the expanding bottom whatsoever.

FYI: aerosol cans, especially ones containing "spray adhesive" or something similar, make very nice fireballs which often rise upward in a nice mushroom effect. Ahhh, the memories.

Hi Stuart

> I bought a very flexible silicone hot-pad. It was square and I cut it round to fit inside my pot. I cut a hole in the center slightly larger than the lip of the fuel canister. When operating, the silicone should protect the cannister from over-heating.
Well, it does not look much different from my heavy alfoil radiation shield, and that does work.
Without making any promises, can I suggest you try it for a meal, but monitor the temperature of the canister underneath the silpad while doing so.

Now, the slit so you can get it off. I don't know how floppy the silpad is. You could of course experiment, and see about hooking the edges together somehow?

Hi Bernard

> I wonder what you with technical knowledge think about the safety of running the new jetboil and reactor stoves at full tilt,

The Jetboil stove was made by Primus for Jetboil, and was made deliberately under-powered for safety reasons (to Jetboil specs). Flat out does not seem to be a thermal problem for that stove, but its carbon monoxide performance is poor.

The Reactor is thermally OK at full power, although that is an awful lot of power. There will be further (extended) discussion of its performance in the Part 3 of the CO series. It's NOT good.

Roger thanks for confirming what I already suspected re gas canisters. I have read most, if not all of your articles on gas stoves and canisters.
My current setup is a Kovea Titanium stove and primus canister with a Snow Peak titanium pot. I always use a heat shield on the canister (circle of aluminium flashing the diameter of a gas canister with a hole in the middle). I then wrap an aluminium windshield tight around the pot, leaving a 50mm gap for the controls and for the air. There is no gap between the windshield and the pot. I always cook on the lowest setting the stove will work at. With this set up I am able to cook a three course meal for two on 26 grams of gas. The meal is usually - boil water for two Cups of soup (500ml), boil 500ml water to rehydrate meat and veggies, boil 500ml water for mashed spuds, boil 300ml water for custard and Mint Slice dessert, and occasioanlly 500ml water for two hot Chocolates. The trick to it is to keep the wind out and run the stove on the lowest setting. Saves gas and weight! Since I have been using this setup the canister has never gotten too hot to leave my fingers on, so I figured it was OK.

Edited by colinpbriggs on 05/03/2007 01:10:11 MDT.

Hi Colin

> I then wrap an aluminium windshield tight around the pot, leaving a 50mm gap for the controls and for the air. There is no gap between the windshield and the pot. I always cook on the lowest setting the stove will work at.

Interesting. I have never used a windshield like this: I always have a 15 - 20 mm gap, but often a much narrower slot at the front for the controls. I have no idea which one would be more efficient.

Yes, I (obviously) agree totally with your comment about running at a low setting for efficiency. Your gas figures sound good to me, although I am not sure anyone who eats custard and Mint Slice dessert should be allowed in ... :-) As for Hot Chocolate after dinner? Oh, the decadence! (Errr --- I may have done it in winter once or twice - too hot for it in the Australian summer though.)

However ...
> Since I have been using this setup the canister has never gotten too hot to leave my fingers on, so I figured it was OK.
This is the message I want everyone to remember! Touchy Feely :-)

Cheers
Roger

Roger,

Your and your colleagues level of professionalism is a big part of what makes BPL a worthwhile informational resource. I especially enjoyed this tidbit:

However, while the Standard specifies that the base should evert before the canister bursts, it does not specify how long beforehand. In this case I estimate the interval was in the millisecond range - the bang was very loud.

Roger,
what influence has the amount of fuel inside the canister on the temperature at which the bottom of the canister pops out? You used 33 grams of the gasmixture but what if you would do the same with a nearly full canister (about 200 grams of fuel I assume). If it doesn't have an influence, why? If it does have an influence, again why?

Or is it as simple as in applying the ideal gas law (PV=nRT) in this form:

(PV)/T (initial) = (PV)/T (final) with V = constant

Edited by Woubeir on 05/03/2007 15:11:55 MDT.

Hi Tom

> what influence has the amount of fuel inside the canister on the temperature at which the bottom of the canister pops out? You used 33 grams of the gasmixture

Yeah, this bit gets technical.

If there is liquid fuel in the canister, the pressure of the vapour is set by the laws of physics - the so-called 'vapour pressure'. The vapour pressure of the liquid is set by the temperature, as shown by the graph in the article. The key thing here is that it does NOT matter how much liquid is in the canister, just that there still is some liquid. I know some people have written otherwise, but they are wrong.

Of course, once your canister gets very low you can't be sure what the remaining liquid fuel is: 30% propane, or maybe just 10% propane, so there is some uncertainty about the pressure. Under most any circumstances I can think of, the concentration of propane will not be greater than the initial concentration, so the pressure may be a little lower than predicted.

In the case of the experiments described, the liquid in the canister would have been very close to 30% propane and 70 % isobutane, because I refilled some empty canisters from a new canister of known mix. However, refilling canisters is complex and something I strongly recommend you do NOT try. I only did it for this article; I do NOT do it for normal use.

If the amount of fuel in the canister is small, such that it all turns to vapour as it heats up, then the pressure will be different - lower in fact. Then and only then does the PV=nKT equation come into play.

Hope this helps.
Cheers

Edited by rcaffin on 05/03/2007 15:23:51 MDT.

Yep, this helps.
Somewhere in my mind I already assumed this, but it helps if someone more knowledgable can confirm this.

Thanks.

"If there is liquid fuel in the canister, the pressure of the vapour is set by the laws of physics - the so-called 'vapour pressure'. The vapour pressure of the liquid is set by the temperature, as shown by the graph in the article. The key thing here is that it does NOT matter how much liquid is in the canister, just that there still is some liquid. I know some people have written otherwise, but they are wrong."

Damn-straight... That's part of what makes the 'everts before failure' kind of stupid... everting will reduce the pressure maybe 10%(?) (approximately whatever the change in volume is from the eversion)... but ONLY if the can contains only vapor.

"If the amount of fuel in the canister is small, such that it all turns to vapour as it heats up, then the pressure will be different - lower in fact. Then and only then does the PV=nKT equation come into play."

Yeah, but you'd have to have an absolutely minute amount of liquid in the container at lower temps to begin with. I can't remember my ratios exactly, but at the temps / pressures we're dealing with the specific volume (volume per weight of material - basically the inversion of density) change during a phase change is somewhere in the neighborhood of 30:1 or 100:1 I believe, meaning you would have to have only a few mls of liquid left in the container.

Roger, I have learned so much from your articles! Thanks so much for all the effort you put into them! Like others, I also found the failure mode very interesting. Knowing that you can't count on an everting canister as a sign of failure is huge and should encourage more vigilance from those who might have taken that "safety" feature for granted. It's also interesting to note that the everting bottom may launch the exploding canister and stove into the air to a height that might maximize the effect of the explosion on those around. Do you recall about how heigh off the ground the canister was when it exploded?

Hi Joshua

> Yeah, but you'd have to have an absolutely minute amount of liquid in the container at lower temps to begin with.
Correct. In a 220 g canister you would need less than about 20 g of fuel for it to all turn to vapour. I deliberately had such a small amount in the first canister I heated.
For the second canister I calculated that about 30 g would leave some liquid at 100 C, so I think I arranged to have about 33 g left in the canister. It blew.

Hi Eric

> Do you recall about how heigh off the ground the canister was when it exploded?

Er ... about one foot WHEN it exploded...
But how high dd it go? Ah well. There was the bang, which always produces some shock and recovery time in the observer. I DO recollect that I heard the canister hit the ground some time later - maybe several seconds later. (So I knew in what direction e to look for it later.) It must have gone at least 100 feet in the air.
The bang was a LOT louder than a rifle shot!

Yeah, the 'evert bottom' idea might have seemed like a good idea at the time (when they wrote the standard), but in practice ... :-)

I could be wrong, but I think that given a height when seated of about 3 feet, an explosion at about 1 foot pretty much maximizes the amount of damage to anyone seated in close proximity to the stove. That's quite a safety feature :o.

Hi Eric

> given a height when seated of about 3 feet, an explosion at about 1 foot pretty much maximizes the amount of damage to anyone seated in close
Well, with any luck it would go straight up, and you might be to one side.
What WOULD cause you some damage imho would be the bang. At that distance I would expect some damage to the eardrums!

Cheers