The Evolution of a Winter Stove – Part 1

[Stephanie Jordan]

Companion forum thread to:

The Evolution of a Winter Stove – Part 1

[Stuart R]

WOW – that looks like a serious investment in equipment!

I'm sure that you have been thru the same make-a-prototype/test-on-a-trip/refine-prototype cycles that I have, but in different circumstances. The one thing I have deliberately made different is the needle valve position. Why? With a really light stove that does not use the canister as a heatsink, the stove itself can get really hot with extended use. That means any o-ring in the valve can be subjected to high temperatures. Viton can handle fairly high temp, but I did not want to get even close to it's limit. So, I have the needle valve at the canister connector, but not the usual crude needle valve that comes with most remote stoves – one with a much finer thread that is actually designed to control liquid. Even so, a very low simmer is tricky with liquid feed, but there is an easy solution – use the canister upright for simmering! The gas flow is low, so there is very little evaporative cooling and little propane gets used. When more heat is required, invert the canister again. I use a narrow bore fuel tube, so there is very little volume of liquid within it. The other thing I prefer about the valve at the canister is that there is no fumbling under a pot of near boiling contents to make adjustments.

Not claiming that my solution is any better at all, just that different experiences can lead to different conclusions!

[Peter S]

And very interesting. Looking forward to seeing the end result.

[Eric Blumensaadt]

I'm still not convinced that using a canister stove in winter – inverted or otherwise – is a good idea.

For now I'll stick with my MSR Dragonfly multi (liquid) fuel stove and TD Sidewinder/Inferno wood stove for winter.

But maybe Roger's ensuing quest will result in a break through design beyond Coleman's engineering.

[Jerry Adams]

Great article – love all the details of how things work – looking forward to parts 2 and 3

If the valve isn't right close to the Lindal valve, then you'll lose a lot of fuel whenever you unscrew it. Maybe part 2 or 3 will resolve that.

[Nick Gatel]

Roger,

You are such a tease, and set up us up for the sequels. You must be reading too many mystery novels. Looking forward to the next installment anyway.

[Mike Cecot-Scherer]

Hi Roger,

Enjoyed your Part1 of the winter stove quest but have found a rather serious mistake in your referenced carbon monoxide article.

Carbon monoxide in camping stoves is caused by exhaust getting into the intake by recirculation currents. The exhaust is obviously very low in oxygen so it causes incomplete combustion despite the stove manufacturer's lean mixture. I got this information from a Professional Engineer that specialized in forensic reconstruction of combustion events (usually for use in lawsuits). BTW a Professional Engineer in the US is a multi-disiplinary engineer whose passed quite a gauntlet to be certified. He QUITE disagrees with the cold pot quenching theory of CO production.

You can easily test this:
1) put an empty pot on the stove, let it get red hot and see if the CO stops
2) deliberately cause some recirculation with an obstruction and see if the CO goes up

This explanation also appears to be completely consistent with less CO generation with raised pots.

Exhaust recirculation also can be seen to be a factor in hanging stove or lantern CO creation as both types form an exhaust pool at the roof from which some can easily get back down to the intake. It also conjures a nightmare scenario where someone operates a stove in their tent beneath a gas lantern – CO in knock-you-out-and-kill-you doses could be quickly produced.

Best,
Mike Cecot-Scherer

[Roger Caffin]

Hi Jerry

> If the valve isn't right close to the Lindal valve, then you'll lose a lot of fuel
> whenever you unscrew it. Maybe part 2 or 3 will resolve that.
A good point, but …
First of all, if the hose is much thinner than average, the volume of fuel in it is low.
If you put a filler cord inside the hose to displace some of that fuel, the loss is even less.
And if you think ahead a wee bit, you can turn the stove off at the canister 10 seconds before you are finished, in which case there may be no loss at all. This soon becomes automatic.

Yep, more soon in Parts 2 & 3. Part 2 has already been submitted, and Part 3 is close.

Cheers

[Roger Caffin]

Hi Stuart

> With a really light stove that does not use the canister as a heatsink, the stove
> itself can get really hot with extended use. That means any o-ring in the valve can
> be subjected to high temperatures. Viton can handle fairly high temp, but I did not
> want to get even close to it's limit.

Safety First!!!

Yes, but just how hot the stove gets does depend on what sort of stove it is and how it is used. Lots of testing needed, and done.

I suspect that if you put several inverted canister stoves next to each other under one vast pot or BBQ plate and ran them for half an hour that way, they might very well get much too hot. I know one such case where the tank on an MSR white gas stove actually blew up, causing severe injury. There were two of them side by side. But this is an unlikely scenario for UL walkers, and is easily avoided.

Under actual field use conditions almost none of my experimental stoves ever got too hot, as far as I can remember. The vortex burner stoves were a bit of an exception: they do get hot and they need to have the O-rings a long way away from the flame. One of those is shown in the second photo, bottom right. It was nice, but hard to make.

Cheers

[Roger Caffin]

Hi Mike

> Carbon monoxide in camping stoves is caused by exhaust getting into the intake by recirculation currents.
I have to disagree with you here. Carbon monoxide results from incomplete combustion: that's basic chemistry.

For there to be incomplete combustion one of two conditions have to apply: flame quenching or insufficient oxygen. In the case of a small camping stove being run in open air or with adequate ventilation, I find it hard to imagine that 'inadequate oxygen' could apply.

> He QUITE disagrees with the cold pot quenching theory of CO production.
Tough. He needs to learn some basic chemistry.

> I got this information from a Professional Engineer that specialized in forensic
> reconstruction of combustion events (usually for use in lawsuits).
Sigh. Meaningless, and valueless.

My wife was on a jury once in a lawsuit about an injury. Both sides had qualified/certified engineers as technical witnesses. The two engineers testified radically differently: one had calculated that the forces involved were X (within OSHA limits) while the other had calculated that they were about double that (well outside OSHA limits). They could not both be right.

Later on I sat down and did the full analysis myself and I could see what each guy had done, and why one of the QUALIFIED engineers was wrong. 'Experts' have been wrong before.

> You can easily test this:
> 1) put an empty pot on the stove, let it get red hot and see if the CO stops
I have done that test many times, with continuous measurement. Yes, when the cold metal in the flame gets red hot the CO production often drops to ~0. Except when the stove is something like a Reactor, for which there is no hope.

> Exhaust recirculation also can be seen to be a factor in hanging stove or lantern
> CO creation as both types form an exhaust pool at the roof from which some can
> easily get back down to the intake.
Abstract theory, and using a false premise.
Practice is that the hot air circulates, and can be seen to drift out of the tent door. (Steam goes with it – makes a good photo with a flash.) If the tent was sealed and there was no wind at all, then you would have a real danger. But under such conditions who seals a tent up?

Sorry, but your theory does not apply to small camping stoves being used in bad weather.

Yours
Roger Caffin (BSc Hons, MSc, PhD, consultant research scientist)

[Jerry Adams]

I noticed that your hose was much smaller than the other hoses.

I know what you mean about flexibility – my old MSR Whisperlite hose is so stiff it's a little difficult to get the fuel bottle and burner with pot on it postioned properly.

[Stephen Parks]

Hi Roger. In case you haven't already become aware of them in the course of your efforts, check out the services offered by protolabs.com. They have two services: firstcut, which is a direct CAD to CNC machining service, and protomold, which uses the same automated programming/machining service to make injection molds that they use to mold your parts. For most applications, they will make the mold from aluminum, but they recently started making stainless molds for higher temp plastics like Ultem and PEEK, which they now offer. They promote their service for lower volume injection molding, say up to 1,000 or 10,000 parts (how big do you think your stove market is?), but if you wear the mold out they will make another at no charge. The prices are reasonable as far as injection molding goes, and the turnaround time is excellent; "parts in one week". They have online tools such as online quotes and mold flow analysis if needed. Having the entire process handled start to finish by one outfit has its advantages. They are in the US, which may not be convenient for you, but you won't have to worry about your IP. You can also request their "Injection Molding for Dummies" booklet, as well as some physical design aids that illustrate best practice and their various finishes. Best,

-Stephen

[Roger Caffin]

Hi Stephen

Thanks! I will look them up immediately.
(Do you have any connection with them?)

Cheers

[Stephen Parks]

I have used them for their rapid machining service several times, but I have never had any injection molded parts made with them. Actually, I haven't had any injection molded parts made from anyone except for some high pressure electrical bulkheads, but that is complicated stuff. Most of the equipment I design does not lend itself toward IM. But I think it would be fun to work on.

One thing protolabs does not do is dimensionally inspect your finished parts. Their rapid machined parts are provided "as is", which is usually within +/-.005". They gave a presentation at my company recently, and it seems like some refinement of the molds is possible if needed, but I don't have any personal experience with that. I can send you some contact info separately if you would like.

Now, given that you have a CNC center on hand that can make small parts efficiently, and given the setup costs for IM (and the fact that you are considering those costs), you must have some very intricate parts or are planning on some significant production volume. I'm very interested to see your future articles!

[Roger Caffin]

Hi Stephen

I trawled through their web site with great interest. However, I suspect that they won't be able to help me. They make no mention of being able to handle threaded parts, and my design uses quite a few threads.

Technical digression to explain: to extract a threaded part from a mold you have to actually unscrew the bit forming the thread. This applies especially to female threads, and there are at least four of those. That's complex, both for the mold itself and for the gear surrounding the mold in the press.
Pity!

Cheers

[Roger Caffin]

Is there enough air circulation in a tent? Obviously not if you seal it up, but what about with the door open?

Here we have an illustration of what I was talking about: steam wafting out the door while I am cooking dinner. Well, yes, it was sub-zero overnight. Combine this view of where the exhaust air is going with the windshield wrapped around the stove to make a little bit of a chimney, and maybe the amount of recirculating air getting back into the air inlets could be predicted to be rather low?

Cheers

[Bob Gross]

"Well, yes, it was sub-zero overnight."

C or F degrees?

–B.G.–

[Tad Englund]

Great work Roger.
Today I was digging around my gear closet and saw my supply of Powermax canisters and wondered what type of stove I would purchase for winter when they ran out. I haven't seen a suitable replacement as of yet, until know….
Keep working on this, you seem to be working through all the issues, I hope it all goes well.
Tad

[Stephen Parks]

That is probably why so many plastic designs use self-tapping screws in smooth-bore holes. Cheap, easy, and won't loosen from vibration. But sometimes you need a real thread form. It can be tapped later, but that still isn't as nice as a molded thread, which isn't as nice as a molded-in insert.

[Eric Blumensaadt]

Roger,

Is it possible (apart from propane & it's attendant container weight) that ther is a different fuel suitable for cartridges or at least a better mix for cold weather?

Failing that could lightweight fuel container technology be devised to handle 100% propane? The ultimate goal would be refillable canisters form a BBQ propane tank.

[Mike Cecot-Scherer]

Hi Roger,

It's hard to know what way to write this. I was pretty sure I said that carbon monoxide results from incomplete combustion CAUSED by the fuel-to-oxygen ratios being shifted to too rich by low oxygen exhaust mixing into a stove's intake. You seem to be disagreeing with this most basic first-order stoichiometric chemistry in favor of a second-order effect flame quenching argument. Let me first point out that these two effects would not be mutually exclusive and one would therefore expect that there would be times when one predominates (see below for a tragic real-life example).

Exhaust recirculation is a a primary cause of carbon monoxide generation in many accidents – people die all the time from this. Cars don't produce CO until they're run in a closed space and the exhaust goes back into the intake. Ice fishing hut heaters work great until the occupant wants to stop loosing that last bit of precious heat and closes the vents – making the exhaust go into the intake. HVAC engineers have learned to be very very careful with exhaust.

Flame quenching is a second order effect. Congratulations on finding it. But I must point out that you have no actual data as to where CO is generated and destroyed in the flame. For all you know the reason that CO production drops with raised pots is exactly because recirculation currents are effectively stopped. Or goes up when the pot is lowered because of strengthened recirculation. A recirculation explanation seems to fit your data quite well.

I know recirculation can cause CO generation because I have made CO with a camp stove or two in just this way. And I've found that wind screens are a terrific way to cause these recirculation currents. Put them on: significant CO generated. Take them off: low to no CO. Isolate the intake from the exhaust and then put on a windscreen: again, low to no CO (all done with a pot of water on the burner). Seems pretty convincing.

Pooling of warm stove exhaust at the top of a tent is easy to verify. That pool may or may not mix well depending on conditions. Just saying it mixes well is not an argument. If you hang a gas lantern high in a closed-up tent and let it run for a while you will see CO production. Probably not a heck of a lot, I've seen 70-100ppm, but definitely not the "theoretical" amount one would find with a false premise and quite enough to give thinking people pause (and to make one worry what might happen if someone, for instance, ran a stove below the lantern).

Your pooh-poohing the advice of a VERY experienced combustion engineer is entirely unsupported and not at all the kind of attitude that the buying public wants to see in their would-be stove designers.

And then there's your "faulty expert" example which brings no information to the table (unless, perhaps, you were intending irony). Nobody knows how many experts that lawyer had to interview to find one that said what he wanted to hear. Doesn't matter either.

I can tell you this: that combustion engineer I referred to and I were asked by a lawyer to figure out why two people died in their tent. They were found in their sleeping bags, pink skinned, their faces laying in vomit – obvious carbon monoxide victims. Their tent was zipped shut but there was a non-closable vent right at the top. They had mis-assembled their lantern (very easy to understand if you saw it) so that the exhaust was being forced down toward the intake which resulted in a plume that registered 500-1000ppm CO. That's a monstrous amount of CO – enough to actually knock someone out. Their tent did not/could not ventilate enough. And we were shocked to find that a wind – which one would have expected to aid ventilation – could actually block outflow. So don't be so nonchalant about adequate ventilation. What happened next? Nothing. The lantern company had gone bankrupt. The wives and children of the victims got nothing.

Looking at the whole outdoor stove configuration, it's kind of preposterous that it normally works so well. I mean, here's an open flame blasting away within a few inches of the intake which is sucking air for all it's worth. The flames and exhaust mostly go up and out – no question – but it's hardly a failsafe design.

Mike,
(just Mike)

[Jerry Adams]

You're saying that CO is produced when the intake air has a reduced Oxygen level?

[Roger Caffin]

Hi Eric

Yes, 100% propane would be much better, but the pressure is much higher and you have a safety issue with the Powermax-style Aluminium containers. This problem is not insurmountable, as you can buy such canisters with 100% propane in Europe. I forget the brand right now, but they are allied to Primus. I have tried to get some, but they can't be bothered shipping them to either Australia or the USA. DoT regulations etc etc. SNARK!

I suspect that they have simply made the containers with a heavier wall thickness. That should be enough to get the higher tensile strength to handle the pressure.

Cheers

[Nick Gatel]

CO is produced when there is incomplete combustion of any hydrocarbon fuel. Perfect air/fuel mixture produces no CO, but this is almost impossible to do. With a backpacking stove there are too many uncontrollable variables.

In a car, perfect combustion would be an air/fuel ratio of 16:1, but that would create excessive HC and NOx emissions. So car manufactures compromise with 14.7:1 ratio and use exhaust gas recycling and catalytic converters with oxygen sensors to provide constant feedback and adjustments to keep all emissions within "acceptable levels." There is no free lunch :)

[Jerry Adams]

So in a car if you have something less than 14.7:1 air to fuel it will start producing CO?

[Author]

Posts