Carbon Monoxide - Myths and Facts

ll the ""good books"" say to not use a stove anywhere in a tent because there is a danger of carbon monoxide (CO) poisoning. You also get this in the product warnings on most stoves, from the company lawyers. However, many of us do this all the time with

All the ""good books"" say to not use a stove anywhere in a tent because there is a danger of carbon monoxide (CO) poisoning. You also get this in the product warnings on most stoves, from the company lawyers. However, many of us do this all the time with no apparent ill-effect. The photo to the right was taken in bad weather in the Pyrenees: the steam and CO (and the CO2) went out the vent at the top. The hot soup was lovely, as was the rest of dinner - every night for 6 weeks. One has to wonder whether the whole thing is a myth perpetuated by companies who simply parrot what other company lawyers have said, or whether there is some substance to the idea.

On the negative side we do have a LOT of articles in journals such as the American Journal of Emergency Medicine and the Journal of Wilderness and Environmental Medicine about the hazards. It seems that some medics have nothing better to do than to bundle several quite unwitting volunteers and a couple of petrol or kero stoves into a sealed tent and see how long they can go before they cark it or the stove goes out. Seems rather unethical to me. However, before getting greatly worried about all these articles, please note two details. The first is that most of these 'experiments' have been run by medics with little or no apparent training in research: the experiments fall very much into the 'fancy that' category. They lack any form of experimental rigor - two subjects, one experiment, and few controls over the conditions, and they are generally solely concerned with what they saw happening - the CO concentration in the subject's blood went up to x ppm. None of them really address the question of why the CO was emitted. I guess the medics knew how to measure CO concentration in blood, but not how to study flame chemistry. Regardless, most of those papers are extremely poor work. One subsequent Letter to the Editor 'complimented' the authors of such a paper (using human volunteers!) thus: 'The authors are to be congratulated on managing to achieve CO levels of 350 to 500 ppm with one as high as 600 ppm. No previous tent or snow cave study has exceeded 300 ppm — the level thought necessary to cause collapse'. Such levels are known to be potentially lethal. Given the rest of that letter and that the author was a (very) senior medic in the Royal Navy, one suspects a heavy dose of incredulous and horrified sarcasm.

On the positive side we have the results of an test run by friends of Ed Huesers of IceTool fame. They fired up some stoves inside an igloo, with the door shut no less, and found little build-up of CO using a CO meter. But while their results were good, we have little knowledge of the actual operating conditions. Did they do something 'right' which kept the CO levels down? It would seem this is possible. It is also very likely that the cold snow they made their igloo out of was really quite porous and could let the CO leak out - this may not happen here in Australia though.

Recently the author has been conducting a major research project for Backpacking Light on the CO emissions from all sorts of stoves. A theory of why CO is emitted from some stoves but not others was developed and verified. The common factor in all stoves is that the fuel mixes with air and burns. One burn pathway is C + O => CO, CO + O => CO2. In the high temperature of a flame, the O2 is dissociated into oxygen (O) atoms and the hydrocarbon fuel into hydrogen (H) and carbon (C) atoms.) Given a chance the hydrocarbon will burn fully to carbon dioxide (CO2) and water (H2O), and will make negligable amounts of carbon monoxide (CO). But if the flame is interrupted and cooled the combustion process may be incomplete, and CO may be allowed to remain. This is known as 'quenching', and this is where the danger lies. What can interrupt the combustion process? Anything which blocks and cools the flame itself - such as a cold pot sitting too low on the stove. Basically, if the pot is placed too close to the flame the combustion cycle can be quenched before all the CO in the flame is converted to CO2. A number of factors in the design of the burner on a gas stove the stoves were varied and the theory checked out. It was also found that some stoves with large surrounding metal trays under the burners can generate a lot of CO: the metal tray tends to block the airflow a bit. Different fuels were also tested, and it was found that petrol, kero and alcohol are all worse for CO emission than gas.

The author's research built on previous research done by another group, where they were able to get both high and low levels of CO from a small number of stoves, burning all major fuels (metho, shellite, gas), by varying the experimental set-up.  The authors of that research found that they could reduce the CO levels from dangerous to safe levels just by raising the pot about 20 mm above the pot supports on a couple of 'bad' stoves. They hypothesised that flame quenching was responsible. Another paper1, also from a member of the British military gives a similar explanation. Several authors have noted that any sign of an orange glow (which comes from hot carbon particles) might be a good indication that the burning process has been interrupted or 'quenched', and that CO may be left behind. One interesting source for this came originally from the Toyota Motor Company, and described the combustion process inside a car engine.

What this means is that any stove which has low pot supports can be dangerous. It also means that running any stove flat out may be dangerous, if the increased fuel flow means the flames are longer, reach up to the pot, and get interrupted and cooled. However, a well designed stove should not suffer from this problem, at least at low altitudes. So not only is full-bore operation wasteful, but it could also be dangerous. On the other hand, cooks who always keep their (well chosen) stove turned down will generally be safe. Note: your pot will heat up quite happily by conduction from the hot air immediately above the flame: it does not need to be buried in the flame. In fact, having the pot in the flame just increases the chance of a burnt pot.

There have been suggestions that melting snow is more dangerous because the pot is so much colder. Let's do a reality check on this one. The flame temperature is going to be well above 1000 C, but let's use 1000 C in this case. Warm water might be at 30 C, while snow might be at -10 C. In one case the temperature differential is 970 C, while in the other case the differential is 1010 C. This is a difference of 40 C in 1000 C: about 0.4%. Come on guys: as far as the flame is concerned, there is no difference between the two!

The general finding is that gas stoves are safer than petrol or kero stoves. In fact, some well-known brands of liquid fuel stoves can be quite bad. We know that the heavier fuel molecules in petrol and kero need more oxygen and take longer to burn; inadequate fuel/air mixing may also make longer flames. This should be a consideration when you select your next stove.

Certainly, the results do mean that you need to make sure your stove has adequate airflow to the mixing region. A very large jet allowing excess fuel flow or very small airholes next to the jet limiting oxygen mixing are dangerous. If using a stove which has different jet sizes, make sure you use the right one - erring on the small side if you are not sure. If the flame is long it may reach up to the pot and be blocked: short flames are better for two reasons: they mean the flame won't reach the pot, and they also mean there is plenty of oxygen getting into the mix. You can of course decrease the flame length by reducing the fuel flow in many cases. But don't think this means you cannot have a windscreen around the stove. Some people (and companies) have claimed that it is dangerous to have more than two or three sides of the stove shielded. Come on: the stove needs some airflow, but not a howling gale! What matters most is the cross-section of the air-holes around the jet as they are usually the limiting factor. I normally leave a vertical 10 mm gap in the windshield with a gas stove, and that must be 100 times the cross-section of the air inlets in the burner. In the picture above the windshield is right open because I have just turned the stove off. Dinner is in the pot, but it is cooking in its own heat, and I wanted the stove to be visible in the picture.

This isn't the whole story of course.  Putting your stove outside the tent in the pouring rain isn't all that smart: trying to cook without getting wet arms will lead, sooner or later, to knocking the whole lot over. This could be just inconvenient, but it could also result in either you or your tent being burnt. Having the stove in the alcove or bell end of the tent with the door open at the top above the stove is a good compromise: just make sure there is plenty of ventilation and that the tent fabric is not too close above the stove. Coleman recommend on their latest multi-fuel stove that you have a vent of at least 10 square inches - that would be about 25 mm by 250 mm. This is not hard to achieve. In practice getting adequate ventilation may be the least of your worries during a storm!


Hydrocarbons - another hazard

Just when I thought carbon monoxide was the only hazard in the 'invisible' class, G Horrocks sent the following to me. I have to say I have not actually run into the problem he describes, but that may be because I have never run a stove in a really enclosed space: it just didn't seem a wise move. On the other hand, my wife used to complain strongly about the pong of kero up the back of the tent when I was using a kero stove near the entrance. Maybe she was right.

You mention CO poisoning and cooking stoves on this page. I have heard some stories regarding CO and CO2 poisoning and cooking stoves, and in my humble opinion I think it is overlooking the real issue with poisoning from cooking stoves. I have no proof or experiments to justify my assertion, just a bit of knowledge of combustion physics and many years of use of my humble Whisperlite. I welcome any comments for or against!

I believe that the biggest issue with cooking with stoves in enclosed spaces is hydrocarbons, not CO or CO2. What's more, it is only an issue when the stove is turned off. When the stove is running it is fine. If you look at the design of the Whisperlite stove, the control valve is on top of the fuel bottle and is connected to the burner by approx 100mm of tubing. It is then heated to vapourise the fuel, mixed with air and burnt. When the burner is running there is no issue, the flame is running leaner than stoichiometric and hence essentially all fuel is burnt and almost no CO is produced. The poisoning issue is when the burner is turned off. You then have several millilitres of liquid fuel in the fuel line still under pressure. The flame goes out as there is not enough fuel to support the flame, so the liquid fuel then gets heated in the burner, evaporates and slowly escapes into the atmosphere unburnt. As the density of vapour is many times less than the liquid this small volume of liquid becomes a large volume of vapour and mixes with the air. The hydrocarbons in the air then cause the problem - essentially a weak form of petrol-sniffing.

I believe I have witnessed this occuring. I remember when I were a lad and my Whisperlite was new, my scout group was preparing for our first extended bushwalking trip, the Overland Track. As part of the preparation we decided to have a night practising preparing the types of food we would be eating on the track so we cooked in the scout hall using our camping stoves. We had all the windows in the hall shut as it was a cold winter night. While the meal was cooking there was no problem. When I turned my Whisperlite off, within a minute everybody in the hall was feeling drowsy and generally off-colour. We opened the windows, the air cleared and we rapidly recovered.

If my theory is correct, then this issue only affects liquid fueled stoves with a long distance between the control knob and where the fuel is vaporised. This would explain why you have not noticed it with your gas stoves.

I think Glenn makes a very good point here. Liquid fuel stoves always stink when you have turned them off. It is interesting that the latest stove from Primus, the Gravity MF, has a rotating connector at the fuel tank, and they recommend that you flip the tank upside down for a short while before you turn the stove off. This is to clear the fuel line of any fuel. Doing this will prevent the above problem