Potential Carbon monoxide Hazards of Cooking in a Tent
by Steffen Wagner
You are on a ski expedition in the Norwegian mountains, and the wind is shaking your storm-proof, tightly guyed tent. The thermometer reads –20°C. Drift snow is piling up in the vestibule, and besides preparing your meal, you also need to melt a heap of snow into drinking water.
Cooking outside is out of the question, so instead the stove is operated in the vestibule or even inside the inner tent, in order to take advantage of the additional heating—and thus drying—effect for clothes that have become damp and clammy.
You think that sounds too extreme and you don’t go on ski tours? Then let’s take another scenario: You are on a kayak trip in the Swedish archipelago and rain is drumming on the tent roof. There is no wind. After a long day on the water, your stomach is growling. The stove hums in the lowered vestibule of your four-season tent. You put tea water on the flame in the new, hyped fuel-saving pot and notice its slightly yellow flame today. When the spaghetti is finally ready, you have a headache and feel slightly nauseous.
In both cases, extreme caution is required, because there is a risk to life: When stoves are operated in poorly ventilated environments, the deadly gas carbon monoxide (CO) can accumulate.
Although campers repeatedly die after leaving a heating lantern burning overnight in their tent, there are only a few documented deaths of people who were cooking inside their tent.
So is cooking inside a tent still an absolute no-go, or is it acceptable if certain precautions are taken?
Among many indigenous peoples, cooking over an open fire inside a tent or hut has always been common practice. With suitable ventilation, it poses no danger. In everyday life, we are often exposed to a certain level of carbon monoxide: A household gas stove, for example, can produce measurable concentrations in a small kitchen when several burners are used without an extractor hood, and on heavily trafficked roads in large cities, values of 100 ppm are measured.
Bengt is a polar guide – this is what professional leaders of commercial polar expeditions are called. Two hundred days a year, the Norwegian crosses the Greenlandic ice sheet with small groups, leads motivated ski-tourists to the North and South Pole, through the Northwest Passage of the Canadian Arctic, through Svalbard in winter, or across the Patagonian ice-field. These are the most extreme ski expeditions on the planet. Safety is Bengt’s business and his expertise.Overnight stays are in expedition tents with snow skirts, and cooking is usually done inside the inner tent – of course with permanent cross-ventilation. To melt and heat enough water for three people, the stove must run for more than two hours each morning and evening. The model most often used is the MSR XGK-EX. It is so loud that you can hardly talk in the tent while it is running.
When I asked Bengt why he chose this particular screamer-stove, he replied: “Because it produces the least carbon monoxide!”
That made me think. A flame is a flame—so why do some stoves produce more CO than others? Are the differences significant? And if so, how can they be explained and reduced?
To find out, I bought a CO measuring device with continuous recording and devised an experimental setup to compare different stoves. I operated them in an aluminium box, closing the lid except for a narrow slit, while bringing a pot of water to a boil inside. Friends lent me a number of different gasoline stoves.
Originally, I had planned to compare stove efficiency as well—how quickly they could bring one liter of water to a boil. However, my PC-connected thermometer broke quickly, and I also realized that the reproducibility of my measurements was not very high due to interfering factors. So I had to abandon quantitative measurements in favour of qualitative comparisons.
In my CO measurements, I expected curves that would rise at different rates depending on the stove, approaching a horizontal equilibrium where CO production and diffusion out of the box balance each other.
The absolute CO values, measured in ppm (parts per million) and plotted on the vertical axis of my graphs, quickly enter the toxic range (see Fig. 1). Of course, results from my aluminium box cannot be transferred directly to a tent. But the values allow comparison between different stoves.
Alongside my own measurements, I tried to gain an overview of scientific and popular publications on the topic of “cooking in tents.”
Source: International Polar Guides Association (IPGA)
This article aims to clarify several factors that promote or reduce the formation of carbon monoxide. We assume that due to weather conditions, a stove is being operated in a vestibule or inside the inner tent. For ventilation, we ideally use two openings on opposite sides of the outer tent, provided it does not have a wide snow skirt. The inner tent must also be opened slightly, preferably on both sides:
What is CO, how does it form, and why is it dangerous?
CO is an odourless, tasteless, and colourless gas produced by incomplete combustion of organic substances. Incomplete combustion is characterized by a lack of oxygen. This occurs, for example, when the flame cools on a cold surface before the second oxidation stage to CO₂ is completed.
For those who want the exact chemistry: Roger Caffin explains it very clearly and understandably.
CO binds about 250 times more easily to haemoglobin in our red blood cells than oxygen does. CO therefore replaces oxygen in the blood. It also attaches to certain proteins in the body. Because of this strong binding, haemoglobin does not immediately release CO again even if we briefly leave the tent and breathe fresh air. Depending on the level of contamination, it can take several hours for symptoms to disappear. In severe cases, only pure oxygen treatment in a hospital helps.
During the experiments for this article, I myself suffered a mild CO poisoning with nausea and headaches in a double garage—so I know what I’m talking about. This despite thoroughly ventilating the garage after each test! As we will see, I also had some real CO spewers in the test.
Factors influencing CO production
- The greater the distance between pot base and flame, the less CO.
Whether there is ice or boiling water in the pot: it is always “cold” compared to the flame. Faster cooling of the flame over a shorter distance leads to more incomplete combustion. In English literature, this distance is called “clearance.” The models I tested ranged from 6 to 35 mm.
- A burner without a pot produces very little CO.
A burner used only for heating or drying can operate in a tent with almost no CO production—though ventilation is still necessary.
- The thinner and more distant the pot support from the flame, the better.
The reason is the same as in point 1. An extreme negative example is the Coleman Peak 1 with its wide metal plates crossing directly through the flame.
- Hands off fuel-saving pots or heat exchanger collars!
To save fuel, the corrugated metal fins used in these designs capture the flames as completely as possible, cooling them efficiently close to the pot. The idea works—but it prevents the complete oxidation of CO to CO₂ so effectively that CO levels skyrocket within minutes.
Larger-diameter pots are better than smaller ones because the fins are farther from the flame.
Still: NEVER use them inside a tent or even in an open vestibule—danger to life!
- Blue flame = good flame, yellow flame = danger!
The flame colour is a direct indicator of temperature. After priming, it should always be blue—for both gasoline and gas stoves.
A yellow/orange flame indicates either low pressure or a dirty fuel line. Since gas cansiters cannot be pumped up, pressure decreases as the canisters empties. Therefore, the canister can never be burned completely empty. Toward the end, the flame turns yellow and produces more CO.
- Pure white gas is the best liquid fuel.
I did not investigate this myself, but the most CO is released when using regular gasoline or kerosene, and the least when using purified fuel.
I usually use alkylate gasoline such as “Aspen 4,” available in hardware stores in canisters and often at gas stations abroad. Heptane is also suitable but more expensive.
- Good adjustability allows CO reduction on the fly.
The higher the burner output, the more fuel is burned—and the more CO is produced. Therefore, it is important to use a stove that can be regulated well.
Many gasoline stoves unfortunately only know “off” and “full blast.” If you are not just boiling water, you can reduce CO production significantly with one adjustment.
Additional recommendations for cooking in tents
- NEVER lie down while cooking!
Fig. 10, CO-meter & warner After a long day, melting large amounts of snow can tempt you to supervise the stove while lying down. If you fall asleep while ventilation fails, you may not wake up. Always sit up while cooking.
- Carry a small CO alarm meter!
It will sound an acoustic warning if limits are exceeded. - Increase clearance with a simple pipe clamp from the hardware store.
This €2 part can significantly reduce CO production.
Summary
Among the stoves tested, I found large differences in CO development. Some are due to design. My measurements show that there is a stove that surpasses the loud XGK-EX in terms of CO reduction: the Primus Omnilite Ti, ideally with the optional silencer burner.
Although it has about 10% less power than the XGK, it is well adjustable, much quieter, and produces less than half the CO. With a pipe clamp, CO production can be reduced even further.
I do not want to issue a free pass—cooking in a tent is potentially very dangerous and should be avoided whenever possible. But with the necessary caution, knowledge of the relationships described here, and appropriate stove equipment, this practice is, in my opinion, manageable and safe.
Epilogue
Source: DALL-E (Open AI)
To round off the topic, I had an AI visualize a stove optimized for CO avoidance according to my specifications. It combines large clearance with pot supports at maximum distance from the flame and vertically arranged outlet holes, which positively affects clearance compared to a flat arrangement.
All referenced publications and also additional ones can be found here: https://wattpaddler.de/kocherpapers/
Disclaimer
The selection of the stove models tested here was more or less random. The focus, however, was on gasoline stoves, since these are the more reliable choice on longer tours and especially in colder conditions. I tested two of my own stoves and several models belonging to friends. The criterion was that they had to be models still commercially available today. The only stove I specifically wanted to include was the MSR XGK-EX, as it was the motive for this examination.
I am impartial toward the manufacturers: I receive no support from any of them, and I had no prior disagreements with any of them.
Since stoves are generally operated outdoors, the CO results presented in this article do not represent an overall evaluation of the stoves, but merely serve as a comparative measurement of a particular characteristic.