Comment by esaym

The short version is more likely CO2, not CO, and the detector was probably fooled by methane. In more detail:

I was motivated to google, and much like radioactivity or electric current its possible to measure "scary" things to low levels that don't usually matter, and according to a 1982 gastroenterology paper that I found, aerobic fecal fermentation will produce an irrelevent yet measurable extremely low CO output while the anerobic fecal bacteria find CO to be a yummy fuel and will rapidly consume it. Unfortunately they cannot consume CO fast enough to scavenge your blood clean if you're suffering from CO poisoning, which is too bad. Carbon monoxide being known as that thing that gives you massive gas would be far preferable to being known as that poison that kills people, but thats how it goes, can't win em all, so despite our helpful fecal bacteria CO does kill people. The paper theorized the anerobes protect the host by eating the extremely small level of CO that normal aerobic gut flora output.

From fooling around with chemical sensors in my youth, quite a few work by heating up an oxidation catalyst and measuring its temp very closely given the carefully measured outside air temp and electricity fed into them. So feed it a fuel like carbon monoxide and it gets microscopically, yet measurably, hotter. You can play molecular weight games to get a catalyst that preferentially responds to light stuff like methane or CO, or heavy stuff like vaporized wax. My semi-educated guess is the aviation CO sensor got fooled by methane in the guy's gas. Methane isn't a serious threat in an airplane fuel tank so its unlike bowel gas analysis was a high priority in the specs. This is in the class of proving things by ruling them out, and you can't rule out a heated oxidation catalyst sensor getting fooled by low molecular weight methane, its a believable scenario.