It is a cold morning at Nanyuki, the kind where the cabin heat gets switched on before the engine has finished warming. Forty minutes into a dual navigation sortie, the student mentions a mild headache and says they slept badly. The instructor, who also got up at five, feels a bit heavy behind the eyes and agrees it has been an early start.
Two people have just explained away the same symptom with the same ordinary reason. Nobody has asked whether something in the cabin is causing it.
The Short Version
- Carbon monoxide (CO) is colourless and odourless, so the absence of a smell tells you nothing (NTSB Safety Alert 070).
- The early signs are vague: headache, nausea, sluggishness. Pilots tend to dismiss them (NTSB Safety Alert 069).
- The FAA brochure on carbon monoxide links higher blood levels to impaired judgment, drowsiness and blurred vision, and notes symptoms can appear sooner or be worse at altitude (FAA carbon monoxide brochure).
- Research for the FAA found that muffler and exhaust systems were the primary source of CO when it caused an accident, and that inadequate maintenance and inspection featured in many of those cases (FAA report DOT/FAA/AR-09/49).
- A detector with an active alert, a proper exhaust inspection and a rule that a smell of exhaust means "assume CO" cover most of the risk.
Why it hides inside IMSAFE
Our IMSAFE check asks a pilot to assess illness, medication, stress, alcohol, fatigue and emotion before the flight. That check happens on the ground, with a clear head. CO is different because it attacks the very faculty doing the assessing, and it does it after take-off.
The FAA's brochure on the subject ties rising levels of CO in the blood to a ladder of symptoms: nothing noticeable at the bottom, then slight headache, then drowsiness, impaired judgment and blurred vision as the level climbs, and it notes that symptoms may come on sooner or be worse at altitude (FAA). The FAA's advisory circular on the subject adds that early signs can include a feeling of sluggishness, warmth and tightness across the forehead (FAA AC 20-32B).
Read that list again as a pilot. Sluggish, slightly warm, a mild headache, a little less sharp. Every item is also a description of a normal Tuesday after a short night. That overlap is what makes CO dangerous: the symptom and the excuse arrive together.
How it gets into the cabin
The engine of a piston single produces CO as a normal part of combustion, and it leaves through the exhaust. The problem begins when some of it reaches the cabin instead. The FAA-commissioned research above found muffler and exhaust systems to be the primary source in accidents where CO was the cause, with the gas reaching the cabin through the heating system or through gaps in the firewall (FAA DOT/FAA/AR-09/49).
That is why the cold morning matters. Cabin heat is the moment the system is most likely to carry a leak towards the occupants, and it is also the moment people are least inclined to open a vent. In Kenya's highlands, where early flights at Nanyuki, Nakuru or Wilson start cool, the heater goes on and the windows stay closed.
The inspection side matters too. NTSB and FAA guidance calls for a thorough exhaust inspection at each 100-hour or annual inspection, and for pilots to look at the muffler ends and for soot during the preflight (NTSB Safety Compass). The same material notes that cracks are easily missed on a visual check. An exhaust problem is a maintenance finding long before it is a cabin problem, which ties it directly to the tech log habits in our piece on the aircraft tech log.
What the NTSB found
The NTSB examined this in its report AIR-22-01. It covered 31 accidents attributed to CO poisoning between 1982 and 2020. Of those, 77 per cent were fatal, with 42 fatalities and four serious injuries, and in 30 of the accident reports no CO detector was found or reported (NTSB press release; report: NTSB/AIR-22-01).
Two things stand out. First, the number is small, which is itself a clue: CO events that end with everyone landing safely and a vague headache never become accident reports. Second, the Board recommended that the FAA require detectors in enclosed-cabin, reciprocating-engine aircraft with an aviation-specific performance standard and an active aural or visual alert, and it also asked industry groups to promote detectors and regular inspections (General Aviation News summary). The FAA has so far recommended voluntary installation rather than a requirement.
A fair reading is this: the regulator has not made detectors mandatory, so nobody will make you carry one. That makes it a personal and school-level decision, not a compliance box.
Detectors: why the alert matters
The NTSB's own safety writing makes a distinction worth passing on to students. A detector that only changes colour depends on someone remembering to look at it, at the moment when that person's attention is the thing being degraded. An electronic unit that sounds an alarm or flashes does not need the pilot to be sharp (NTSB Safety Compass).
For a flying school, a detector in each aircraft is a cheap item compared with the cost of the fleet, and it protects the people least able to judge the situation, which is to say students and passengers. It is also a good line for the school's own safety management system to track: is a detector fitted, is it checked, and does anybody know what an alarm means?
What to do when you suspect it
FAA guidance is direct. If you smell exhaust, assume CO is present, shut off the cabin heater and close other openings from the engine compartment. The brochure also advises maximum fresh air, landing as promptly as possible, seeking medical attention afterwards, and having the aircraft inspected by a certified mechanic before it flies again (AC 20-32B; FAA brochure). Remember the earlier point: no smell does not mean no CO, so the same actions apply to an unexplained headache or drowsiness in flight, especially if the heater is on.
For an instructor, the decision to turn back is easier when it has been agreed on the ground. Brief it as a rule: if either of us feels a headache or unusual drowsiness with the heater running, heat off, vents open, and we head for the nearest suitable aerodrome. Nobody has to prove anything first.
The five-minute habit
- Preflight: look at the exhaust and muffler ends for soot or damage, and read the tech log for any exhaust or heater defect.
- Detector: if one is fitted, check it is powered and in date; if none is fitted, raise it with the school.
- Briefing: agree the trigger and the response before start-up.
- In flight: a headache, drowsiness or "I feel off" gets the same response as a smell: heat off, fresh air, land.
- After landing: write it in the tech log and ground the aircraft until an engineer has inspected it. Get checked by a doctor too.
AngaBrief records the IMSAFE answers a student gives before a flight and puts them in front of the instructor. A tool cannot detect CO, and it is not a dispatch authority: final go or no-go rests with the pilot in command, which includes the call to turn back when the cabin is the problem.
Key Takeaways
- CO cannot be smelled, and its early symptoms look like ordinary tiredness.
- Exhaust systems are the main source, so the preflight and the 100-hour inspection are the first line of defence.
- The NTSB found no detector in 30 of 31 CO accident reports it reviewed, yet fitting one is voluntary. Choose to fit one.
- Agree on the ground what a headache or unexplained drowsiness means: heat off, vents open, land.