CO₂ on board: What did our in-flight air quality measurement reveal?
Protronix case study – measuring CO₂ in a commercial aircraft cabin on the Prague–Zurich–USA route.
Introduction
An aircraft cabin is a very specific environment – a high occupant density in a small volume, with air that must be mechanically supplied and conditioned by the environmental control system (ECS). Cabin air quality has been a topic of discussion practically since the first commercial flights, and we decided to measure it ourselves.
We enjoy taking measurements in the most varied environments – schools, wine cellars, industrial halls and offices. In this particular case, the key parameter worth monitoring is CO₂ itself – a small space with a high occupancy. So we took a CO₂ sensor on board and recorded the complete profile of two flights – from Prague via Zurich to the USA. What did the data show?
1. Boarding the Prague–Zurich flight
While waiting for take-off clearance, the cabin air felt distinctly stale and uncomfortably hot. The sensor confirmed this impression: the CO₂ concentration climbed to 4,500 ppm within just a few minutes – a level that causes very unpleasant symptoms such as headache, dizziness and nausea.
This is a value that is:
within the range where people may experience headaches, dizziness, fatigue, impaired concentration and nausea,
well above a level we would consider unacceptable for occupancy in any other indoor space.
At around 10:00 a.m. the ventilation was switched on at full capacity and the values dropped to a more tolerable level of approximately 1,500 ppm. That is a significant improvement, but still above the recommended limit of 1,000 ppm – see chart on the right.
Chart – boarding the Prague–Zurich flight
Why are CO₂ levels so high during boarding?
This is neither a mistake nor cost-cutting by the airlines, but a technological limitation. When the aircraft is parked at the gate with engines off, the air conditioning packs have no source of compressed air – they have to be supplied either from the on-board auxiliary power unit (APU) or from a ground-based unit connected to the aircraft by a hose (ground pneumatic source or pre-conditioned air, PCA). Neither is often sufficient to cover the thermal and ventilation load of a full cabin. During take-off, the bleed air taken from the engine compressors is throttled or shut off entirely so that the engines can deliver maximum thrust.
The phases of boarding, taxiing and take-off are therefore demonstrably the worst part of the flight from a ventilation standpoint. The International Air Transport Association (IATA) therefore recommends running the air conditioning packs for at least 10 minutes before passenger boarding and keeping them in operation throughout boarding and deplaning.
2. Ventilation switched on – the Prague–Zurich flight itself
As mentioned – at approximately 10:00 (after take-off and flight stabilisation) the ventilation was switched on fully and within a few minutes the values dropped to around 1,500 ppm.
During the rest of the flight the values fluctuated within the 1,500–2,000 ppm range. This roughly corresponds to what published studies consider the typical average for transport aircraft cabins:
Study
Average measured CO₂ concentration
Cao et al. 2019 (179 flights, USA)
1,353 ± 290 ppm
Thermal Condition Investigation 2021 (China)
1,440 ± 111 ppm
Yıldırım et al. 2023 (Turkey, B737/A320neo)
1,399 ppm (max. 3,000 ppm)
Our Protronix measurement (cruise phase)
1,500–2,000 ppm
Our values are therefore in the upper half of what the published literature reports.
Cruise phase of the Prague–Zurich flight – values stabilised in the 1,500–2,000 ppm range.
Protronix note: The regulatory limit of 5,000 ppm and the recommended hygienic limit of 1,000 ppm are two completely different things. The first defines when air is safe. The second defines when it is healthy and comfortable. Aircraft meet the regulations with a wide margin, but exceed the comfort threshold practically all the time.
We understand that some operating conditions are unavoidable – for example reduced ventilation during landing or temporary shut-off at take-off for safety reasons. From the passenger’s point of view, however, these processes significantly degrade the in-flight experience and can cause real health issues in more sensitive individuals.
3. Connection – boarding the flight to the USA
After the connection in Zurich the scenario repeated itself, although the values were not quite as dramatic. While boarding the second aircraft, the sensor again showed values well above the recommended limit – just over 2,600 ppm.
Boarding the second aircraft in Zurich – values around 2,700 ppm
Transatlantic flight – values stabilised in the 1,500–1,800 ppm range
4. The transatlantic flight itself
After take-off and flight stabilisation the values once again settled within the range typical for the cruise phase – around 1,500–1,800 ppm.
The environmental control system works most efficiently at cruise altitude: outside air is drawn from the compressor stages of the engines at a temperature of 150–280 °C, cooled in the air conditioning packs and mixed with recirculated air, which passes through HEPA filters capturing ≥ 99.97 % of particles, including bacteria and viruses.
The full volume of cabin air is exchanged approximately every 2–3 minutes, which is a significantly faster air-change rate than in any office building.
From a CO₂ perspective, however, this does not translate into low values – HEPA filters do not capture CO₂ (CO₂ is a gas, not a particle). The only way to reduce it in the cabin is to increase the supply of fresh outside air, which costs fuel. Modern aircraft therefore operate with a ratio of approximately 50 % fresh and 50 % recirculated air – a compromise between fuel consumption and air quality.
5. Approach, landing and waiting to deplane
The final phase of the flight is once again among the worst. During landing and taxiing to the terminal the ventilation is throttled, and after parking at the gate the same scenario as during boarding occurs: passengers sit in a full cabin, the air conditioning runs in a reduced regime from the APU or a ground source, and the doors often open only after several minutes of waiting.
The values measured in this phase rise sharply again, similarly to boarding.
Approach to landing + landing itself, and renewed rise of values above 2,800 ppm
Summary of the measured data
Flight phase
Measured CO₂ concentration
Assessment
Boarding – Prague–Zurich flight
~4,500 ppm
Critical – beyond the comfort threshold and into the mild health-impact range
Cruise phase – short-haul flight
1,500–2,000 ppm
Above the recommended limit, in line with the published literature
Boarding – flight to the USA
~2,800 ppm
Well above the hygienic limit
Cruise phase – long-haul flight
1,500–1,800 ppm
Above the recommended limit, but stable
Landing and waiting
significant peaks
A repeat of the boarding situation
Expert conclusion by Protronix s.r.o.
Our measurement confirmed what the published literature has been documenting for a long time, but what rarely reaches the average passenger:
Aircraft do comply with regulatory limits. The 5,000 ppm CO₂ limit set by FAA/EASA (CS/FAR 25.831) was not exceeded in any phase. From a safety perspective, everything is in order.
From a comfort and well-being perspective, however, it is not. The hygienically recommended value of 1,000 ppm is exceeded practically throughout the entire flight, and in the critical phases – boarding, take-off, landing – by several multiples.
The main culprit is not the airlines, but physical and operational constraints. A parked aircraft has no source of compressed air in sufficient quantity; during take-off part of the bleed air is reallocated to engine thrust; the air conditioning is tied to available energy sources.
The question of monitoring. CO₂ is not continuously monitored as standard in commercial aircraft cabins today – even though it is a simple, relatively inexpensive and highly informative indicator of air quality and ventilation performance.
What does this mean for our industry?
If even extremely tightly regulated environments such as aircraft struggle with these problems, it is all the more important to have continuous CO₂ monitoring in places where regulation is far weaker – in offices, classrooms, meeting rooms, lecture halls, gyms, restaurants and medical practices.
Our CO₂ sensors and combined indoor air quality (IAQ) sensors do exactly that: they integrate CO₂ measurement with HVAC and BMS control, so that the supply of fresh air is governed by actual room occupancy rather than by a theoretical design assumption.
Key takeaways for HVAC designers and facility managers
The recommended value of 1,000 ppm CO₂ should be the design target for HVAC systems, not just an academic reference.
Without continuous measurement, there is no way to guarantee that the ventilation system is performing as designed – this applies in an aircraft just as much as in a meeting room.
Demand Controlled Ventilation (DCV) based on CO₂ reduces energy consumption while improving air quality during occupied hours.
CO₂ is a strong proxy indicator – it reflects not only air quality in terms of exhaled CO₂, but also indirectly the risk of airborne disease transmission.
The aircraft is a fascinating technological environment – it keeps passengers alive in conditions that would otherwise kill them within seconds. At the same time, it is an environment where, despite enormous technical sophistication, we still struggle with the fundamental rule of all indoor spaces: enough fresh air per person. And precisely where this rule is hardest to meet, it makes sense to measure. Not to estimate, not to rely on the design alone, but to measure.
Our CO₂ sensors perform this role every day in thousands of buildings across Europe. This in-flight experiment was just a small illustration of how quickly a sensor can reveal a problem that the human body senses but cannot name.
