The Classroom Feels Stuffy but the CO2 Sensor Says Fine: Ventilation Rates from CO2 Data

Key Takeaways
  • A single CO2 reading describes only one point in the room at one moment, so it cannot by itself judge whether ventilation is adequate.
  • A true NDIR sensor measures CO2 directly, while an eCO2 estimate infers it from other gases and can be misled by cleaning products or fresh paint.
  • Automatic baseline calibration can silently bias readings low in a school, because classrooms used for evening events may never empty long enough to re-zero correctly.
  • The indoor-outdoor CO2 gap is what describes ventilation, so a drifted baseline or a wrong outdoor assumption makes the reading misleading even when the display looks plausible.
  • Mounting the sensor in the breathing zone, away from windows, doors, supply outlets, and people, and checking it against fresh air each term are what make the number on the wall agree with the air in the room.
Table of contents

    It is 10:40 on a Tuesday morning, and the Year 9 classroom feels thick. Twenty-eight students are working through a maths worksheet, the windows shut and the air heavy enough to make everyone yawn. The teacher glances at the small white box beside the door: 720 ppm, green. "The sensor says the air is fine," she tells the caretaker at lunch. He shrugs, and nothing changes.

    This is the most common failure in indoor air quality monitoring, and it is not a broken sensor. A single CO2 number describes one point in the room at one moment. It does not say whether the room is ventilated well enough. The fix is understanding what the sensor measures, where it should sit, and how a reading becomes a ventilation rate.

    How an NDIR sensor turns air into a number

    Most wall-mounted CO2 sensors use NDIR, short for non-dispersive infrared. An infrared light source shines through a small chamber of sampled air toward a detector. Carbon dioxide absorbs infrared light very strongly at one wavelength, around 4.26 micrometres, so a filter only lets that wavelength through. The more CO2 in the chamber, the more light is absorbed and the weaker the signal. The sensor converts that signal drop into parts per million. NDIR is the standard choice because it measures the gas itself, not a proxy.

    A cheaper type of sensor does not measure CO2 at all. So-called eCO2 sensors measure volatile organic compounds and estimate a CO2 value from the result. It roughly tracks real CO2 when people are the main source of both gases, but fails when cleaning products or fresh paint add other compounds. If the goal is to judge ventilation, an NDIR sensor is the right tool, and an eCO2 estimate is not.

    The main weakness of NDIR sensors is drift. The infrared source ages and dims, dust settles on the optics, and the baseline shifts with temperature. Many sensors counter drift with automatic baseline calibration, re-zeroing themselves to outdoor background levels whenever the room empties overnight. That works in an office that empties nightly, but a classroom used for evening events may never empty for long enough. If the sensor re-zeros while people are still in the room, every reading afterwards sits too low.

    From a CO2 reading to a ventilation rate

    Every person in the room is a small, continuous CO2 source. A seated adult at rest exhales roughly 0.3 litres of CO2 per minute, so a full classroom is a substantial source. Ventilation brings in outdoor air and carries room air away. The concentration settles at a balance point: outdoor background plus the CO2 generated indoors divided by the ventilation rate. Double the fresh air, and you roughly halve the amount by which indoor CO2 rises above outdoor levels.

    Outdoor air typically sits around 400 to 420 ppm of CO2, and this background is the starting point for every judgement indoors. A reading of 720 ppm against a 400 ppm background means about 320 ppm of extra CO2 from occupants; the same 720 ppm against a 450 ppm background means only 270 ppm extra. That indoor-outdoor gap is what actually describes ventilation, so a drifted sensor baseline or a wrong outdoor assumption makes it wrong even when the display looks plausible.

    Ventilation engineers express the result as air changes per hour, or ACH: the volume of fresh air supplied per hour divided by the volume of the room. A 60-square-metre classroom with a 3-metre ceiling holds about 180 cubic metres of air, so one air change per hour means supplying 180 cubic metres of fresh air hourly. More people per square metre means a higher air change rate to hold CO2 steady. A CO2 reading is only useful once you know the background level, the number of occupants, and the room volume — and the sensor knows none of those things. Its job is to report concentration; turning that into a ventilation decision is yours. That also explains the stuffy room: warmth, humidity, odours, and still air all play a part, and CO2 is only one ingredient.

    What to look for when choosing a sensor

    Start with the sensing principle: choose a true NDIR sensor and avoid eCO2 estimates. Then look at how the sensor behaves over time, not just its headline accuracy figure. A sensor quoted at a few dozen ppm of accuracy can be off by far more a year later without drift compensation, or if its automatic baseline calibration cannot work in your building's usage pattern. Ask whether it supports manual calibration against a known reference, and whether baseline calibration can be disabled or scheduled — in a school, that control matters more than a slightly better accuracy figure.

    Next, think about how the reading will be used. A display on the wall is enough if a teacher is expected to open a window when the number climbs. But to set fan speeds, prove a ventilation rate, or spot a pattern across a term, the sensor needs to log data and talk to a building management system — typically over RS485 and Modbus. Be honest about the limits, too. One sensor measures one point in one room; it cannot describe the whole building. A CO2 sensor is also not a life-safety device: it is a ventilation tool, not a substitute for a proper gas alarm where one is required.

    Installation pitfalls: where the sensor sits decides what it sees

    Mount it at the wrong height and the reading is wrong. Air in an occupied room is well mixed by body heat and movement, so measure in the breathing zone — roughly 1.1 to 1.7 metres above the floor, about the height of a seated or standing person's nose and mouth. Near the ceiling it samples warm air up high; near the floor, cool air by the skirting board.

    Keep it away from windows and doors. A cracked window sends a jet of outdoor air across the room, and a sensor sitting in that jet reads close to outdoor background even while the far corner is stagnant. The sensor should represent the room as a whole.

    Keep it away from people. A sensor above a teacher's desk, or on the wall behind the front row, sits inside the plume of exhaled breath from the nearest occupants. Keep at least a couple of metres between the sensor and the nearest regular seating position.

    Do not ignore airflow. A sensor next to a supply diffuser samples freshly delivered air before it has mixed with the room, so it reads low. A sensor in a dead corner behind a bookcase samples air that barely moves, so it reads high and responds slowly. Best is an internal wall away from outlets and corners.

    Watch out for heat and sunlight. A sensor on a sun-facing wall warms up through the morning, and the temperature rise can shift its readings and age its components faster. Keep the sensor out of direct sun and away from heat sources such as radiators and projectors.

    Verification and calibration: proving the number is real

    Once the sensor is installed, do not take its first reading on trust. The simplest check costs nothing: leave the room empty overnight with the ventilation running normally, then read the sensor first thing in the morning before anyone arrives. In a well-ventilated empty room, the reading should sit close to the outdoor background, typically around 400 to 420 ppm. Far above that, either overnight ventilation is too weak or the baseline has drifted.

    The stronger check is a comparison against a reference. Borrow or hire a recently calibrated handheld CO2 meter, place it next to the wall sensor at the same height, and let both settle for at least fifteen minutes before comparing. If the two agree within a modest margin, the wall sensor is probably fine; if they disagree by a large amount, and the handheld is the trusted instrument, the wall sensor needs calibration or replacement. A useful extra test is the breath check: exhale gently toward the sensor and watch the reading climb, then fall back — proof it responds to real CO2 rather than sitting frozen.

    How often should you recheck? A practical rhythm is a fresh-air check at the start of each term, a comparison against a reference instrument once a year, and a check any time the sensor is moved, the room is refurbished, or the readings stop matching how the room feels. Automatic baseline calibration is a help, not a substitute: it only works if the room genuinely empties for long enough. If a sensor's readings keep drifting between checks, or calibration no longer brings it back into line with the reference, replace it rather than keep adjusting it.

    Troubleshooting

    Symptom Likely cause What to check
    Room feels stuffy but the sensor reads low and steady Sensor in a fresh-air jet or near a supply diffuser Move it to an internal wall and compare readings over a full school day
    Reading spikes whenever someone stands nearby Sensor inside the breath plume of the nearest occupants Move it a couple of metres from the nearest desk and watch the spikes disappear
    Reading never falls near outdoor background in an empty room overnight Baseline drift, or re-zeroing while the room is occupied Do a morning fresh-air check after an empty night; compare against a handheld meter
    Reading barely moves all day as the room fills and empties Frozen output, long averaging, or a sensor in a dead-air corner Do a breath check; check the averaging settings and move it out of the dead corner
    Readings climb through the term and never come back down Gradual drift from infrared source ageing or dust on the optical path Compare against a reference meter; if the gap has grown, calibrate or replace it
    Two sensors in similar rooms disagree by a large amount Different mounting positions, different ventilation, or one sensor drifting Swap the sensors between rooms for a day and see whether the difference follows the sensor

    Checklist: getting a trustworthy CO2 reading in a classroom

    • Choose a true NDIR sensor, not an eCO2 estimate, so the reading reflects real CO2 rather than a guess from other gases.
    • Mount the sensor in the breathing zone, roughly 1.1 to 1.7 metres above the floor, on an internal wall in the main body of the room.
    • Keep the sensor away from windows, doors, and supply diffusers so it samples representative room air, not a fresh-air jet.
    • Keep at least a couple of metres between the sensor and the nearest seats so exhaled breath cannot spike the reading.
    • Keep the sensor out of direct sunlight and away from radiators, projectors, and other heat sources.
    • Note the local outdoor background level and the number of occupants, so a reading can be turned into a ventilation judgement rather than taken at face value.
    • When the reading and the room disagree, trust the room first and investigate the sensor's position and baseline before changing the ventilation.

    If the classroom in the opening scene had followed this approach, the story would have ended differently. The caretaker would have known the green light only described the air at one point on one wall. A sensor in the breathing zone, checked against fresh air each term, would have turned a vague complaint into a ventilation rate — something a facilities team can act on. The same NDIR principle sits at the heart of the RS485 greenhouse CO2 sensor, carbon dioxide measuring instrument: it is marketed for greenhouse use, and this article makes no claim it is rated for classroom deployment, but the lessons about placement, background, and calibration apply to any NDIR sensor used for indoor monitoring. Install it where it can see the room honestly, check it against fresh air, and the number on the wall will finally agree with the air in the room.

    Need help choosing the right sensor?

    Our engineers can help you pick the right model, check compatibility, and get your installation right the first time.

    RS485 Greenhouse CO2 Sensor - Carbon Dioxide Measuring InstrumentRS485 Greenhouse CO2 Sensor - Carbon Dioxide Measuring InstrumentView product →
    Email our engineers Contact us
    Back to blog

    Leave a comment

    Please note, comments need to be approved before they are published.