What a CO2 monitor tells you about an ordinary room

As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.

The question a room CO2 reading actually answers

An adult sitting still produces somewhere around 18 to 19 litres of carbon dioxide an hour. That gas mixes into the room and is diluted by whatever outdoor air arrives. Once production and removal balance, the room settles at a steady concentration, and that concentration is simply the outdoor value plus the production rate divided by the ventilation rate. Because a litre of gas in a cubic metre of air is a thousand parts per million, the sum is easy to do in your head.

That table is where the familiar benchmark comes from. The widely quoted target of a thousand parts per million is not a toxicity threshold; it is the reading a room produces when each person in it is receiving roughly eight to ten litres of outdoor air per second. Read the table in reverse and any steady reading becomes a ventilation measurement. A bedroom holding at 1,700 ppm with one sleeper is delivering about four litres per second, which is half of what guidance asks for.

This is also why banded displays are coarse by nature. A device such as the INKBIRDPLUS NDIR monitor states three stages, normal from 400 to 700 ppm, warning from 700 to 1500 ppm and dangerous from 1500 to 5000 ppm, with a user set alarm point. The word dangerous there describes ventilation, not poisoning, and the band is wide enough to cover everything from a slightly stuffy study to a sealed room with four people in it. An accuracy figure and a measurement range are not stated in those listing bullets, so the stage boundaries are the only published numbers to work with.

Outdoor baseline, and why the number never starts at zero

Every reading is a sum of the outdoor background and what the occupants have added. Outdoor air now sits somewhere above 420 parts per million, and local values run higher in traffic, higher again in a courtyard where air does not move and briefly much higher downwind of a flue. A monitor that reads 480 ppm in an empty room is not faulty; it may be telling you about the street.

That background matters because of how these sensors hold their zero. An NDIR sensor measures absorption, and absorption measurements drift as the lamp ages and the optics collect dust. Consumer units compensate with an automatic baseline routine that assumes the lowest value seen over a rolling window must have been outdoor air. In an office that empties every night the assumption holds. In a bedroom that is never aired, or a plant room, or a small flat with the windows shut through winter, it quietly rebases the sensor to a wrong zero and the readings all shift down. That is why both INKBIRDPLUS units describe the same remedy: switch the unit off, take it to a ventilated place and hold three buttons for 200 seconds to run a manual calibration against outdoor air. Our guide to calibrating a carbon dioxide monitor covers the routine for other designs.

Reading a room: occupancy, volume and the rate of rise

The steady value is only half the information in the trace. The slope carries the rest. A small sealed bedroom climbs quickly because the production is spread over few cubic metres; a large draughty living room with the same two people climbs slowly and settles lower. Watching the climb tells you the room volume and the leakage working together, which is something no single reading can.

The decay after everyone leaves is even more useful, because it isolates ventilation from occupancy. Once production stops, the excess above outdoor decays exponentially, falling to about a third of its starting excess in one air change. If a room sits at 1,400 ppm when you leave it and reads about 750 ppm an hour later, the excess has fallen from roughly 980 to roughly 330, which is one air change in an hour. Repeat it with the window open and you have measured what the window is worth. A monitor with a long battery life makes this practical, and the Inkbird IAM-O2 three in one monitor states three to six months of use on a charge, a measurement range of 0 to 9999 ppm, and temperature and humidity accuracy of plus or minus 0.2 degrees Celsius and plus or minus 1.8 percent relative humidity, so it can be left in place for the week of observation this exercise needs. A worked routine for the overnight case is in our guide to using a carbon dioxide monitor in a bedroom, and the Qingping carbon dioxide monitor review looks at what logging to a phone adds.

What the CO2 figure says nothing about

Carbon dioxide is a proxy for one thing only: the rate at which human occupancy is being diluted. Several of the problems people buy these devices for are invisible to it.

• Particles from cooking, candles or outdoor air. These come and go independently of occupancy, and opening a window can raise them while lowering carbon dioxide.
• Carbon monoxide from a faulty appliance, which is a different gas entirely and needs an electrochemical alarm sited to a standard.
• Volatile compounds from paint, flooring and cleaning products, which continue in an empty room.
• Radon, which is driven by pressure differences and soil rather than breathing.
• Mould, which no air monitor detects. Its precondition is sustained surface dampness, which is a humidity and temperature question.

Multi channel units bolt the missing sensors into the same case. The INKBIRDPLUS six in one monitor states carbon dioxide, PM2.5, PM10, an index value, temperature and humidity, with an audible alert once the index passes 100. That is one case holding at least three distinct sensing technologies, and each has its own error. A dedicated particle instrument is a different proposition again: the Eujgoov dust particle counter states six size channels from 0.3 to 10 micrometres and a built in micro vacuum pump, and it reports counts as well as a mass concentration. Counting particles in size bins and estimating a mass figure are not the same measurement, which is the distinction our guide to choosing a PM2.5 monitor starts from.

Where to put the monitor and what to change

Placement decides whether the number means anything. Put the unit at seated or sleeping head height, at least a metre from the nearest person, out of the direct path of a vent, a window or a door, and not above a radiator or in direct sun. A monitor on a desk beside your face reads your plume. A monitor on top of a wardrobe reads a stagnant layer.

Then act on the shape of the trace rather than the peak. A room that climbs steadily all evening and does not recover overnight needs a permanent opening, a trickle vent or a mechanical supply, not a window opened for five minutes before bed. A room that spikes only when it is full needs a rule for that occasion. A room that reads high the moment you walk in needs the sensor checked outdoors first, because a drifted baseline looks exactly like poor ventilation. Once the number responds the way the arithmetic above predicts, you have a working instrument and a measurement you can act on.

Frequently Asked Questions