Good ventilation is an important mitigation strategy for COVID-19 transmission as it reduces exposure to small aerosols that can remain airborne for long periods in shared indoor spaces. However it can be challenging to determine how effective ventilation is in many buildings. CO2 monitoring is an approach that can be used to both assess and manage ventilation. CO2 is in our exhaled breath and emitted through our normal activities. In an indoor environment, providing there are no other significant sources (e.g. combustion devices) it represents the fraction of air that has been exhaled by people in that space. The higher the CO2 value, the more of the air has previously been breathed by people in the room. The amount of CO2 in the air increases with the number of people in the space and the time that they spend there, and is reduced by ventilation with outdoor air. A CO2 measurement in an indoor space can therefore be used as a proxy for both occupancy and ventilation. It is important to remember that, although improving ventilation can reduce transmission risk, the level of CO2 in the air is not a direct measure of infection risk. CO2 measurement is already widely used in mechanically ventilated buildings and some naturally ventilated buildings with automated controls – the CO2 value is used as a measure to control the ventilation system. This section focuses on stand alone use of CO2 monitors rather than those installed as part of building management systems.
SAGE-EMG have looked at the relationships between ventilation and infection risk and considered guidance being offered worldwide. Their current consensus is:
It is also recommended that spaces where there are enhanced risks, such as those where there is regular sustained singing/speaking or spaces where aerobic activity takes place, that higher ventilation rates may be needed, and it is more important to maintain CO2 < 800ppm.
CO2 monitors can be used in two different ways in buildings
However the sensors are used, it is important to have a strategy for what to do if readings suggest poor ventilation – it is no use measuring if action is not taken. This may be a staged plan where initial actions focus on guidance for managing ventilation or changing occupancy (lower numbers, more frequent breaks) with a second stage to take more significant action if the environment can’t be improved.
CO2 measurements are a useful indicator, but should not be treated as exact readings or a direct measurement of ventilation rate unless used by a professional as part of a well-controlled ventilation measurement assessment.
Multiple factors can influence the CO2 reading within a space including:
Figure 1: Measured CO2 concentration in a well mixed room (black line) with occupancy (red bars).
In most spaces the CO2 level will fluctuate with the occupancy, activity and variation in ventilation. Occasionally sensors will record short spikes which are much higher than the average reading – this may be a short increase in occupancy, or someone located too close to a sensor for a short period. In using CO2 data to assess ventilation, mean values typically over periods of 15-30 minutes are generally more representative than short duration fluctuations. Sensors should be positioned in the occupied region of the room as far as possible, and ideally at breathing height. It is a good idea to position them away from windows, doors and ventilation supply points. They should also be positioned at least 50 cm away from people. If sensors are used to actively manage ventilation then greater consideration needs to be given to the display, location and robustness of the sensor. Those that use mains power are likely to be more reliable as they will not require maintaining as frequently. There may also be benefits in fixing sensors to a wall or other surface so they are not removed as easily. A maintenance plan should be in place to regularly check sensors are working and to recalibrate if necessary. Spurious readings may indicate a faulty sensor. It is important that there is clear responsibility for the sensors. Not all spaces in buildings are suitable for CO2 monitoring. They are likely to be most effective in medium sized spaces with regular occupancy by a reasonably consistent number of people. These may include offices, meeting rooms, classrooms, restaurants/bars, some retail, indoor sports in smaller spaces and many workspaces. Spaces with very low numbers of occupants or those that are occupied very transiently are less likely to give reliable readings. Large spaces are unlikely to be fully mixed and may require several sensors to get an appropriate measurement. This is described in more detail by SAGE-EMG1. In spaces which have high levels of filtration/air cleaning, such as through application of a stand alone air cleaner or HEPA/UVC in ducting, a CO2 monitor will not be representative of the air from a viral risk perspective. A HEPA filter or UVC device will remove virus and other contaminants from the air, but will not remove gases such as CO2.