Thread by @CathNoakes

A Sunday thread on ventilation metrics and exposure as they are not quite as straightforward as you may think 1/

Under steady state conditions (ie people have been in the space for a while and the ventilation is constant) then concentration in the air of anything that remains airborne (CO2, the smallest virus particles) is emission rate divided by total ventilation rate 2/

The total ventilation rate is the total volume of fresh air provided to the room, which is normally measured in m3/s, l/min or a similar metric. You can also get this from the ventilation rate per person (l/s/p) multiplied by the number of people 3/

For CO2, everyone exhales it and it stays airborne, so you can estimate the emission rate from the amount of CO2 they exhale (g/s/person or l/s/person) x number of people. You then have to do some conversions to get from volume or mass to ppm 4/

This is the first complication. When people are fairly passive, the CO2 concentration at 10 l/s/person ventilation rate is around 800ppm. But if people are active, it could easily be 1200ppm or higher at the same vent rate 5/

Similarly smaller people with lower metabolic rates (i.e young children) may only result in CO2 of 650ppm at 10 l/s/p (but as their breathing rates are lower they will also likely exhale and inhale less virus too) 6/

So although in most spaces we can assume that 10 l/s/person is around 800ppm, we have to be careful as it is not always the case. Now what about air changes per hour? 7/

Air changes per hour is a way of describing ventilation rate relative to the size of a room. A ventilation rate of 1 ACH essentially means that the ventilation flow in 1 hour is equivalent to the whole volume of the room 8/

ACH is very useful when you are looking at how quickly something gets removed from a space as it tells you about the turn over time for the air. But there are some things to watch 9/

1 ACH doesn’t mean all the air in the room has been replaced in that hour. The clean air mixes with the stale air and dilutes any contaminants. Each ACH removes ~63% of the contaminants that are in the room (exponential relationship) 10/

So when contamination ceases, at 1 ACH 63% will be removed in 1 hour, at 2 ACH 86% will be removed in 1 hour, at 3 ACH 95% in 1 hour, at 6 ACH 99.7% in 1 hour 11/

But when the contamination is always present (the infector stays in the room) then we get something slightly different. Increasing from 1 to 2 ACH will reduce exposure by 50%, going from 1 to 3 ACH will reduce exposure by 66%, 1 to 6 ACH reduces by 83% 12/

This difference can be important when looking at air cleaners. If a product claims to reduce by 90% in a particular room size in 1 hour, then it is equivalent to 2-3 ACH, and would only reduce steady state exposure by ~60% 13/

OK lets look at ACH compared to CO2 concentrations. This very much depends on room size. If you have a 180 m3 high school classroom with 30 people in it, then 10 l/s/p is approx. 800ppm and approx. 6 ACH. 14/

But if you take those same people and put them doing the same activity in a very large space (e.g. sports hall) at 4000 m3, then 10 l/s/p is still 800ppm, but the ACH is now 0.27 ACH 15/

Although you would want your sports hall to have a higher vent rate than 10 l/s/p to deal with the emissions during higher activities, you would be unlikely to ever get 6 ACH – this would be 222 l/s/p for 30 people and would be v hard to achieve 16/

What does all this mean for virus risk? First thing to remember is that CO2 is not a direct proxy for risk. Everyone exhales CO2 but only infected people exhale virus. CO2 tells us about the ventilation and gives clues about the far field airborne risk 17/

If we assume well mixed air in a room, we can estimate far field airborne risk from the Wells-Riley equation. This relates exposure to the total ventilation rate – so if we have ACH we also need room volume and if we have l/s/p then we need number of people 18/

If we are considering risk in shared air where people spend a long period time together (classroom, office, restaurant…) then CO2 concentration is probably the most useful ventilation metric. 19/

The reason for this is that it auto scales to some extent with activity and breathing rate and so is a more realistic measure of effective ventilation that is already adjusted for the activity and people in the setting 20/

In small spaces that are transiently occupied, CO2 is a less good measure. Values often can’t build up to steady state in the time and so it can give a false sense of assurance. People may also more likely to be at close proximity 21/

ACH is a more useful metric when you are considering small spaces that are transiently occupied and you want to know how quickly virus will be removed from the air before the next person – shops, toilets, locker rooms, GP rooms, dental rooms 22/

But remember with ACH that the size of the space does matter too. In a small space with high ACH concentration of virus may be much higher than in a very large space with low ACH as there is less air to dilute it 23/

Some other caveats to bear in mind. All of this assumes that air and contaminants in rooms are instantly and uniformly mixed. In reality this is NOT the case and so ventilation measures are approximate indicators 24/

Ventilation is hard to measure – all these metrics are tricky and there are always fluctuations from the ventilation itself, the number of people in the room and their activity and the measurement method 25/

If you are aiming for 800ppm and you are at 850ppm, that’s probably good enough. If you get some short higher spikes don’t be too concerned. Its sustained high concentrations that matter 26/

Viral aerosols are more concentrated close to the infectious source. The smallest aerosols can disperse throughout the whole space, but larger aerosols will travel shorter distances and deposit more quicky - but can still hang about long enough to be inhaled 27/

Ventilation has limited impact on exposure close to a source, and has very limited effect on larger aerosols. Therefore the ventilation rate or CO2 concentration only represents part of the risk. 28/

Mixing of aerosols in a room will depend on the particular ventilation strategy and activities in the room. There will be regions of higher or lower concentration and they don’t necessarily match the CO2 distribution 29/

We also know there is a huge range in viral emissions from people depending on viral load and activity. The ventilation can’t adapt to this and so all we can do is improve to reduce the risk – there is no such thing as guaranteed safe 30/

Overall, use the right metric for the right space and remember numbers don’t always apply to every space. Also remember ventilation is just one part of the mitigation strategy. Vaccines, testing, masks, distancing and good hygiene are also needed to keep this virus in check. 31.