We know the values of NOx, CO and soot production for vehicles based on the PIARC reports (for fluent traffic, as well as for traffic jams), the traffic flow from a detailed traffic forecast analysis and the age/fuel type of the cars as brackets from national statistics. This allows us to estimate mean/peak emissions within the tunnel with reasonable accuracy. We know what is the airflow velocity / ventilation capacity for all of the traffic scenarios as well. Now, we have to propose a filtration solution, for which we first need to know, how can we translate our emissions into PM2.5 / PM10.
A very easy assumption for our further modelling, would be that the most of the particulates generated in the tunnel are formed from the soot. I'm not sure if it is possible for larger particulates to form from NOx / O3, as the tunnel is constantly ventilated (either through traffic, or by mechanical ventilation in a traffic stop). I would also assume that mass of soot >>> mass of particulates emitted by breaks, tyres etc., which means we could ignore the latter (especially for traffic stop scenario).
Please share your experience in this (particulate) matter.