Penetration of particles into buildings and associated physical factors. Part I: Model development and computer simulations

The PM2.5 standard proposed by the U.S. Environmental Protection Agency (EPA) has stimulated research on the relationships between particulate matter concentrations and the exposures and subsequent health responses of sensitive subpopulations, such as the elderly. Since individuals in these subpopulations may spend more than 90% of their time indoors, understanding the relationship between outdoor particle concentrations and those found in indoor microenvironments is critical. This research resulted in a time-dependent indoor air quality model incorporating all potential particle sources and loss mechanisms. Monte Carlo simulations of the model identified the mechanisms, such as particle loss during penetration through the building envelope, that modify the outdoor particle size distribution during transport into the interior of a building, calculated indoor-to-outdoor (I/O) concentration ratios, and estimated penetration factors as a function of particle size. Indoor particle generation and transport of outdoor particles through the HVAC system were the most important contributors to the indoor concentration in residential and commercial buildings, respectively. The most significant removal mechanisms included ventilation through and particle removal by the HVAC filter if an HVAC system was present, or particle deposition on indoor surfaces if an HVAC system was not present. The modeled I/O concentration ratios varied between 0.05 and 0.5, depending on particle size and type of ventilation system, and agreed well with published experimental results, Penetration factors less than unity were calculated for particles with aerodynamic diameters larger than 0.2 mum if the air exchange rate and steady-state I/O concentration ratio were correlated during the simulations. An additional correlation between the air exchange rate and particle deposition velocity is required if penetration factors less than unity are to be modeled for particles with aerodynamic diameters smaller than 0.2 mum. These results support the possibility that appropriate experimental studies will yield penetration factors less than unity.