Modeling pollutant penetration across building envelopes

As air infiltrates through unintentional openings in building envelopes, pollutants may interact with adjacent surfaces. Such interactions can alter human exposure to air pollutants of outdoor origin. We present modeling explorations of the proportion of particles and reactive gases (e.g., ozone) that penetrate building envelopes as air enters through cracks and wall cavities. Calculations were performed for idealized rectangular cracks, assuming regular geometry, smooth inner crack surface and steady airflow. Particles of 0.1-1.0 mum diameter are predicted to have the highest penetration efficiency, nearly unity for crack heights of 0.25 mm or larger, assuming a pressure difference of 4 Pa or greater and a flow path length of 3 cm or less. Supermicron and ultrafine particles are significantly removed by means of gravitational settling and Brownian diffusion, respectively. In addition to crack geometry, ozone penetration depends on its reactivity with crack surfaces, as parameterized by the reaction probability. For reaction probabilities less than similar to 10(-5), penetration is complete for cracks heights greater than similar to1 mm. However, penetration through mm scale cracks is small if the reaction probability is similar to 10(-4) or greater. For wall cavities, fiberglass insulation is an efficient particle filter, but particles would penetrate efficiently through uninsulated wall cavities or through insulated cavities with significant airflow bypass. The ozone reaction probability on fiberglass fibers was measured to be 10(-7) for fibers previously exposed to high ozone levels and 6 x 10(-6) for unexposed fibers. Over this range, ozone penetration through fiberglass insulation would vary from > 90% to similar to 10-40%. Thus, under many conditions penetration is high; however, there are realistic circumstances in which building envelopes can provide substantial pollutant removal. Not enough is yet known about the detailed nature of pollutant penetration leakage paths to reliably predict infiltration into real buildings. (C) 2001 Elsevier Science Ltd. All rights reserved.