Factors influencing variability in the infiltration of PM_(2.5) mass and its components

摘要

The infiltration of particles into homes can vary seasonally, between homes in a community and between communities. However, few studies have examined the day to day variability across multiple homes. We used continuous data collected from a 2-year (2005-2006) personal exposure study conducted in Windsor, ON to estimate daily infiltration factors (F_(inf) )for fine particulate matter (PM_(2.5)), Black Carbon (BC), and ultrafine particles (UFP) as well as the ambient personal exposure factor (F_(pex)) for PM_(2.5). In addition, the daily ambient and non-ambient generated components of indoor and personal concentrations were estimated. Median daily F_(inf) estimates ranged from 0.26 to 0.36 across seasons for PM_(2.5); from 0.28 to 0.59 for BC; and from 0.15 to 0.26 for UFP. Median daily F_(pex) estimates ranged from 0.24 to 0.31 across seasons. Daily PM_(2.5) and UFP F_(inf) and F_(pex) estimates were higher in summer than winter, although BC showed the opposite trend. Predictors of daily infiltration were typically related to window-opening behaviours, air conditioning, meteorological variables, and home age. In addition, use of electrostatic precipitators and stand alone air cleaners was associated with significantly reduced infiltration factors, indicating that these devices may provide a cost effective mechanism of reducing human exposures to particles of ambient origin. The majority of indoor PM_(2.5) (median 57-73%) and indoor BC (median 90-100%) was of ambient origin across seasons, while both personal PM_(2.5) and indoor UFPs had significant non-ambient contributions (median 60-65%). Factors that were found to increase non-ambient particle concentrations were typically related to cooking, candle use, supplemental heating, cleaning, and number of people in the home. Factors that were found to decrease non-ambient particle concentrations were open windows, and air cleaner use. This work has several implications to both epidemiologic studies and risk management. A better understanding of the factors influencing F_(inf) and F_(pex) can improve exposure assessment and contribute to reduced exposure misclassification in epidemiologic studies. Furthermore, by increasing our knowledge of non-ambient and ambient exposures, risk associated with PM exposure can be managed more effectively.