Using central site measurement data to predict personal exposure to particulate matter (PM) is challenging, because people spend most of their time indoors and ambient contribution to personal exposure is subject to infiltration conditions affected by many factors. Efforts in assessing and predicting exposure on the basis of associated indoor/outdoor and central site monitoring were limited in China. This study collected daily personal exposure, residential indoor/outdoor and community central site PM filter samples in an elderly community during the non-heating and heating periods in 2009 in Tianjin, China. Based on the chemical analysis results of particulate species, mass concentrations of the particulate compounds were estimated and used to reconstruct the PM mass for mass balance analysis. The infiltration factors (Finf) of particulate compounds were estimated using both robust regression and mixed effect regression methods, and further estimated the exposure factor (Fpex) according to participants' time-activity patterns. Then an empirical exposure model was developed to predict personal exposure to PM and particulate compounds as the sum of ambient and non-ambient contributions. Results showed that PM mass observed during the heating period could be well represented through chemical mass reconstruction, because unidentified mass was minimal. Excluding the high observations (>300μg/m3), this empirical exposure model performed well for PM and elemental carbon (EC) that had few indoor sources. These results support the use of Fpex as an indicator for ambient contribution predictions, and the use of empirical non-ambient contribution to assess exposure to particulate compounds.
Keywords: Empirical exposure model; Infiltration factor; Mass balance; PM(10); Particulate compound; Personal exposure factor.
Copyright © 2016 Elsevier B.V. All rights reserved.