Assessment of contributions of indoor and outdoor sources to total PM-10 exposure: people living in shop houses in Bangkok

Abstract

The relationship between personal, indoor and outdoor PM-10 concentrations to which individual subjects were exposed, over a number of days was investigated for people living in roadside buildings. Nine repeated measurements covering 3 seasons were conducted among 28 non-smoking participants, living in 14 shop houses on Sukhumvit road, Bangkok. The averages of the means of personal, indoor and outdoor PM-10 concentrations were 81.6, 74.6 and 130.7 (µg/m3 respectively. The overall mean of outdoor concentrations exceeded both indoor and personal PM-10 concentrations and the levels were higher in winter rather than in summer or the rainy season. Variations of indoor PM-10 concentrations were found from floor to floor, and were especially notable at the first floor of the shop house. Nonetheless, the result showed that personal PM-10 exposure levels were correlated with the outdoor concentrations. The relationship between personal and outdoor PM-10 concentration was moderate, with a median Pearson's R correlations of 0.706. Excluding one house with car parking inside, the median Pearson’s R correlation was increased to 0.760. Aside from the outdoor PM-10 concentrations, personal PM-10 concentrations tended to be higher under conditions of incense burning, exposure to tobacco smoke from non-residents, door opening, and winter season. However, sleeping in bedrooms with an air conditioning system tended to lower the personal PM-10 concentrations. This finding supported a conclusion that outdoor PM-10 concentration could be used as a conservative exposure surrogate in the health impact epidemiological studies for this life-style group of people. The results from factor analysis showed that there were probably five major sources contributing to personal exposure PM-10 concentrations. Moreover, multiple regression analysis revealed that there were two significant attributable sources of personal PM-10 mass which showed elemental compositions frequently attributed to soil/crustal and steel industry sources. The estimated source contributions to personal PM-10 concentrations were 40%-50% for soil/crustal type and 3%-5% for steel industry type.