Indoor air quality modeling for PM10, PM2.5, and PM1.0 in naturally ventilated classrooms of an urban Indian school building

Abstract

Assessment of indoor air quality (IAQ) in classrooms of school buildings is of prime concern due to its potential effects on student's health and performance as they spend a substantial amount of their time (6-7 h per day) in schools. A number of airborne contaminants may be present in urban school environment. However, respirable suspended particulate matter (RSPM) is of great significance as they may significantly affect occupants' health. The objectives of the present study are twofold, one, to measure the concentrations of PM 10 (<10 μm), PM2.5 (<2.5 μm), and PM 1.0 (<1.0 μm) in naturally ventilated classrooms of a school building located near a heavy-traffic roadway (9,755 and 4,296 vehicles/hour during weekdays and weekends, respectively); and second, to develop single compartment mass balance-based IAQ models for PM10 (NVIAQM pm10), PM2.5 (NVIAQMpm2.5), and PM 1.0 (NVIAQMpm1.0) for predicting their indoor concentrations. Outdoor RSPM levels and classroom characteristics, such as size, occupancy level, temperature, relative humidity, and CO2 concentrations have also been monitored during school hours. Predicted indoor PM10 concentrations show poor correlations with observed indoor PM10 concentrations (R 2 = 0.028 for weekdays, and 0.47 for weekends). However, a fair degree of agreement (d) has been found between observed and predicted concentrations, i.e., 0.42 for weekdays and 0.59 for weekends. Furthermore, NVIAQMpm2.5 and NVIAQMpm1.0 results show good correlations with observed concentrations of PM2.5 (R 2 = 0.87 for weekdays and 0.9 for weekends) and PM1.0 (R 2 = 0.86 for weekdays and 0.87 for weekends). NVIAQMpm10 shows the tendency to underpredict indoor PM10 concentrations during weekdays as it does not take into account the occupant's activities and its effects on the indoor concentrations during the class hours. Intense occupant's activities cause resuspension or delayed deposition of PM10. The model results further suggests conductance of experimental and physical simulation studies on dispersion of particulates indoors to investigate their resuspension and settling behavior due to occupant's activities/movements. The models have been validated at three different classroom locations of the school site. Sensitivity analysis of the models has been performed by varying the values of mixing factor (k) and newly introduced parameter R c. The results indicate that the change in values of k (0.33 to 1.00) does not significantly affect the model performance. However, change in value of R c (0.001 to 0.500) significantly affects the model performance. © 2010 Springer Science+Business Media B.V.