Abstract
Current studies on modeling indoor exposures use averaged values of contaminant concentrations in ventilated rooms, which is limiting for a detailed study of the indoor exposure. Lack of an appropriate validation of indoor exposure models is another limit in the modeling approach. In addition, the current experimental approach uses point sources, which are not a representative type of sources that exist in the building environment. The objective of this dissertation is to address these gaps.Studies on the effects of airflow patterns on indoor exposures were carried out by both the experimental approach and the modeling approach. An area source was constructed in full-scale rooms ventilated by four different ventilation systems. With the experimental data, we evaluated a common turbulence model in predicting the contaminant dispersion. The model was validated qualitatively by a graphical method. Quantitative evaluation using the statistical method shows that in order to predict more accurate exposures, the airflow field has to be well predicted.The validated model has been used to further study the contaminant dispersion. The distribution depends on the airflow pattern, the source height, and the source type. Displacement ventilation generally creates a stratified concentration distribution, with lower concentrations in the occupied zone when the source is uniformly distributed in the room. This stratification may disappear when the passive concentrated source is located at the lower level or close to the supply diffuser. The general exposure level could be estimated through analyzing the relative source positions in the airflow pattern. The location of the exhaust outlet may not very much affect the airflow pattern, but can significantly affect the exposure level in the room.Exposure caused by a dynamic source (decane and paint) was also studied in rooms with a mixing type of ventilation. We found that the distribution pattern is established soon after the application of the source and its change with time is negligible. This indicates that the dynamic characteristics of a source and the resulted distribution in the room could be studied separately, which will greatly simplify the study. The data obtained from dynamic source tests were used to validate a source model. A step-by-step procedure was developed to validate the model to predict emissions in a full-scale chamber with information obtained from small chamber tests.Finally, we developed a regression method to obtain emission parameters based on small chamber emission data. This method is simple and easy to apply. By this method we have obtained emission parameters for twenty dry materials. The sensitivity analysis shows that measurement error contributes greatly to the uncertainties and that this method may not be appropriate to obtain partition coefficients of small values.