Abstract

Understanding airflow and contaminant transport patterns in building helps engineers to design systems that effectively ventilate all spaces. Computer simulation models are used to predict air and contaminant flow as inexpensive alternatives. Several airflow and contaminant dispersion models have been developed to study air distribution and indoor air quality in buildings. A comprehensive research was conducted to validate the three multi-zone airflow models--COMIS, CONTAM and ESP-r. The validation process was carried out at different levels: fundamental comparisons of three models, validation with experimental data collected in a controlled environment test in a laboratory, and validation with field measurement data of two single-family houses in Ottawa. An approach was developed to distribute the measured air leakage characteristics of the whole house with the garage through cracks and gaps on exterior walls and roof, before the simulation with field measurement data. After that, COMIS and CONTAM were used to predict indoor air contaminant dispersion in the field measurement houses caused by car exhaust in the attached garages. The validation comparisons indicate that there are good agreements between the predictions made by COMIS, CONTAM and ESP-r; there are, however, some differences between these models' predictions and the measured data. The contaminant concentrations predicted by COMIS and CONTAM illustrate that the air-tightness of garage/house interface wall has significant impact on the contaminant dispersion in the rooms of houses.