Estimation of Air Quality Criterion-based Building Setback Distance From Urban Roads and Freeway Using Microscopic Simulation of Traffic, Emissions, and Pollution Dispersion

Abstract

Emissions from road vehicles and other sources that exceed well-defined concentrations affect human health. To protect people, especially vulnerable members of population, ambient air quality standards and guidelines should be met. According to projections, the internal combustion engine vehicles are likely to maintain their dominance for many decades and only marginal improvements in emission reduction can be expected. Therefore, it is necessary to investigate separation distance between vulnerable receptors and high traffic roads and highways. Available guidelines on setback distance from urban roads and highways and the associated land use considerations do not account for pollutant dispersion and air quality standards. Also, methods are not available to predict pollutant concentration at various setback distances and heights under worst case of traffic and wind condition. This research is aimed at the development of methods for determining air quality criterion-based building setback distance from urban roads and freeways. The research approach integrates microscopic models for simulating vehicle trajectories, emission factors and pollutant dispersion. Prior to application, the models were verified with field data on vehicle trajectories and emissions. In addition, data on background pollution concentrations were investigated. The integrated simulation model when applied to selected building locations in Ottawa produced logical results. To meet the objective of developing tools that could be used to check pollutant concentrations at various distances and heights or to find setback distance for specified conditions, the simulation design methodology was developed and applied while meeting the statistical reliability requirements. Simulation results were used to develop statistically significant predictive models for CO, NO2, and PM2.5 concentration. The non-linear regression models are better than linear regression models and the ANN models compare well with non-linear models. The new predictive models are designed to provide 1-hour pollutant concentration on the worst case basis. Example applications show logical results of potential interest to professional persons and researchers. In addition, ideas are advanced for future research. It is contended that the products of this research can be of immediate assistance to the real-world planners and policy analysts in establishing setback guidelines that take into account air quality standards.