A Method of Estimating Sub-micrometer Particle Concentrations from Roadside Measurements in an Urban Environment

This paper reports on a co-operative research project between the Queensland University of Technology (QUT) and Queensland Transport to predict the contributions that motor vehicle emissions from various fleet compositions and traffic conditions make to the level of airborne sub-micrometer particles. Extensive data relating to sub-micrometer particle concentration levels and the corresponding contributing vehicles have been collected at two significantly different Brisbane sites. One site is a straight section of the Pacific Motorway, with average free flowing peak hour flows of 1650 vehicles per hour per lane, of which 15% were LCVs and HCVs. The other site is a signalised intersection on a major urban freight arterial route outside a public hospital where the traffic has a significantly larger proportion of HCV undergoing stop/start conditions. These sites have significantly different fleet compositions and driving cycles resulting in differing particle concentration levels and speciation. Particle concentration data at each site was collected over several weeks and matched to the corresponding vehicle counts. Fleet composition was determined from analysis of video and other data. Meteorological data (eg wind speed and direction, temperature, precipitation, humidity) and terrain data at the sites were also recorded. From an analysis of this data, the effect of various meteorological factors on the measured concentration levels was determined and thence an estimate of the amount of tailpipe particle emissions for the particular fleet compositions and driving cycles was made. Functional dependencies between time of a day (for week and weekend days) and speciated vehicle flow were developed. Flow rates for specific time of day /day of week, were statistically analysed in order to predict the average flow for those periods of interest within a 95% confidence level. Vehicle flow rates were further categorised by fuel type. From this analysis, the flow rates for different fuel categories for the period of interest can be predicted. From the statistical analysis of this data, models have been developed that can predict worst case levels of sub-micrometer particle concentrations with a satisfactory level of confidence. The models also indicate how the concentrations are affected by changes in the proportions of petrol/diesel vehicles and increases in CVs. This has significance for use in modelling to assess environmental impacts of alternative transport options.