Influence of virtual changes in building configurations of a real street canyon on the dispersion of PM10

•PM10 simulations in a street canyon are carried out using a CFD code, FLUENT.•Influence of different building configurations on peak PM10 levels is studied.•Modelled CFD results are compared with the measured PM10.•Virtual opening in buildings reduced peak PM10 concentrations.•Building configurations found to be important for controlling PM10 concentrations.

Four geometrical configurations of a real street canyon in Barreiro city (Portugal) are considered to study their influence on the dispersion of PM10. These configurations include actual architectural layout of the street (Option 1), and three virtual cases (Options 1–3). Option 2 includes the modification of real geometry by including 4 m gaps between the buildings situated on the southern part of the street canyon. Option 3 considers 6 m gaps between buildings as opposed to 4 m gaps in Option 2. Option 4 assumes the same height for all buildings on the southern part of the street canyon, with no gaps between buildings. Computational fluid dynamics code (CFD), FLUENT, is used to simulate the detailed flow and turbulence characteristics in three-dimensional domain of chosen street canyon, together with the PM10 dispersion for both the summer and winter seasons. The modelled PM10 concentrations were then compared with the measured data at seven different locations in the street canyon. Our results indicate up to 23% lower PM10 concentrations at 1.5 m above the road level during the along-canyon wind direction due to the channelling of flow, compared with those observed during the cross-canyon wind direction. Detailed inspection of the results obtained from the Options 1–3 indicated that the spacing between the buildings tend to increase particle dilution during the cross-canyon winds, resulting in up to 20, and 22% reduced concentrations for options 2, and 3 respectively, compared with the actual configuration (Option 1). The largest improvement (∼7%) in the PM10 concentrations was given by Option 2, while other options showed modest changes. Possible reasons for these changes under varying meteorological conditions are explained in the context of changing building configurations and their implications in city planning.