Development of decision making tools for street planning in high density cities

Abstract

In recent decades, a vast number of tall buildings have been built in response to rapid process of urban city developments. However, the construction of tall buildings in urban streets has led to many environmental problems, in particular poor air quality inside urban canyons. A number of measures have been proposed for mitigating the air quality inside urban streets. Among them, street and building configurations inside isolated deep canyons form the major focuses of the first part of thesis. Prior to identification of suitable street and building configurations, the air pollutant exposures for the major population subgroups present in canyons were estimated and compared using the indirect exposure approach. The indirect approach was applied to estimate the exposures of three different population subgroups (i.e. pedestrians, shop vendors and residents) in a canyon by multiplying the air pollutant concentrations with the duration of exposure within the canyon. The air pollutant concentrations were predicted by Computational Fluid Dynamics models based on the Reynolds-Averaged NavierStokes (RANS) equations with the standard k-{464} turbulence model. Models were constructed to simulate 162 different street and building configurations in both tree-free and tree-lined canyons with aspect ratios of 2, 4 or 6 (i.e. AR2, AR4 or AR6). The findings from this part of the study indicated that the personal exposures of shop vendors were extremely high if they were present inside a canyon without any setback or separation between buildings and when the prevailing wind was perpendicular to the canyon axis. Their exposures were doubled if they stayed in a canyon with tree planting. Building separation and building setbacks were found to be effective in reducing personal air exposures in tree-free or tree-lined canyons under perpendicular wind despite their effectiveness being varied with the configurations. If the reduction in the total development floor area (the total floor area permitted to be developed within a particular site area) was also taken into consideration, vertical setbacks were recommended for tree-free canyon AR2 and horizontal setbacks were recommended for tree-free canyon AR4 and AR6. However, the recommendation would be different for tree-lined canyons where building spacing was more preferable for AR2 and horizontal setbacks were more preferable for AR4 and AR6. The findings arising from this part of the study were subsequently used to formulate a hierarchy decision flow chart which lays out the priority orders of factors for consideration during street planning after taking into account both air pollutant health exposure and loss in floor development area. The priority orders were determined to be as follows: prevailing wind direction, aspect ratio, building design element (i.e. building separation and building setbacks), and tree planting location (in case a tree-lined street). Alternatively, if people's preferences for street and building configurations had also been taken into account together with the air pollutant exposure, a decision making tool employing quadrant analysis was developed to help planners to identify the optimum configurations which gave lower air pollutant exposures and were highly preferred by public. The development of this tool was based on the exposures resulted from by CFD and people's preference responses obtained for three different streetscape elements including permeability (i.e. spacing between buildings), aspect ratio (i.e. ratio of street width to building height), and tree planting through a questionnaire survey designed using Discrete Choice Experiments. The survey results from 509 respondents in Hong Kong indicated that individuals holding different perceptions of the impact of trees on air quality were found to affect their preferences for the streetscape attributes. Permeability was the most preferred attribute while tree planting and aspect ratios were less preferred. The findings arising from the survey were finally input to the quadrant analysis for identifying the best options, which were subsequently determined to be the ones with low aspect ratios, tree planting on both sides of the street and either low or high permeability. In addition, this thesis also demonstrated a successful development of procedures to formulate decision making tools for urban street planning and it serves as a valuable linkage between engineering sciences and urban street planning.