Modelling Interactions Between Lot-Scale Decentralised Water Infrastructure and Urban Form – a Case Study on Infiltration Systems

Abstract

Modelling the design and implementation of urban water infrastructure (particularly decentralised systems) for strategic planning and policymaking requires detailed information of the spatial environment and quantitative knowledge of social preferences. Currently available models, however, mostly use land use, population and impervious cover data without much regard for detailed urban form or society. This study develops an algorithm for determining urban form from minimal spatial data input by incorporating local planning regulations. The interaction between urban form and implementation of lot-scale infiltration systems under different social, biophysical and climate constraints is then investigated, firstly by looking at how this varies in different residential land uses and subsequently in a case study of a typical Melbourne residential subdivision of mixed land uses. Feasibility of infiltration and its downstream impact (runoff volume, frequency and pollution) were assessed for a range of social preferences (quantified as allowable garden space) and climate scenarios (30 % increase/decrease in rainfall and evapotranspiration). Performance indicators were determined through long-term simulation with the MUSIC software. Results show how different biophysical, planning, social and climate conditions affect infiltration feasibility as well as system performance. High infiltrating soils, for example, allow smaller, well-performing and socially less-imposing systems. Low infiltrating soils lead to larger system sizes, occupy much of the allotment’s garden space, but nevertheless provide the benefit of runoff frequency reduction. Overall, climate impact was not significant except for areas with poorly infiltrating soils. Joint consideration of social, planning, climate and water management aspects potentially allows more efficient policymaking, as an array of system configurations can be tested against different multi-faceted scenarios. Such models can help facilitate better participatory planning and policymaking.