Society for Sustainability and Environmental Engineering Conference November 2009

- John Baulis, Lisa Lloyd, Fiona Paton & Ben Staniford, School of Civil, Environmental and Mining Engineering, The University of Adelaide
A major challenge this century is how to reliably supply water to urban areas under the increasing pressures of population and income growth, urbanisation and climate change. Substantial research in this field has been undertaken but this has largely focused at the local and unit scales, rather than the regional scale. This study presents a novel approach to assess the sustainability of and to optimally design an urban water supply system (UWSS) at the regional scale, over a range of planning horizons. This firstly includes a risk based performance assessment to measure a proposed system’s reliability, resilience and vulnerability, which can help governments and water planning authorities to sustainably manage the UWSS. Secondly, a multi-objective optimisation process is used to design the optimal system in terms of economic cost and greenhouse gas (GHG) emissions. The multi-objective genetic algorithm NSGAII is used to solve the optimisation problem and produce a Pareto Front, illustrating the optimal trade-offs between the two conflicting objectives, which can be used by decision makers to help identify the optimal design of the UWSS. Finally, a sensitivity analysis tests the robustness of the solutions. The approach is found to be sound when applied to a case study based on Metropolitan Adelaide’s Southern water supply system for three planning horizons; 2020, 2060 and 2100. Findings of the risk based performance assessment indicate that future plans are adequate for 2020 but inadequate for 2060 and 2100. However, the sensitivity analysis demonstrates that this performance significantly depends on demand, rainwater tank size, the River Murray supply constraint and climate change impacts. Results of the design process show that a small trade-off exists between economic cost and GHGs and that the optimal range of rainwater tank size is fairly robust, whereas the optimal range of the desalination plant size is greatly influenced by demand.