Omid Seyedashraf

Sustainable urban drainage systems are nature-based engineering assets used to enhance urban flood resilience by increasing the local infiltration rate. The design of sustainable drainage infrastructure must consider several co-benefits and performance objectives, which makes the identification of optimal trade-offs a daunting task. In this work, a many-objective optimisation approach is used in conjunction with state-of-the-art visualisation techniques to facilitate the analysis of trade-offs between several stakeholders’ objectives. To this end, a genetic algorithm based on a controlled non-dominated sorting approach is linked to an urban rainfall-runoff model that evaluates the hydrologic response of the system. The search engine explores possible asset types and combination of assets, their area extent and spatial distribution, and determines a Pareto-optimal set of design options that minimises system flood volume, total suspended solids, average flood duration, peak runoff, and capital costs. An interactive parallel plot is used to present the results and identify sustainable drainage designs based on stakeholders’ interests. It is shown how the approach can facilitate collaboration between decision-makers and help them maximise the benefits of sustainable drainage infrastructure. Moreover, a pairwise correlation analysis is carried out between non-monetary objective functions, which shows that flood duration has a poor correlation with other objectives and should therefore be considered as an independent performance metric in the design of sustainable urban drainage systems.