A resilience-based methodology for improved water resources adaptation planning under deep uncertainty with real world application

This is the author accepted manuscript. The final version is available from Springer Verlag via the DOI on this record.Resilience of a water resource system in terms of water supply meeting future demand under climate change and other uncertainties is a prominent issue worldwide. This paper presents an alternative methodology to the conventional engineering practice in the UK for identifying long term adaptation planning strategies in the context of resilience. More specifically, a resilience based multi-objective optimization method is proposed that identifies Pareto optimal future adaptation strategies by maximizing a water supply system’s resilience (calculated as the maximum recorded duration of a water deficit period over a given planning horizon) and minimizing total associated costs, subject to meeting target system robustness to uncertain projections (scenarios) of future supply and demand. The method is applied to a real-world case study for Bristol Water’s water resource zone and the results are compared with those derived using a more conventional engineering practice in the UK, utilizing a least-cost optimization analysis constrained to a target reliability level. The results obtained reveal that the strategy solution derived using the current practice methodology produce a less resilient system than the similar costing solutions identified using the proposed resilience driven methodology. At the same time, resilience driven strategies are only slightly less reliable suggesting that trade-off exists between the two. Further examination of intervention strategies selected shows that the conventional methodology encourages implementation of more lower cost intervention options early in the planning horizon (to achieve higher system reliability) whereas the resilience-based methodology encourages more uniform intervention options sequenced over the planning horizon (to achieve higher system resilience).This work was financially supported by the UK Engineering and Physical Sciences Research Council, HR Wallingford and The University of Exeter through the STREAM Industrial Doctorate Centre. We thank Bristol Water for allowing the use of their data and information, which is available from their publicly available water resources management plan