Keywords
Multi-sector participatory planning; Decision Making Under Uncertainty; Robust Many Objective Optimization
Start Date
15-9-2020 7:20 PM
End Date
15-9-2020 7:40 PM
Abstract
Increasing water scarcity and uncertainty in future conditions can lead to growing competition between different water consuming sectors. In response, decision makers are increasingly employing Decision Making Under Uncertainty (DMU) methods which combine simulation models with search algorithms to identify robust portfolios of supply, demand management and policy interventions and quantify their efficient performance trade-offs. In this study we describe a multi-year regional multi-sector participatory planning effort considering the public water supply, environment, agriculture and energy sectors undertaken in the East of England, one of the driest parts of the UK. A large regional infrastructure and environmental system simulator was built and connected to many-objective heuristic search to optimize six aggregated regional planning objectives representing multi-sector priorities under several climate, demand and hydrological scenarios. The simulation model tracked approximately 1000 performance metrics representing local sub-regional performance in each sector. The Pareto-approximate portfolios of interventions and their implied trade-offs were interrogated by stakeholders using web-based interactive trade-off analysis. Using minimum performance thresholds on optimization objectives and tracked metrics to filter portfolio performance stakeholders identified a balanced and efficient portfolio that met each sector’s needs. The selected portfolio was found to use a moderate level of storage and transfer capacity. Tracked metrics revealed local trade-offs between agricultural supply, deviations from required environmental flows, and public water supply and helped with portfolio selection. Further analysis using stakeholder performance thresholds revealed two alternative infrastructure development visions for the region, using new storage dams and water transfers to differing extents.
Planning regional multi-sector water resource systems in East England
Increasing water scarcity and uncertainty in future conditions can lead to growing competition between different water consuming sectors. In response, decision makers are increasingly employing Decision Making Under Uncertainty (DMU) methods which combine simulation models with search algorithms to identify robust portfolios of supply, demand management and policy interventions and quantify their efficient performance trade-offs. In this study we describe a multi-year regional multi-sector participatory planning effort considering the public water supply, environment, agriculture and energy sectors undertaken in the East of England, one of the driest parts of the UK. A large regional infrastructure and environmental system simulator was built and connected to many-objective heuristic search to optimize six aggregated regional planning objectives representing multi-sector priorities under several climate, demand and hydrological scenarios. The simulation model tracked approximately 1000 performance metrics representing local sub-regional performance in each sector. The Pareto-approximate portfolios of interventions and their implied trade-offs were interrogated by stakeholders using web-based interactive trade-off analysis. Using minimum performance thresholds on optimization objectives and tracked metrics to filter portfolio performance stakeholders identified a balanced and efficient portfolio that met each sector’s needs. The selected portfolio was found to use a moderate level of storage and transfer capacity. Tracked metrics revealed local trade-offs between agricultural supply, deviations from required environmental flows, and public water supply and helped with portfolio selection. Further analysis using stakeholder performance thresholds revealed two alternative infrastructure development visions for the region, using new storage dams and water transfers to differing extents.
Stream and Session
false