Optimizing desalination for regional water systems: Integrating uncertainty, quality, and sustainability

Abstract

Regional water supply systems in arid regions worldwide face growing challenges due to increasing demands under conditions of water scarcity. For coastal regions, the addition of seawater desalination as an alternative water source can hedge against uncertainty stemming from demands placed on the system and hydrological inputs to the natural resources. In order to effectively integrate desalination into a water supply system, water managers utilize optimization models that solve for the optimal mix of desalination and natural resources to meet consumer demands at a sufficient water quality while maintaining sustainable levels in the natural resources. The complexity of managing both quantity and quality for a water supply system facing uncertainties is increased by contractual limitations between desalination plant owners and water managers that dictate the desalination production decisions to occur at the start of the year before the uncertainties have been revealed. To incorporate the misalignment between decision timelines and the order in which the uncertainty is revealed, a two-stage optimization approach is utilized where management decisions are split into implementable decisions representing the desalination production volumes, and recourse decisions representing supply volumes from the remaining sources and water allocations in the supply network. An iterative process is then undertaken in which in the first stage the implementable decisions are solved for potential uncertain scenarios, and then introduced to the second stage as inputs where the recourse decisions are iteratively solved for new uncertain scenarios. This framework is applied to the Israeli national water supply system, where the unique challenges of managing mixed salinity sources and asynchronous decision timelines in the face of uncertainty are addressed. A multicriteria decision analysis method is then applied where tradeoffs are found between the upfront investment in desalination and the expected system operational cost and robustness to failure. The inclusion of multicriteria decision analysis to analyze the multiple desalination production alternatives allows for including stakeholders preferences in model-based decision-making.