Stochastic Economic Dispatch for Combined Water–Energy Microgrid Systems Considering Data-Driven Scenario Generation Approach

Numerous bilateral opportunities exist within the water–energy nexus, leading to an increased focus on the efficient operation of combined water–energy microgrid systems. Emphasis is being placed on developing methods that prescribe the most economical and efficient use of distributed energy resources and the design of energy-efficient water systems. However, achieving this objective is often complicated by significant uncertainties in the availability and variability of renewable energy sources, along with fluctuations in water demand, which introduce complexities into the optimization process and impact the reliability of traditional deterministic models. This work develops a k-means algorithm with a scenario reduction method for scenario generation to solve a two-stage stochastic economic dispatch problem. The algorithm is tested for Telford Borough in Pennsylvania. The proposed framework provides reliable economic power dispatch for water supply systems while accounting for uncertainties in future energy demands, which are directly linked to fluctuating water demand. The study demonstrates the advantages of employing a stochastic scheduling approach while comparing it to the deterministic power dispatch results. The findings demonstrate the effectiveness of the method in enhancing operational stability, achieving up to 12% operational cost savings. They also highlight the importance of accurate forecasting for robust energy dispatch in large-scale integrated water–energy microgrid systems.