Two-stage robust scheduling of a wind-powered reverse osmosis desalination plant with detailed energy recovery device modeling

Freshwater scarcity has become a critical global challenge, making seawater desalination an increasingly important solution for sustainable water supply. Among various desalination technologies, reverse osmosis (RO) desalination is widely adopted due to its high efficiency and technological maturity. However, RO desalination plants are highly energy-intensive, resulting in considerable operating costs and carbon emissions. To address this issue, integrating renewable energy sources into RO desalination systems has attracted growing attention. However, existing studies on the operational scheduling of RO desalination plants often either neglect the energy recovery device (ERD) in the desalination process or adopt an overly simplified representation of the ERD. Moreover, uncertainty-aware operational scheduling of wind-powered RO desalination plants has received relatively limited attention. Using an RO desalination plant as a representative case, this paper develops a scheduling framework that integrates a detailed ERD representation with two-stage robust optimization (TRO) to account for wind power uncertainty. A detailed ERD model, together with its operational constraints, is incorporated into the economic dispatch framework of the RO desalination plant. Wind power uncertainty is addressed within the TRO framework, which is solved using the column-and-constraint generation (C&CG) algorithm. In addition, a robustness evaluation model is constructed to assess the effectiveness of the proposed method in managing system uncertainties. Case studies show that the detailed PX model yields an energy recovery efficiency of approximately 95%. Furthermore, for wind power forecast errors of up to 45%, the proposed robust scheduling strategy maintains advantages in both economic performance and robustness compared with the deterministic approach.