Abstract

The impacts of climate change such as hydrological droughts have motivated significant changes in reservoir management strategies. The operation of multipurpose reservoirs is vital for efficiently utilizing stored water resources, serving various needs such as electricity generation and agricultural irrigation. Despite substantial efforts to support decision-making within each economic sector, there is a notable lack of comprehensive integration across these sectors in concurrent analyses. To address this gap, we propose an integrated approach that combines a large-scale hydrothermal scheduling model with a basin-scale water resources model to comprehensively analyze the operations of both the power and agricultural systems. This approach enables the assessment of operational policies for multipurpose reservoirs and their performance at both local and regional scales under diverse hydrological scenarios. An essential modification analyzed here is the prioritization of water allocation to agricultural users. We rigorously evaluate the impacts of this modification across various hydrological conditions, using a prominent Chilean basin as a case study. Applying this methodology to Laja Lake, the largest Chilean multipurpose reservoir with significant hydroelectric capacity and extensive agricultural areas, we analyze the situation in 2025 and find that hydrological variations directly affect both the electrical and agricultural aspects of performance. During drought conditions, there is a noticeable increase in thermal generation, costs, emission intensity, and water deficits. Furthermore, the adjusted policy reveals intricate trade-offs between emissions from the power sector and agricultural water deficits. In drier scenarios, improving agricultural reliability incurs minimal additional operational costs and reduces power sector emissions, supporting the adoption of a policy aligned with net-zero objectives.