OPERATE

Artificial reservoirs are widespread across the world, and their use ranges from the traditional management of the water resource to flood reduction operations. The importance of existing artificial reservoirs has gradually increased over the years as these structures are highly impacting and their management costs can be remarkably high. Existing reservoirs can be precious to ensure a proper management of water resources to achieve several social, financial, and environmental goals. Indeed, the actual climate crisis and the increased occurrence of extreme events (i.e., fewer and more intense rainfall events, prolonged periods of drought, etc.), leading to a higher flood frequency, urge a prompt action. In addition, artificial reservoirs undergo severe sediment yield during their lifetime, leading to a reduction of the useful storage capacity. To this regard, the optimal management of existing artificial reservoirs could considerably mitigate these effects, thus also providing benefits to managing authorities and users, as well as to the communities living nearby. The aim of this project proposal is to develop an innovative methodology for the optimal management of the water resource stored within existing artificial reservoirs to maximize hydroelectric power production and optimize water resources utilization, leading to a maximized income for managing authorities that could be used for the management of the infrastructure itself. In particular, the concept of reliability will be overstretched within the optimization model, uncertainty quantification and analysis, while stochastic models and Scientific Machine Learning (SciML) techniques for both simulation and prediction will be developed coupled with the analysis and quantification of the uncertainties. A software suite will be created containing different models for both long term optimization and short-term optimal control of the water volume within the reservoir.