MOP: first industrial application of Stochastic Dual Dynamic Programming for optimization of the Uruguayan power system

This paper introduces MOP, a development that implements the first application of Stochastic Dual Dynamic Programming (SDDP) for optimization of the Uruguayan power system. The study addresses the challenge of efficiently managing a system characterized by a large hydroelectric component and an increasing integration of non conventional renewable energy sources.A detailed case study of the Uruguayan system is conducted to demonstrate the effectiveness of the SDDP algorithm. The analysis highlights robust convergence and compares the performance of SDDP with standard Stochastic Dynamic Programming methods (SDP). The results show that the SDDP algorithm effectively manages system complexity and yields accurate optimization outcomes. Additionally, the successful application in the current system suggests the scalability of the method to systems with more state variables.