The deliverable consists of two distinct parts. The first part focuses on advancements in a preventive control algorithm for the power system to ensure safety with probabilistic constraints (chance constraints), taking into account uncertainties in injections. The active power control of the Phase Shifting Transformer (PST), suitably modeled, is integrated into the options available to the algorithm. The optimization problem underlying the control is reformulated using Benders decomposition, making it more efficient for large networks. The integration of the control problem with stability constraints, expressed in terms of steady-state variables, is also demonstrated.
Two different formulations for linearizing AC load flow equations are considered, and the one providing the most accurate results is used in conjunction with the optimal dispatch for security, to ensure constraints on node voltages, reactive power exchanged by generators, and reactive power flows on branches. This approach is more comprehensive as it addresses both active and reactive power. Simulations conducted on large networks confirm that the Benders decomposition formulation is more efficient than the original one and is the only one manageable by computational tools in the case of a high number of contingencies in a large power system. The second part of the document addresses methods for improving resilience analysis in large power networks: a three-step technique is proposed, based on hierarchical clustering, which considers the correlation between weather events on lines and information related to network topology. This technique aggregates lines into groups where there is a higher probability of multiple contingencies occurring together. This facilitates a more efficient selection of multiple contingencies for resilience analysis. Simulations conducted on two large sets of lines from two areas of the Italian transmission network demonstrate that the proposed technique identifies sets of lines that have indeed tripped together for the same past events.