This report describes the final results of the activity devoted to developing an algorithm for the detection and localization of faults in distribution networks using measurements from Phasor Measurement Units (PMU). Based on what has already been developed, the algorithm has been improved so as to overcome the critical issues that emerged, especially from the identification and localization of single-phase faults in networks with compensated neutral. The positioning criterion of the measurement devices has also been studied in depth and the definition of clusters (network portions) within which the fault is located has been formalized, in the case fewer PMUs than network nodes exist. In order to extend the functionality deriving from measurements from PMUs, the possibility of using such data for the automatic identification of the network topology (status of disconnectors and circuit breakers) has also been studied in depth, relying both on a bibliographic analysis and on simulations.
The latter have been performed via the MATLAB/Simulink software using two network models (one of which composed of 103 nodes) and considering different cases such as different degrees of imbalance, imperfect knowledge of the network parameters, and different configurations of the neutral state; they have shown a performance improvement compared to those obtained using the previous version of the algorithm.
However, some critical issues remain in the localization of single-phase earth faults, in the case of network operation with inductance-compensated neutral.
Furthermore, the proposed method for the automatic identification of the network topology has been validated through simulations, following an analysis of the reference bibliography. Preliminary simulations have shown that the method is able to correctly detect the state of the switching devices and therefore the current topology.