Earth fault location for medium voltage distribution grids with distributed additional voltage measurements

This PhD thesis tries to design an efficient fault location algorithm tailored for long and highly ramified medium voltage distribution feeders. The impedance-based method solves the multiple estimation problem faced by single-end methods by leveraging some additional optimally placed voltage measurements. It has been demonstrated that this method is able to locate any type of earth faults into a connected set of nodes being the solution area, the size of which only depends on location and number of said additional measurements. In order to deal with possible measurement errors and zero-sequence grid impedance values uncertainties, a probabilistic formulation is proposed. This formulation achieves to be robust to both sources of errors at the cost of a minor increase in the size of the solution area. Finally, a technical-economic analysis is carried out in order to search for the most profitable number of additional measurements to deploy. The potential economic gains for the system operator from an increase in the reliability are estimated, assessing the profitability of the method for the operator.