Thesis (M.Sc.) - Cairo University - Faculty Of Engineering - Department of Electrical Power and Machines

Medium voltage distribution networks in electric power systems are the link between high voltage transmission lines and low voltage networks and can be classified into overhead feeders and underground cables. Overhead MV distribution networks are characterized by presence of several laterals, short length feeders, single end measurements at HV/MV substation, and affected by the load variations in low voltage networks. So fault location in distribution systems has several challenges in order to mitigate outage time and fast supply restoration. A digital protection technique is implemented for MV overhead distribution network in this thesis not only to identify the faulted section but also to calculate the fault distance from the measuring point at MV substation. The proposed scheme is based on the voltage signals measured at the substation only. It uses fault passage indicators (FPI) at strategic points to identify the faulted section. Within the software technique, the algorithm converts the measured voltage signals from phase domain to modal domain using Clarke transformation, then frequency analysis is performed in order to determine the characteristic frequency of travelling wave generated due to fault and received at MV station clearly. This frequency is used to accurately determine the fault distance according to mathematical formulas represented by some developed equations during a training phase using curve fitting. Each equation is a relation between the estimated fault distances and their corresponding frequencies for a specific fault type and path. By getting the characteristic frequency using fast Fourier transform upon fault occurrence and substituting in the appropriate equation, the accurate fault location can be estimatede fault inception angle, fault resistance, and load variation.

Issued also as CD