Detection of high impedance faults in M.V. mesh distribution network

This paper examines the problem of locating single-line-to-ground faults through resistance that may occur anywhere in an existing multi-ring electrical distribution network that belongs to a Kuwaiti oil fields, in order to improve reliability and to reduce the shutdown time of this network. The developed method is based on using Artificial Neural Network techniques as the algorithm to save time in fault location, as well as estimating the value of the fault resistance. This technique is applied to an existing 13.8-kilovolt-distribution network, which caters to an oil production field spread over an area of approximately sixty km2. The nature of this distribution system, which consists of multiring configuration, illustrates the complexity of the network feasibility and the effectiveness of the suggested fault location method. Simulation results for several study cases have shown the feasibility and the effectiveness of the suggested fault location method.