Fault detection and location algorithm for DG‐integrated distribution systems

Abstract

The overcurrent protection on the traditional distribution system uses local measurements and settings as a basis for decision-making. However, connection of multiple distributed generators with different short-circuit characteristics makes the local measurements unpredictable, leading to possible loss of protection coordination. In order to solve this problem, this study proposes an algorithm that achieves its protection function through load flow and sensitivity calculations based on voltage measurements. The algorithm is built on a modelling approach that splits the distribution network into a series of overlapping protection zones. The overlapping feature is realised logically through peer-to-peer communications between intelligent agents located in adjacent zones. A sensitivity-based electrical signature was identified that is generated when a fault occurs within a zone. Simulation results show that the proposed protection algorithm is able to identify the zone in which the fault has occurred and the specific faulted node or line section in that zone. The fault current contributions, or no contributions, of the distributed generators (DGs) do not impact the operation of this algorithm.