A Sampling Approach for Intentional Islanding of Distributed Generation

This paper proposes an innovative way of making a decision to island (or synchronize) distributed generators when operating conditions of a distribution system are varied. Islanding decisions are made when a network undergoes disturbances. In this approach, there is a decision-making mechanism, which samples states of the network operating condition to either synchronize or island in an identified period of time. Decision making is achieved through extended sequential sampling, which manages generator states within specific time periods. The approach can minimize the duration of islanded operation and manage generators' re-connection to the grid. Use of the proposed mechanism and sampling approach offer benefits over conventional relays; incorrect operations due to errors in measurements are reduced because the operational actions are not based on a single decision sample, but instead rely on extended sequential sampling over a number of samples. A case study is investigated using the IEEE 30-bus test system and the results are presented for this new approach