Techno-Economic Analysis of WAMS Based Islanding Detection Algorithm for Microgrids with Minimal PMU in Smart Grid Environment

With large scale deployment of non dispatchable renewable energy sources, distributed generators (DGs) have paved way for multiple microgrids. Existing standards stipulate disconnection of DGs in the event of any fault in either utility or microgrid side. However, to ensure reliable power supply from these microgrids, the proposed scheme operates an islanding detection algorithm (IDA) employed at the microgrid control center (MGCC) that acts on the switch at the utility point of common coupling (PCC). The whole microgrid is switched over to an islanded mode on detection of an islanding event. Phasor measurement unit (PMU) data from the utility PCC enables accurate islanding detection using islanding detection monitoring factor (IDMF) and rate of change of inverse hyperbolic cosecant function of voltage (ROCIHCF) along with voltage at the PCC. Mathematical morphological filters are employed to detect any persistent short circuit fault in the microgrid side which may island the DG. For such faults, decision for disconnecting the DG or islanded operation is based on probability of power balance (PoB) and probability of islanding duration in the sub-microgrids, computed at the MGCC. Further, performance of the IDA in microgrids is assessed using a proposed, microgrid performance index (MGPI) considering the uncertainties in NDRES. Suitability of the proposed indices to predict events leading to islanded operation in real time is also validated using decision tree (DT) method. The discrimination capability between islanding and other transient events of DTs, yielded an accuracy of approximately 99.9% for minimum detection time, which proves the prowess of the method in real time scenario. Compared to existing methods, the proposed method promises reduced islanding detection time. Another distinct feature is that time for islanding detection remains same irrespective of power mismatch ratios. The proposed method prevents false alarms for critical non-islanding events with zero non detection zone. Utilizing the proposed method, any redundant DG outage can be avoided, minimizing their down time. An economic analysis of the proposed islanding detection method using wide area monitoring system (WAMS) with minimal PMU deployment has also been studied to reduce the cost of PMU installation without sacrificing the reliability benefit for the customers.