A Novel Algorithm to Mitigate Protection Challenges in a Distribution System Integrated with Inverter-Based Distributed Energy Resources

Abstract

To overcome the challenges on global warming due to fossil-fuels based generation, renewable distributed energy resources (DERs) like inverter based DERs (IBDERs) have been significantly integrated into distribution systems. However, during a short-circuit fault, the fault current contribution from IBDER is very low due to strong control of the inverters. The low fault current creates sensitivity issues in the overcurrent relay of IBDER which can create protection failure. To overcome this issue, a new way of implementing machine learning based algorithm named Radial Basis Function Neural Network (RBFNN) will be proposed. This method will use the time series data to detect fault current contribution from IBDER fast and accurately. In a distribution system, there could be a recloser on the feeder between a feeder breaker and IBDER. An ideal scenario is the feeder breaker and recloser to trip for any faults between them. In this case, the overcurrent relay of IBDER should not operate to avoid any unnecessary outages to the customers between the recloser and IBDER. However, if RBFNN algorithm is implemented in the overcurrent relay of IBDER then it will trip for all faults on the feeder along with feeder breaker due to its fast operation. To avoid such operation, this paper is proposing the RBFNN algorithm for both recloser relay and IBDER relay which will trip the recloser relay instead of the IBDER relay for any faults between the feeder breaker and recloser. However, the RBFNN algorithm of recloser relay will be blocked for any faults between the recloser and IBDER. Simulations have been performed on a distribution system with feeder breaker, recloser and IBDER for various fault scenarios to prove the benefits of this algorithm. This paper also shows the coordination of RBFNN algorithms between the recloser and IBDER for faults between feeder breaker and recloser to avoid any miscoordination.