A novel wavelet-based feature extraction from common mode currents for fault location in a residential DC microgrid

DC community and residential microgrids are recognized as effective means for electrification of remote areas as a result of monumental efforts in many parts of the world - most significantly in India - but also through similar efforts in Nepal, Cameroon, New Guinea and Nigeria. So far modular approaches have been developed that enable construction of scalable microgrids based on PV and battery storage. However, as these systems proliferate, it will be necessary to develop safe and reliable methods for fault protection. Ground faults are of specific concern because, in many cases, cables are buried underground. At the same time, microgrids include current monitoring and processing capability wherever an energy resource interfaces to the microgrid through a power electronic converter. This paper discusses methods for identifying ground fault behavior within standard DC microgrid structures and proposes methodologies for extracting specific information about the location and type of fault using wavelets. The sensing hardware, sampling rates and processing requirements that are needed are also presented.