Advanced Grid Operations: The Real-Time Simulation of Power System Faults to Innovate Protection Methodology

This research answers the question, “With the current knowledge of power systems engineering, would protection problems be solved the same way?” Today’s protection methodologies follow a very narrow approach that relies on overcurrent detection, and these are traditionally used in power systems with radial power flow. In other words, when a protection scheme senses a higher than intended current, it limits or disables current flow; however, when this type of problem is approached with today’s innovative problem solving solutions, a lens is provided to see issues differently. In Advanced Grid Operations research, a Real-Time Digital Simulator is used to simulate, in real-time, several different types of faults on different types of power systems to create a repository of fault data used to train a Machine Learning algorithm. Machine Learning Protective Relaying aims to ultimately provide additional information that can be leveraged to solve protection problems regarding traditional protective relays, renewable energy, and other roadblocks.