Electric grid modernization efforts are directed toward enabling a resilient, secure, sustainable, and reliable power system. Modern real-time, data-driven tools support wide-area monitoring and provide decision support to control room operators, toward the goal of grid resiliency. West Virginia University’s Smart Grid REsiliency and Analytics Lab (SG-REAL) models cyber-power systems using a Real-Time Digital Simulator (RTDS) and a real-hardware-based cyber network. The Schweitzer Engineering Laboratories (SEL) Software Defined Network (SDN) Switches are used to network the SEL Relays and SEL Real-Time Automation Controller (RTAC) at the substation level of the SG-REAL testbed. A wide area network (WAN) has been emulated via Network Simulator-3 (NS3) with hardware and software-based SDN devices for hierarchical substation-level networking. A detailed model of the Substation automation system (SAS) is developed using RTDS GTNETx2 GOOSE/SV modules and standalone hardware as per IEC 61850. The physical power system, modeled after a synthetic Washington state transmission network, provided by PowerSimulator™, is developed in the RTDS with hardware-in-the-loop (HIL). The network consists of 40 buses and 61 transmission lines and is modeled over 3 RTDS cores. The integrated cyber-power setup is used to (1) perform ethical cyber-attacks with HIL and (2) generate a variety of cyber and power data, enabling the development and validation of tools for both power system anomaly detection and cyber resiliency and security. The data from the testbed is utilized to develop advanced Machine Learning (ML) applications for grid resiliency. These include various supervised, unsupervised, and physics-informed ML approaches. Further, we can distinguish unsolicited cyber-attacks from normal faults, events, and scheduled maintenance operations using advanced ML tools. Overall, the SG-REAL testbed allows us to develop and validate software tools, analyze advanced attacks, and devise appropriate mitigation scenarios to enhance situational awareness and improve grid resiliency.
Vasavi Sivaramakrishnan, West Virginia University