Semi-Autonomous Facility Emergency Response (SAFER) for Energy-Water Resilience

The focal area of this whitepaper is enhancing cyber security for integrated energy-water systems. The near-term opportunity is to provide a decision support system (controller) to assist the hydroelectric plant operators during communications outages or during a cyber event (cyber-attack or relay mis-operation) where communications and control commands are spoofed.

The proposed research is to develop a digital twin (DT) of the electric grid as observed from a hydroelectric plant's hard-wired measurements of the power lines connecting it to the power system (i.e., observations taken at the hydroelectric facility switchyard's main power transformer high side interconnect). The research start point is to use this point of measurement as a "shadow model" of the power grid. The objective is to detect anomalies between control commands the hydro facility receives and commands the hydro plant could expect to receive given the observed state of the grid. The purpose is to alert the facility operator if control commands or communications for the hydro plant have been spoofed (cybersecurity defense). Another practical use of this measurement-based DT is to be able to dispatch the hydro facility to respond to grid status during a SCADA failure (resilience during communications outage) based on possible cyber-events such as cyber-attacks or relay mis-operations (human/setting errors).

Project result will be proof of concept and a SAFER controller that will include a DT of a utility system as observed from one hydroelectric facility that has black start capability. The developed local controller will be successful if it can react quicker than human power plant operators, better estimate system conditions (islanded or intact, f and v regulation capabilities, etc.), and/or anticipate disruptive actions or deteriorating conditions, demonstrating its ability to improve the speed and efficacy of control actions during widespread or total communications outages. Another expected outcome will be a successful test, involving a simulated cyber intrusion where the SAFER controller will detect attempted dispatch of improper commands directed to the hydropower generator, raise an alarm, and formulate control commands appropriate for the status of the utility system in its current operating state. The ultimate goal is for the SAFER local controller design strategy to be generalized so it can be applied to other power generating facilities and major nodes, relays, switches, tap changers, etc. Future applications of a direct-measurement-based DT could include management of non-powered dams, navigable waterways, water draws for irrigation and potable water, and municipal water distribution systems.