Cyber-Physical Resilient Distribution System (CYPRESS)

Advances in smart grid technology have resulted in a higher reliability of service to consumers. Although faults and outages are inevitable, service must be restored to normal in the event of a blackout. Events such as hurricane, earth-quakes and floods are high impact low probability (HILP) events that can cause wide spread blackouts. However, deployments of Microgrids (MGs) and distributed energy resources (DERs) at the distribution level provide opportunities for service restoration in islanded mode. Bulk power distribution networks have limited number of switchable lines and loads. Hence, the network can often be sectioned into large bus blocks. Switching multiple loads simultaneously can result in severe frequency and voltage oscillations especially in inverter dominated weak inertial systems. Therefore, limits on frequency and transient voltage of DERs shall be incorporated as constraints of the restoration problem. 

This work is an effort to validate a multi-layer sequential service restoration framework on an experimental setup at AIT Energy Systems Lab. An IEEE-123 node test feeder was emulated in OPAL-RT 5600/5700 with a controller in the loop (CIL) setup under the ERIGrid Transnational Access (TA) program. eMegaSim (real-time electromechanical solver for power systems) was used to emulate the test feeder at Ts = 100uSec. Moreover, multi-agent system (MAS) based on Rpi controller was used to implement distributed secondary control for frequency and voltage regulation during the restoration process.

During the online (17 days) and physical (17 days) access period several tests were conducted to validate the framework for different scenarios. 

• Grid-following inverters (GFL) with and without grid-support functions as defined in IEEE 1547:2018 for Freq/Watt and Volt/Var regulation. 
• Centralized automatic gain control (AGC) for frequency and voltage regulation. 
• Role of distributed secondary control in frequency and voltage regulation during the black-start process. 
• Consensus control law in the distributed controllers. 
• Random location of faults and MG formation.