RT22 | MG Simulation CHIL via FPGA IEC 61850 GOOSE Subject to Cyber-Attacks

This work develops Hardware-in-the-loop (HIL) simulation against cyber attacks. The test setup in Opal-EXata cyber attack example ( paper), is based on the software-in-the-loop (CSIL) MGC, however, the hardware-in-the-loop HIL testing is left for the developer. Thus, we design a light-weight intelligent electronic device (IED) that performs MGC, interfaces are developed based on IEC 61850 GOOSE protocol from/to the real-time simulation and the MGC. They are executed on two equipment stages, FPGA and BeagleBoneBlack. The concept behind using these boards is based on their low cost, flexibility, support for various interfaces/protocols, I/O pins, and ease of configuration. CSIL versus CHIL tests are used to evaluate the MG behavior against different cyber attacks. We also evaluate the MGC designed control function in accordance with IEC 61850 GOOSE protocol. Two scenarios was carried out, In the first scenario, the MG will be islanded in second 1, in this case the MGC will check the power balance and implement the power balance operation emergency condition (the difference not exceeds 3MW). If the check emergency condition becomes true, the MGC attempts to immediately disconnect the sheddable Load 4 to maintain the MG stability. Be that as a delay attack is introduced to the GOOSE trip command packets sent from the MGC to Load 4, the load shedding function may fail to operate in the required timeframe. In this case this will cause severe unbalance between generation/load relationship and oscillations on MG nominal operation parameters such as e.g., frequency, voltages etc. Through the C code available delay function within the MGC initial code, a one-second delay attack is applied to the MGC GOOSE message command. MGC based on its normal operation will receive measurements that is sent from each MG assists via GOOSE. Be that as a Man-in-the-middle attack is introduced to the load measurements data. Before the load measurements data being received by MGC. The data is manipulated in the middle of its way to the MGC. In this case, the MGC may take incorrect actions based on these received non-critical measurements. According to the test scenario 2, the active power measurement from Load 2 is duplicated by applying a packet manipulation attack to the GOOSE message. In this case, the MGC will take the incorrect action (false tripping) because it perceives the controlling operation emergency condition is true (total load will be more than 3MW greater than the total generation). As a result, a trip command is sent to disconnect Load 3. . It also causes oscillations on MG nominal operation parameters such as e.g., frequency, voltages etc. Comparison between both achieving testing results, SIL and HIL will be made within the rest of this work. The results are The development and performance of an MGC against cyber attack control schemes have been implemented in this paper. These are done by design and deployed on a light-weighted intelligent IED. The MGC control solution and its relevant communication system have been designed in compliance with the IEC 61850 and executed on two equipment stages, FPGA and BeagleBoneBlack. CSIL versus CHIL tests are used to evaluate/assess the MG behavior against different cyber attack scenarios. Moreover, we also evaluated IEC 61850 GOOSE protocol implementation, processing and finally control action performance. The obtained results demonstrate that the light-weight MGC approach and data modeling of various IEC 61850 predefined data object LNs are correct for the design of the power balance control/protection function against cyber attack. In addition, they demonstrate the successful implementations of the designed control/protection function and the modeled MGC LNs in various cyber-attack case studies on reliable detection of the emergency condition. Further work on the analysis of the data received by MGC, implementation of different cyber attacks and power balance detection algorithms is needed