Robust Optimal Power Flow under Industrial Cyberattacks and Renewable Uncertainty
The digitalization of electric-power operation couples optimal power flow (OPF) with measurement systems, communication networks and increasingly variable renewable generation. Power system operators may have to make dispatch decisions using load measurements that have been altered by false-data injection attacks. At the same time, actual renewable power generation may differ from what was forecast. When these two problems occur together, the system may need more balancing capacity than would be expected if each uncertainty were considered separately. This paper examines both effects through a scenario-based robust direct-current optimal power flow (DC-OPF) approach. The proposed model determines generator schedules and upward and downward reserves with ensuring that the system can respond to bounded combinations of corrupted load measurements and renewable forecast errors. The approach focuses on reducing their operational consequences. It assumes that the operator does not know the exact amount of corruption measurement but can define reasonable limits for measurement errors and renewable uncertainty. The study uses the publicly available PGLib-OPF v23.07 IEEE 14-bus and IEEE 30-bus networks, with additional renewable generation and cyberattack scenarios introduced for testing. The results show that the robust approach increases preventive operating costs by 0.98% for the 14-bus system and 1.97% for the 30-bus system compared with a deterministic OPF using a standard 10% balancing-reserve rule. In 3,000 out-of-sample scenarios for each network, the robust schedule avoided emergency load shedding within the defined uncertainty limits. By comparison, the deterministic approach required load shedding in 62.77% of the IEEE 30-bus scenarios, with an average of 2.56 MW and a maximum of 18.93 MW of unserved power. These results suggest that the benefit of robust OPF is not simply a change in how generation is scheduled. More importantly, it helps place reserve capacity where it can be used effectively, reducing the risk that corrupt measurements and unexpected renewable fluctuations will prevent the power system from operating within its physical limits.