Date of Award
Open Access Dissertation
Doctor of Philosophy in Electrical Engineering (PhD)
Administrative Home Department
Department of Electrical and Computer Engineering
Committee Member 1
Committee Member 2
Committee Member 3
Insurance businesses for the cyberworld are an evolving opportunity. However, a quantitative model in today's security technologies may not be established. Besides, a generalized methodology to assess the systematic risks remains underdeveloped. There has been a technical challenge to capture intrusion risks of the cyber-physical system, including estimating the impact of the potential cascaded events initiated by the hacker's malicious actions.
This dissertation attempts to integrate both modeling aspects: 1) steady-state probabilities for the Internet protocol-based substation switching attack events based on hypothetical cyberattacks, 2) potential electricity losses. The phenomenon of sequential attacks can be characterized using a time-domain simulation that exhibits dynamic cascaded events. Such substation attack simulation studies can establish an actuarial framework for grid operation.
The novelty is three-fold. First, the development to extend features of steady-state probabilities is established based on 1) modified password models, 2) new models on digital relays with two-step authentications, and 3) honeypot models. A generalized stochastic Petri net is leveraged to formulate the detailed statuses and transitions of components embedded in a Cyber-net. Then, extensive modeling of steady-state probabilities is qualitatively performed. Methodologies on how transition probabilities and rates are extracted from network components and actuarial applications are summarized and discussed.
Second, dynamic models requisite for switching attacks against multiple substations or digital relays deployed in substations are formulated. Imperative protection and control models to represent substation attacks are clarified with realistic model parameters. Specifically, wide-area protections, i.e., special protection systems (SPSs), are elaborated, asserting that event-driven SPSs may be skipped for this type of case study.
Third, the substation attack replay using a proven commercially available time-domain simulation tool is validated in IEEE system models to study attack combinations' critical paths. As the time-domain simulation requires a higher computational cost than power flow-based steady-state simulation, a balance of both methods is established without missing the critical dynamic behavior. The direct impact of substation attacks, i.e., electricity losses, is compared between steady-state and dynamic analyses. Steady-state analysis results are prone to be pessimistic for a smaller number of compromised substations.
Finally, simulation findings based on the risk-based metrics and technical implementation are extensively discussed with future work.
Yamashita, Koji, "ESTABLISHMENT OF CYBER-PHYSICAL CORRELATION AND VERIFICATION BASED ON ATTACK SCENARIOS IN POWER SUBSTATIONS", Open Access Dissertation, Michigan Technological University, 2020.