Dynamic probabilistic risk assessment for cyber security risk analysis of the electric grid

Electric power systems are a major part of critical infrastructures. Rapid addition of digital systems in the electric power sector has increased opportunities for malicious entities to cause serious consequences. Therefore, it is necessary to understand the risk profile of the overall system such that decision makers have appropriate information to make judicious investments in defensive measures. Current practices rely heavily on subject-matter experts. In this research, we propose a dynamic probabilistic risk assessment (DPRA) framework for the risk analysis of the electric grid. The results of the analysis can be used to support decision-making in security resource allocation. DPRA provides a probabilistic tool to consider the uncertainties in system evolution. In this research, we focus on uncertainties related to the attacker's and the operator's behaviors, as well as system failure or degradation. The large number of system evolution trajectories resulting from the above uncertainties may lead to differing consequences, and therefore the risk metrics of interest can be quantified. We demonstrate the proposed DPRA framework on a modified version of the IEEE 14-bus system, which exhibits behavior representative of a part of a North American electric power transmission system. The potential cyber-attacks on the system and the uncertainties in system evolution are described. The results, including the representative system trajectories and the risk estimations, are presented.