b'Risk and Resilience AssessmentAdvancing the scientific understanding needed to develop a methodology of Cyberattacks on Electricfor evaluating risk in high-consequence cyber-physical systems that make up Grids: Informing Riskcritical infrastructure.Characterization Using DynamicU nderstanding risk in high-consequence cyber-physical systems required characterizing the complex interactions between physical systems and Probabilistic Risk Assessment general-purpose computing components that are distributed across those systems. Current approaches rely on subjective assessment based on subject matter expert input and do not have a firm scientific basis. There are mature and accepted methods for understanding risk in physical systems that require a thorough understanding of the components in the system, how those components interact to perform their intended functions, and the likelihood of failure of the components. TOTAL APPROVED AMOUNT:Cyber-physical systems have another layer of complexity because they can aggregate $1,560,000 over 3 years previously unknown interactions between components in the physical system through intentional or unintentional misuse of the general computing capability of PROJECT NUMBER:the cyber technologies embedded in the system. When the threat is an intelligent 19P45-020 adversary, whose goal is to compromise or sabotage the system, the complexity PRINCIPAL INVESTIGATOR:of determining what can happen and how likely it is becomes ever more difficult. Katya Le Blanc The horizontal application of the same or similar cyber technologies creates a scale in attack surface complexity that can overwhelm the present high-fidelity n-k CO-INVESTIGATORS: physical failure models with orders of magnitude greater numbers and modalities. Craig Rieger, INL More computational power is necessary but not sufficient to provide the solutions Thomas Ulrich, INL that will be needed by investment decision makers. A scientific basis for decision Timothy McJunkin, INL tools and sensitivity analysis is crucial to guide the development and application of Brian Johnson, University of Idaho higher-fidelity physical effects models to the most consequential assets and to guide Carol Smidts, The Ohio State University critical choices with uncertainty in data-limited scenarios. This project addressed these critical gaps by developing a theoretical framework for characterizing risk and uncertainty, collecting and generating data using modeling and simulation techniques, and developing repeatable methodology that can scale in complexity and application for risk quantification.Dynamic probabilistic risk assessment framework.122'