This technology introduces a novel influence graph-based vulnerability assessment framework for Integrated Power and Gas Systems (IPGS), designed to identify critical components that initiate cascading failures.
The method first models cascading failures using fault chain theory, where sequential component outages are characterized by transition probabilities derived from overload conditions. Unlike conventional approaches that treat electric and gas networks separately, this framework develops a unified fault chain model for the entire IPGS. Both electric power flow (AC power flow) and dynamic gas flow models are integrated to simulate realistic interdependencies between subsystems.
From the generated fault chains, an influence graph is constructed. In this directed and weighted graph:
To identify the most critical branches, the framework applies eigenvector centrality, which evaluates not only how often a component fails, but also how influential its neighboring components are. This allows detection of components that may not fail frequently, yet trigger severe system-wide impacts when they do.
Validation on a 39-bus, 29-node IPGS model demonstrates that the method accurately isolates high-impact branches whose simultaneous failure leads to blackout conditions—outperforming traditional centrality-based approaches.