Topological Performance Measures as Surrogates for Physical Flow Models for Risk and Vulnerability Analysis for Electric Power Systems

Abstract

Critical infrastructure systems must be both robust and resilient in order to ensure the functioning of society. To improve the performance of such systems, we often use risk and vulnerability analysis to find and address system weaknesses. A critical component of such analyses is the ability to accurately determine the negative consequences of various types of failures in the system. Numerous mathematical and simulation models exist that can be used to this end. However, there are relatively few studies comparing the implications of using different modeling approaches in the context of comprehensive risk analysis of critical infrastructures. In this article, we suggest a classification of these models, which span from simple topologically-oriented models to advanced physical-flow-based models. Here, we focus on electric power systems and present a study aimed at understanding the tradeoffs between simplicity and fidelity in models used in the context of risk analysis. Specifically, the purpose of this article is to compare performance estimates achieved with a spectrum of approaches typically used for risk and vulnerability analysis of electric power systems and evaluate if more simplified topological measures can be combined using statistical methods to be used as a surrogate for physical flow models. The results of our work provide guidance as to appropriate models or combinations of models to use when analyzing large-scale critical infrastructure systems, where simulation times quickly become insurmountable when using more advanced models, severely limiting the extent of analyses that can be performed.

Review and Evaluation of the J100‐10 Risk and Resilience Management Standard for Water and Wastewater Systems

“…However, there is debate in the risk analysis literature regarding whether assigning static probabilities is even feasible. One group (see Aven & Guikema, ; Aven & Renn, ; Brown & Cox, ; Cox, ) argues that the intelligent nature of the adversary makes assigning meaningful and useful probabilities problematic if not impossible. Bayesian probabilities of attack can be elicited through experts, but are misleading because the defender and attacker act on different knowledge bases.…”

“…Articles by Alderson et al (2015b) and Alderson, Brown, Carlyle, and Anthony Cox (2013) emphasize the use of physical infrastructure models rather than simple topological representations to provide the most accurate reflections of network performance. Larocca, Johansson, Hassel, and Guikema (2015) compared a range of topological metrics and physical models to measure power system performance, and found that combining graph theory with physical flow models provided the most accurate insights.…”

Complexity and Vulnerability Analysis of Critical Infrastructures: A Methodological Approach

“…Automatic fare collection system (12) Screen doors (13) Escalator (14) Environment control system (15) Water supply and drainage system (16) Electrical and mechanical control system (17)…”

“…With the increasing of complexity and hazards, there is a need to develop new perspective or approach for infrastructure safety. In the context of the existing literature, some attention is already paid to vulnerability of infrastructure system [11][12][13]. Vulnerability analysis is applied within the management of CIs and some achievements have been made by exploring the impacts of random disturbances, deliberate attacks, and natural disasters.…”

Comparing Network-Centric and Power Flow Models for the Optimal Allocation of Link Capacities in a Cascade-Resilient Power Transmission Network

“…Some studies [23], [25], [27] have provided qualitative comparisons between complex network theory models and power flow models -identifying similarities and differences, and evaluating advantages and disadvantages. Most recently, Correa and Yusta conclude on the appropriateness of graph theory techniques for the assessment of electric network vulnerability by comparison to physical power flow models [28].…”