Browse by author
Lookup NU author(s): Dr Manuel HerreraORCiD
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Assessing the reliability of critical infrastructure networks, such as urban systems essential for city functions like electricity and water, is essential for robust operation and risk management. Traditional methods for reliability estimation, such as minimal cut-sets and path enumeration, often become computationally infeasible for large-scale, complex networks due to the need to evaluate all possible node-to-node paths. This paper introduces a novel approach based on landmark nodes – critical nodes essential for maintaining network connectivity – to estimate reliability more efficiently. Instead of analysing all paths between nodes, the method focuses on those connecting regular nodes to landmark nodes, significantly reducing the number of paths considered and improving computational efficiency. The network is first decomposed using a graph clustering algorithm, producing internally dense subgraphs. Reliability is then evaluated through intra-subgraph and inter-subgraph paths. A bipartite network model is also employed to represent inter-cluster structure, accounting for failures in both nodes and links. This supports a multi-scale reliability analysis across local areas and the full network. The methodology is validated using benchmark power distribution networks to ensure reproducibility. To demonstrate practical relevance, it is also applied to a real-world case study involving the water distribution system of Pavia, Italy. This application highlights how key urban areas and components can be efficiently identified to prioritise maintenance and guide resource allocation, contributing to more resilient and sustainable infrastructure management.
Author(s): Herrera M, Giudicianni C, Sasidharan M, Wright R, Creaco E, Parlikad AK
Publication type: Article
Publication status: Published
Journal: Reliability Engineering & System Safety
Year: 2026
Volume: 265
Issue: Part A
Print publication date: 01/01/2026
Online publication date: 16/08/2025
Acceptance date: 06/08/2025
Date deposited: 18/08/2025
ISSN (print): 0951-8320
ISSN (electronic): 1879-0836
Publisher: Elsevier Ltd
URL: https://doi.org/10.1016/j.ress.2025.111563
DOI: 10.1016/j.ress.2025.111563
Data Access Statement: Data will be made available on request.
Altmetrics provided by Altmetric
