Novel methodology for resilience assessment of critical infrastructure considering the interdependencies: A case study in water, transportation and electricity sector

Abstract

Critical Infrastructures (CI) are vital for societal and economic stability, yet their resilience against disasters remains inadequately understood with the increasing interdependencies among the CIs. A better understanding of these interdependencies and the dynamic nature of CI functionalities is crucial for advancing disaster resilience assessment within engineering systems. This paper introduces a novel approach using a Dynamic Bayesian Network (DBN) to assess resilience in interdependent CI systems. The DBN method enables a probabilistic evaluation of system resilience by incorporating interdependencies and capturing the temporal dynamics of system capacities. This approach offers a more detailed perspective on resilience by modelling system functionality using expected values of different functionality states over time. Using a case study in Sri Lankan electricity, water distribution, and road infrastructure sectors and 34 experts, this study examines the complex network of CIs. It demonstrates the applicability of the proposed methodology. P-values of the Chi-Square test performed between the variation of model-predicted resilience and expert assessments are significantly less than 0.05, confirming the model's validity. Additionally, this study explores the expansion of the methodology for resilience assessment under multiple hazards, emphasizing its real-world effectiveness. The findings highlight the efficacy of the proposed methodology and its potential to assist asset managers, owners, and decision-makers in informed resilience planning and optimization strategies. This comprehensive approach fills critical gaps in existing methodologies, offering a robust framework for assessing CI resilience in a dynamic and systematic nature.