Assessing IRMA Impacts and Subsequent Service Restoration to Reveal the Interdependencies and the Resilience of Saint-Martin’s Critical Infrastructures.

Abstract:

Following IRMA, a Category 5 hurricane on the Saffir-Simpson scale, the French Saint-Martin Island was disconnected from the world, isolated, without electricity, water or any means of communication. In response to this situation, an evaluation of Critical Infrastructure (CI) networks failures must be performed to identify vulnerabilities, interdependencies, recovery, and ultimately the resilience of these networks. Nonetheless, the term resilience has been contested in its definition for over a decade now. Consequently, the numerous definitions of the term underlie the heterogeneity and nonuniformity of the methods used for measuring resilience. In an effort to bridge this gap, the presented study proposes a method to implement, in an operational way, the concept of resilience by analyzing the disruption and the return to service of Saint-Martin Island’s CI: electrical, drinking water, sewage, telecommunications and transportation (roads, airports and seaports) networks following IRMA. CI data was collected from various sources: interviews, geospatial platforms, press releases, community reports, published and grey literature, etc... The impacts of IRMA on the CI functioning were analyzed by studying CI robustness, failures (power outs, cut offs) and rapidity of service resumption. Following this step, cascading failures and interdependencies were determined. To assess CI recovery, spatial and temporal tracking of CI restoration was performed to determine the time taken for each component to return to its normal state. Key findings revealed resilience curves that pinpoint the vulnerabilities and interdependencies underlying the observed network failures and thus offered valuable inputs for the construction of decision support models for long-term, multi-risk “build back better” projects of Saint-Martin’s CI. Moreover, the comparison of these curves highlighted the networks that were the least resilient during IRMA, and subsequently provided insights that can be used as a support for prioritization during decision making to improve CI resilience. The proposed methodology offered a robust tool for measuring operational resilience of CI based on actual on-site observations, clarifying some assumptions often encountered in predictive resilience measurement methods.