DEFINING THE NETWORK RESOLUTION FOR THE RELIABILITY AND RESILIENCE ASSESSMENT OF CRITICAL INFRASTRUCTURE

Abstract eng:
The catastrophic events of the past years revealed the high vulnerability of our communities. In particular, disruptions of the critical infrastructure had significant impacts on the economic security, social welfare, public safety, and the overall well-being of the affected regions. An accurate assessment of the reliability and resilience of critical infrastructure is of utmost importance. However, the modeling of critical infrastructure is challenging because they typically extend over large areas and are constituted by a large number of elements. As a result, it is important to properly define the resolution of the networks considered in the analyses. Furthermore, it is important to model the changes in the infrastructure capacities and demands following a disrupting event as well as the multiple dependencies and interdependencies among different networks. After a disrupting event, networks may be not able to satisfy pre-event demands, and the post-event demands is challenging to model because of the uncertainties in the human behavior/response (e.g., evacuation or the decision to relocate.) At the same time, dependencies and interdependencies can be responsible of significant and typically unanticipated functionality losses even when a specific network does not experience direct physical damage. An accurate description of these aspects requires a proper definition of the network resolution. On one hand, a detailed representation of the networks is possible at the cost of a great amount of required input data (which are often scarce in the immediate aftermath of a damaging event) and large computational burden. On the other hand, a simplified or skeletonized representation of the networks can also be used with typically less detailed inputs requirements and computational costs, but at the possible expense of the accuracy and possible usefulness of the analyses. It is important to identify an optimal tradeoff between granularity and accuracy of the network resolution. This paper focuses on the definition of the optimal network resolution in the assessment of the reliability and resilience of critical infrastructure. A sensitivity analysis allows to assess: i) whether the resolution of simplified networks is sufficient for assessing the network performance at a community level; ii) up to what extent simplified models can capture post-event variations in capacity and demand; and iii) how dependencies and interdependencies can be integrated and modeled at different levels of resolution. This paper considers a potable water network, modeled with different levels of resolution, subject to seismic hazard. However, the principles and procedures are general and can be extended and applied to any critical infrastructure subject to natural or anthropogenic hazards.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket