Post-earthquake recovery of a community and its electrical power supply system

Abstract

Research from the past ten years shows that the seismic fragilities of different infrastructure networks are better understood and estimated than the recovery of a community subjected to these fragilities. Modeling the recovery process for a specific infrastructure system-community-hazard combination is a challenging task due to the lack of data and because the recovery process is subject to a large set of interdependent factors and variables. The primary objective of the present work is to understand how typical modeling assumptions affect the estimated seismic resilience of a community, whose electrical power supply network has been affected by an earthquake. The recovery model relies on component recovery probability functions conditioned on the initial post-disaster damage state of the single components. The loss of resilience (LOR) of a set of systems, composed by an electric power supply system and a community demanding electric power is assessed using a compositional approach to quantify the loss of resilience of the combined system. The modeling assumptions that are considered are: a) the time-varying recovery function, and b) the probabilistic model for recovery. For the recovery function, the adopted probability model of the infrastructure system is found to have a strong influence on the loss of resilience, while it is relatively insensitive to that of the community. LOR is also observed to have a significantly higher sensitivity towards the mean of the used recovery functions than towards their variance.