Quantification of Lifeline System Interdependencies after the 27 February 2010 M_w 8.8 Offshore Maule, Chile, Earthquake

Abstract

Data on lifeline system service restoration is seldom exploited for the calibration of performance prediction models or for response comparisons across systems and events. This study explores utility restoration curves after the 2010 Chilean earthquake through a time series method to quantify coupling strengths across lifeline systems. When consistent with field information, cross-correlations from restoration curves without significant lag times quantify operational interdependence, whereas those with significant lags reveal logistical interdependence. Synthesized coupling strengths are also proposed to incorporate cross-correlations and lag times at once. In the Chilean earthquake, coupling across fixed and mobile phones was the strongest per region followed by coupling within and across telecommunication and power systems in adjacent regions. Unapparent couplings were also revealed among telecommunication and power systems with water networks. The proposed methodology can steer new protocols for post-disaster data collection, including anecdotal information to evaluate causality, and inform infrastructure interdependence effect prediction models.

Probabilistic study of cascading failures in complex interdependent lifeline systems

“…However, the EA approach is in general unable to provide insights about the mechanisms that yield the measured responses. Instead, empirical studies tend to inform the calibration of computational or physical models [48].…”

Percolation in multiplex networks with overlap

“…Each layer corresponds to a network with a specific function, but the entire complex system requires multiple layers operating in a coupled way and forming an interacting set of networks. For example, analyses of the disruptions provoked by an earthquake to large infrastructures show the relevance of interdependence among power transmission and telecommunications [16] and the different resilience of coupled networks such as the power grid and the water system [17].…”

Post-disaster infrastructure restoration: A comparison of events for future planning

“…Such functions tend to suit those datasets where recovery is modeled back to the initial pre-disruption conditions due to observed datasets commonly showing restoration rates slowing over time [20,56,80]. However, as discussed, the proposed methodology seeks to only address the response phase meaning that functions that asymptotically tail-off are undesired.…”