Large‐scale systems resilience: A survey and unifying framework

Abstract

A unifying framework in defining and measuring resilience has been an intense research topic. In this paper, resilience is measured as a function of intrinsic capacities of a system, the effectiveness of recovery, and the extrinsic random shock process. Some existing resilience measures are analyzed as special cases of the proposed unifying measure. Then, we develop a framework in which the key constituents in achieving resilience are identified. Resilience is represented by four key dimensions: reliability, robustness, recovery, and reconfigurability. Finally, some practical and specific strategies are proposed to enhance the resilience of critical infrastructure systems under the proposed framework.

Quantitative Assessment of System Response during Disruptions: An Application to Water Distribution Systems

“…1. System response is determined by different variables, including the external disruption process and the intrinsic capabilities of the system (Shen et al 2020). These include absorptive capability (i.e., the ability to minimize the impacts of disruptions), adaptive capability (i.e., the ability to self-organize for recovery of performance), and recovery capability (i.e., the ability of a system to be repaired) (Vugrin et al 2011).…”

“…However, if a disruption occurs, then part of the desired demand might remain unsatisfied not only at the disrupted node, but also at the nodes connected to it, which might also suffer pressure losses to compensate for the additional flow. Therefore, the timedependent demand served d i is a direct consequence of the intrinsic capacities of the WDS in terms of pressure built up and damage sustained (pressure loss) (Shen et al 2020). Accordingly, here the average satisfied demand is used as a proxy for the demand delivery service to monitor the overall system performance when node i is disrupted:…”

“…Brentan et al (2018), for example, observed slow pressure losses for a WDS subject to nodal leaks at certain locations. Tarani et al (2019) found that the loss of performance, measured in terms of demand deficiency over time, was delayed with respect to the start of a flood wave, and that it was quickly recovered following system restoration. Similarly, Cimellaro et al (2015) assumed a two-step response for a WDS subject to pipe bursts, characterized by a delayed loss of performance, i.e., number of households with water, and a sudden recovery following complete restoration.…”

“…Overall, empirical evidence from previous studies show that WDS behavior is determined by many factors, such as the type of disruption and the monitored measure of performance. Currently, these behaviors have been simply evaluated using empirical data (Shen et al 2020), and a model for their characterization, which could provide further insight for resilience assessment and enhancement, is still lacking.…”

The Effect of Time-Varying Value on Infrastructure Resilience Assessments

“…As with performance, time thresholds describe an (often implicit) non-linear transition in stakeholder value. Examples include acceptable recovery time [66], latency limit [60], [61], maximum acceptable disruption time [14], maximum acceptable recovery time [59], [67], maximum allowed recovery time [68], and maximum tolerable period of disruption [69], [70]. Time thresholds can vary across scenarios and stakeholders.…”