Characterizing multi-event disaster resilience
Introduction
The potential for significant loss of life resulting from improper preparedness and/or response to sudden-onset disasters calls for effective quantification of system resilience. Such resilience can be considered both at a local level and on a larger scale, and in particular, the resilience of systems such as key supply chain networks is critical to the sustainability of society. These networks provide vital human services including water and medical supplies, energy resources (i.e. electric power, oil, gas, fuel, etc…), transportation capabilities, information and communication technologies, and currency exchange functionality [1], [2]. The ability to quantify their resilience is instrumental in improving both emergency planning and response by being able to compare various configurations and to assess their relative returns and risks for effective emergency management [3].
The focus of this paper is on quantifying system resilience to sudden-onset disasters that are followed by multiple related sub-disasters. Such situations include (1) natural disasters such as earthquakes and their aftershocks, along with related events such as delayed building collapses; and hurricanes with levee failures and the associated disruptions in transportation networks and power grids [4], [5], (2) human-made disasters such as terrorist attacks [6], [7] and information security attacks [8] on stock exchanges, which could lead to serious price fluctuations and result in panic and financial meltdown, and (3) cascading network effects such as supply chain failures [9] associated with sudden price increases in essential materials such as wheat and rice, thus increasing hunger and causing serious economic disruptions. Situations such as these are very complex and they typically involve the need to balance multiple criteria in order to take effective actions to increase the systems' resiliency.
Godchalk [10] defined preparedness to sudden-onset disasters as the set of “actions taken in advance of an emergency to develop operational capabilities and to facilitate response in the event that an emergency occurs” (p. 136). A system’s operational risk does not only increase with the threat level, but also with the degree of vulnerability of a system [11]; the definition of resilience therefore should not only incorporate post-event consequences, but also pre-event preparedness and strategic planning. Resilience engineering calls for proactively looking for “ways to enhance the ability of organizations to explicitly monitor risks and to make appropriate tradeoffs between required safety levels and production and economic pressures" [12]. In resilience engineering, success is defined by “the ability of the system to monitor changing risk profile and take timely action to prevent the likelihood of damage”, while failure is the “absence of this ability" [12].
The approach taken in this paper for quantifying multi-event resilience is based upon the concept of the disaster resilience triangle [13]. Originally introduced by Bruneau et al. [13], the underlying construct of the disaster resilience triangle (illustrated in Fig. 1) was extended by Zobel in Refs. [3], [14] to provide an approach for calculating a multi-dimensional measure of predicted disaster resilience. This work also provided a means for visually comparing the tradeoffs between the multiple criteria from which the resilience measure was constructed [14], and an approach for representing decision makers' preferences between these criteria [3].
Both Refs. [3] and [14], however, only addressed predicted resilience for a single disaster event, and the models are not sufficient to assess the resilience of a system affected by multiple related events. If a system has not had a chance to recover fully by the time the next related sub-event occurs, then the characteristic shape of the disaster curve will tend to look more like Fig. 2 than like Fig. 1. In order to gain the advantages of computing an overall resilience to multiple related events, therefore, and thus to compare the predicted resilience of different systems, this paper extends the original concept of predicted disaster resilience to fit this new paradigm.
The rest of the paper is organized as follows. Section 2 presents an overview of prior work on disaster resilience, and Section 3 provides a re-interpretation of predicted resilience using the partial resilience associated with each of a number of sub-events. Section 4 suggests an approach for more clearly characterizing the component resilience values of these sub-events, and 5 Illustrative example, 6 Discussion provide an illustrative example of the overall technique and a discussion of the results. The paper concludes in Section 7 with a brief review of the contributions of the technique in support of effective decision making for multi-criteria emergency management.
Section snippets
Background
Methods to mitigate the consequences of random system failures resulting from sudden-onset disruptions have been widely investigated, and many such studies [15], [16], [17] focus on characteristics and costs specific to individual components of the systems. In contrast, the systems-based literature tends to propose vulnerability mitigation strategies for systems as a whole, rather than just individual links. For example, Moghtaderi-Zadeh and Der Kiureghian [18] prioritized investments in
Re-interpreting predicted resilience
As discussed above, the concept of predicted resilience [14] provides a means for analytically representing the relationship between the initial impact of, and the subsequent recovery time from, a sudden onset disaster event. If we let X represent the percentage of functionality lost, and we let T represent the time needed to recover “normal” operations, then the area of the triangle in Fig. 3 can be interpreted to represent the overall amount of time-varying loss suffered by the system due to
Characterizing partial resilience
Given that the derived concept of () is a surrogate measure that may correspond to a large number of different possible scenarios, it is also important to look at additional means of characterizing multi-event resilience in order to be able to differentiate between such scenarios. One such approach involves introducing the concept of partial resilience. Given the area above the curve for a particular sub-event, it is relatively straightforward to define the resilience associated with that
Illustrative example
As discussed above, there are a number of different examples of systems that suffer the impacts of multiple events within the context of a larger disaster situation. In order to illustrate the application of the technique introduced above, we focus on the example of an earthquake in a heavily populated area that is subject to potential landslides and possible flooding.
Earthquakes are typically associated with aftershocks. The 2010 earthquake in Chile, for example, had 304 aftershocks of
Discussion
Fig. 12 is more immediately useful than Fig. 11 as an analytical aid because it focuses on just the overall multi-event resilience values. In comparing the five scenarios, we can see that scenario 5 not only has the greatest overall resilience but also it provides a good balance between the amount of damage suffered and the time to recovery. Scenario 3 has the next highest resilience value and represents a slightly quicker process of recovery, even though that facility actually suffered more
Conclusions
We proposed in this paper a multi-criteria approach for capturing the tradeoffs between the robustness of a system and the rapidity of its recovery, in situations involving the occurrence of multiple related disaster or emergency events. The new concept of partial resilience captures the relative time of occurrence and impact of these related events, and it provides the opportunity to both analytically and graphically represent the multi-dimensional and multi-criteria nature of a system under
References (33)
- et al.
Evaluating information assurance strategies
Decision Support Systems
(2005) Representing perceived tradeoffs in defining disaster resilience
Decision Support Systems
(2011)- et al.
Decision support for network disruption mitigation
Decision Support Systems
(2008) - et al.
A two-stage stochastic programming model for transportation network protection
Computers & Operations Research
(2009) - et al.
A bilevel mixed-integer program for critical infrastructure protection planning
Computers & Operations Research
(2008) - et al.
Government preparedness: using simulation to prepare for a terrorist attack
Computers & Operations Research
(2008) - et al.
Real-time supply chain control via multiagent adjustable autonomy
Computers & Operations Research
(2008) - et al.
Prioritization of bridges and tunnels in earthquake risk mitigation using multicriteria decision analysis: application to Lisbon
Omega
(2008) - et al.
Modeling s-t path availability to support disaster vulnerability assessment of network infrastructure
Computers & Operations Research
(2009) - et al.
Pre-disaster investment decisions for strengthening a highway network
Computers & Operations Research
(2010)
Decision support for improvisation in response to extreme events: learning from the response to the 2001 World Trade Center attack
Decision Support Systems
Resiliency-Connectivity metric in wireless sensor networks with key predistibution schemes and node compromise attacks
Physical Communication
A multicrieria decision support system for bank rating
Decision Support Systems
Evaluation and analysis of logistic network resilience with application to aircraft servicing
IEEE Systems Journal
Analysis of electric grid security under terrorist threat
IEEE Transactions on Power Systems
Disaster mitigation and hazard management
Cited by (189)
Community resilience to wildfires: A network analysis approach by utilizing human mobility data
2024, Computers, Environment and Urban SystemsA general design-oriented resilience measurement and evaluation method for engineering systems: Resilience cube
2024, Reliability Engineering and System SafetyThe resilience evaluation of unmanned autonomous swarm with informed agents under partial failure
2024, Reliability Engineering and System SafetyNetwork Resilience: Definitions, approaches, and applications
2024, Journal of King Saud University - Computer and Information SciencesResilience evaluation and optimal design for weapon system of systems with dynamic reconfiguration
2023, Reliability Engineering and System SafetyCharacterisation of resilience metrics in full-scale applications to interdependent infrastructure systems
2023, Reliability Engineering and System Safety