Comparison of static and dynamic resilience for a multipurpose reservoir operation

Abstract

Reliability, resilience, and vulnerability are the traditional risk measures used to assess the performance of a reservoir system. Among these measures, resilience is used to assess the ability of a reservoir system to recover from a failure event. However, the time‐independent static resilience does not consider the system characteristics, interaction of various individual components and does not provide much insight into reservoir performance from the beginning of the failure event until the full performance recovery. Knowledge of dynamic reservoir behavior under the disturbance offers opportunities for proactive and/or reactive adaptive response that can be selected to maximize reservoir resilience. A novel measure is required to provide insight into the dynamics of reservoir performance based on the reservoir system characteristics and its adaptive capacity. The reservoir system characteristics include, among others, reservoir storage curve, reservoir inflow, reservoir outflow capacity, and reservoir operating rules. The reservoir adaptive capacity can be expressed using various impacts of reservoir performance under the disturbance (like reservoir release for meeting a particular demand, socioeconomic consequences of reservoir performance, or resulting environmental state of the river upstream and downstream from the reservoir). Another way of expressing reservoir adaptive capacity to a disturbing event may include aggregated measures like reservoir robustness, redundancy, resourcefulness, and rapidity. A novel measure that combines reservoir performance and its adaptive capacity is proposed in this paper and named “dynamic resilience.” The paper also proposes a generic simulation methodology for quantifying reservoir resilience as a function of time. The proposed resilience measure is applied to a single multipurpose reservoir operation and tested for a set of failure scenarios. The dynamic behavior of reservoir resilience is captured using the system dynamics simulation approach, a feedback‐based object‐oriented method, very effective for modeling complex systems. The results of dynamic resilience are compared with the traditional performance measures in order to identify advantages of the proposed measure. The results confirm that the dynamic resilience is a powerful tool for selecting proactive and reactive adaptive response of a multipurpose reservoir to a disturbing event that cannot be achieved using traditional measures. The generic quantification approach proposed in the paper allows for easy use of dynamic resilience for planning and operations of various civil infrastructure systems.

Using system dynamics simulation for assessment of hydropower system safety

“…Once performance measures have been selected for the system of interest, these can be used to estimate system resilience, which represents the last step in the overall methodology (Figure ). Many resilience definitions in the literature have been developed for different domains—for example economics [ Pant et al ., ], natural or man‐made disasters [ Tierney and Bruneau , ; Simonovic and Peck , ; Ayyub , ; Rosati et al ., ], and hydropower system benefits [ Simonovic and Arunkumar , ]. Ayyub [] provides a comprehensive overview of resilience metrics used for a wide range of applications.…”

“…A small group of researchers have noted the following shortcomings associated with traditional risk assessment approaches in the dams industry and other industries [ Regan , ; Hartford et al ., ; King et al ., ]. The focus of traditional risk analysis tends to be on extreme events at the edge of the design envelope in terms of structural loads and inflows, while failures may occur well within the design envelope due to a combination of events which individually may not be uncommon [ Komey et al ., ; Hartford et al ., ] Selection of events and failure modes, as well as assigning probabilities to events can be a very subjective process due to the lack of data relating to dam failures and the variety of infrastructure components and other contributing factors involved in each incident [ Hartford and Baecher , ; Leveson , ] Events are often assumed to be completely independent of one another, despite the fact that common cause events are possible [ Putcha and Patev , ; Leveson , ; Komey et al ., ] Traditional analysis techniques such as fault trees and event trees often assume a linear progression of events, ignoring component interactions that can lead to nonlinear system behavior [ Regan , ; Leveson , ; Thomas , ] Systems are decomposed into more manageable subsystems for analysis and the interactions between them are completely ignored or simplified [ Regan , ; Leveson , ; Thomas , ] Traditional risk assessment tends to focus on the generation of singular values representing the probability of event occurrence and the consequence, when in reality probabilities are subjective and consequences are dynamic as the level of reduced performance can vary in time before the system recovers fully [ Leveson , ; Simonovic , ; Simonovic and Arunkumar , ]. …”

Disaster-Related Resilience as Ability and Process: A Concept Guiding the Analysis of Response Behavior before, during and after Extreme Events

“…All four "properties" as called by Bruneau et al "help" or serve as basis for resilience but it is claimed here that they are not directly part of it as a process. These four properties: robustness, redundancy, resourcefulness and rapidity, introduced by Bruneau et al [6] resurface, among others, in the publication of Simonovic and Arunkumar [10], however with a somehow different interpretation.…”

“…But especially in research aiming to apply the concept of resilience, be it qualitative or quantitative assessments such as risk indices or damage functions, more scientific rigor for narrowing down resilience should be observed. For example, new, aggregated resilience metrics have been developed and are proposed as replacements for risk-based performance assessment [9] [10]. While the proposed metric enables the dynamic change of "resilience" to be monitored and hence contributes to the comprehensive assessment of how a disaster and the recovery unfold, it is different from the resilience metric (time elapsed between the disaster and the achievement of pre-disaster performance level, see also Figure 1) as proposed by Hashimoto et al [11] and Duckstein et al [12] for water resources systems.…”

Understanding the Resilience of Soil Moisture Regimes

“…The concept of resilience has been applied in many areas, including ecosystems (Botter et al, ; Holling, ; Srinivasan & Kumar, ), water resources systems (Hashimoto, ; Simonovic & Arunkumar, ; Simonovic & Peck, ), and infrastructure development (Bruneau & Reinhorn, ; Nguyen et al, ). Peterson et al () defined resilience as the difference in system state between a stable state and the threshold of a regime shift.…”