Min Ouyang

Review on modeling and simulation of interdependent critical infrastructure systems

Modern societies are becoming increasingly dependent on critical infrastructure systems (CISs) to provide essential services that support economic prosperity, governance, and quality of life. These systems are not alone but interdependent at multiple levels to enhance their overall performance. However, recent worldwide events such as the 9/11 terrorist attack, Gulf Coast hurricanes, the Chile and Japanese earthquakes, and even heat waves have highlighted that interdependencies among CISs increase the potential for cascading failures and amplify the impact of both large and small scale initial failures into events of catastrophic proportions. To better understand CISs to support planning, maintenance and emergency decision making, modeling and simulation of interdependencies across CISs has recently become a key field of study. This paper reviews the studies in the field and broadly groups the existing modeling and simulation approaches into six types: empirical approaches, agent based approaches, system dynamics based approaches, economic theory based approaches, network based approaches, and others. Different studies for each type of the approaches are categorized and reviewed in terms of fundamental principles, such as research focus, modeling rationale, and the analysis method, while different types of approaches are further compared according to several criteria, such as the notion of resilience. Finally, this paper offers future research directions and identifies critical challenges in the field.

A methodological approach to analyze vulnerability of interdependent infrastructures

Abstract The infrastructures are interconnected and interdependent on multiple levels, the failure of one infrastructure can result in the disruption of other infrastructures, which can cause severe economic disruption and loss of life or failure of services. A methodological approach to analyze vulnerability of interdependent infrastructures has been introduced in this paper, two types of vulnerability are studied: structural vulnerability and functional vulnerability. Infrastructure topologies are only used for analysis on structural vulnerability while operating regimes of different infrastructures are further considered to analyze functional vulnerability. For these two types of vulnerability, interdependent effects are mainly studied and the effects of interdependence strength between infrastructures have also been analyzed. The analysis on structural vulnerability will be helpful to design or improve the infrastructures in the long run while the discussion on functional vulnerability will be useful to protect them in the short term. The methodology introduced in this paper will be advantageous to comprehensively analyze the vulnerability of interdependent infrastructures and protect them more efficiently.

An approach to design interface topologies across interdependent urban infrastructure systems

This paper proposes an approach to design or retrofit interface topologies to minimize cascading failures across urban infrastructure systems. Four types of interface design strategies are formulated based on maximum network component degree, maximum component betweenness, minimum Euclidean distance across components and component reliability rankings. To compute and compare strategy effectiveness under multiple hazard types, this paper introduces a global annual cascading failure effect (GACFE) metric as well as a GACFE-based cost improvement (GACI) metric. The GACI metric quantifies the improvement of the strategy effectiveness per kilometer increment of interdependent link length (ILL) relative to a reference strategy with minimum ILL. Taking as examples the power and gas transmission systems in Harris County, Texas, USA, optimum interface designs under random and hurricane hazards are discussed. Findings include that the strategy based on reliability rankings minimizes the GACFE metric, and decreases the GACI value relative to a reference practical strategy by 10–15% under different power grid safety margins. Such metrics will contribute to coupled utility system design or retrofit given that current guidelines or recommended practices in the utility industry mostly rely on minimum Euclidean distances and are yet to include interdependent effects in their provisions.

Resilience assessment of interdependent infrastructure systems: With a focus on joint restoration modeling and analysis

As infrastructure systems are highly interconnected, it is crucial to analyze their resilience with the consideration of their interdependencies. This paper adapts an existing resilience assessment framework for single systems to interdependent systems and mainly focuses on modeling and resilience contribution analysis of multi-systems’ joint restoration processes, which are seldom addressed in the literature. Taking interdependent power and gas systems in Houston, Texas, USA under hurricane hazards as an illustrative exmaple, five types of joint restoration stategies are proposed, including random restoration strategy RS1, independent restoration strategy RS2, power first and gas second restoration strategy RS3, gas aimed restoration strategy RS4, and power and gas compromised restoration strategy RS5. Results show that under limited restoration resources, RS1 produces the least resilience for both systems, RS2 and RS3 both generates the largest power system resilience while RS4 is the best for the gas system; and if quantifying the total resilience as the evenly weighted sum of two systems’ individual resilience, RS5 produces the largest total resilience. The proposed method can help decision makers search optimum joint restoration strategy, which can significantly enhance both systems’ resilience.

Vulnerability assessment and mitigation for the Chinese railway system under floods

The economy of China and the travel needs of its citizens depend significantly on the continuous and reliable services provided by its railway system. However, this system is subject to frequent natural hazards, such as floods, earthquakes, and debris flow. A mechanism to assess the railway system vulnerability under these hazards and the design of effective vulnerability mitigation strategies are essential to the reliable functioning of the railway system. This article proposes a comprehensive methodology to quantitatively assess the railway system vulnerability under floods using historical data and GIS technology. The proposed methodology includes a network representation of the railway system, the generation of flood event scenarios, a method to estimate railway link vulnerability, and a quantitative vulnerability value computation approach. The railway system vulnerability is evaluated in terms of its service disruption related to the number of interrupted trains and the durations of interruption. A maintenance strategy to mitigate vulnerability is proposed that simultaneously considers link vulnerability and number of trains using it. Numerical experiments show that the flood-induced vulnerability of the proposed representation of the Chinese railway system reaches its maximum monthly value in July, and the proposed vulnerability mitigation strategy is more effective compared to other strategies.

Comparisons of complex network based models and real train flow model to analyze Chinese railway vulnerability

Recently numerous studies have applied complex network based models to study the performance and vulnerability of infrastructure systems under various types of attacks and hazards. But how effective are these models to capture their real performance response is still a question worthy of research. Taking the Chinese railway system as an example, this paper selects three typical complex network based models, including purely topological model (PTM), purely shortest path model (PSPM), and weight (link length) based shortest path model (WBSPM), to analyze railway accessibility and flow-based vulnerability and compare their results with those from the real train flow model (RTFM). The results show that the WBSPM can produce the train routines with 83% stations and 77% railway links identical to the real routines and can approach the RTFM the best for railway vulnerability under both single and multiple component failures. The correlation coefficient for accessibility vulnerability from WBSPM and RTFM under single station failures is 0.96 while it is 0.92 for flow-based vulnerability; under multiple station failures, where each station has the same failure probability fp, the WBSPM can produce almost identical vulnerability results with those from the RTFM under almost all failures scenarios when fp is larger than 0.62 for accessibility vulnerability and 0.86 for flow-based vulnerability.

Comparisons of purely topological model, betweenness based model and direct current power flow model to analyze power grid vulnerability

This paper selects three frequently used power grid models, including a purely topological model (PTM), a betweennness based model (BBM), and a direct current power flow model (DCPFM), to describe three different dynamical processes on a power grid under both single and multiple component failures. Each of the dynamical processes is then characterized by both a topology-based and a flow-based vulnerability metrics to compare the three models with each other from the vulnerability perspective. Taking as an example, the IEEE 300 power grid with line capacity set proportional to a tolerance parameter tp, the results show non-linear phenomenon: under single node failures, there exists a critical value of tp = 1.36, above which the three models all produce identical topology-based vulnerability results and more than 85% nodes have identical flow-based vulnerability from any two models; under multiple node failures that each node fails with an identical failure probability fp, there exists a critical fp = 0.56, above which the three models produce almost identical topology-based vulnerability results at any tp ≥ 1, but producing identical flow-based vulnerability results only occurs at fp = . In addition, the topology-based vulnerability results can provide a good approximation for the flow-based vulnerability under large fp, and the priority of PTM and BBM to better approach the DCPFM for vulnerability analysis mainly depends on the value of fp. Similar results are also found for other failure types, other system operation parameters, and other power grids.

Vulnerability analysis of complementary transportation systems with applications to railway and airline systems in China

Abstract Most of existing studies on vulnerability analysis of multiple infrastructure systems mainly focus on negative effects of interdependencies, which mean that failures in one system can propagate to other systems and aggravate the initial damage. In reality, there also exist positive effects of interdependencies, which are shown in complementary systems and mean that if one system fails another system can provide alternative services to satisfy customers׳ demands. Different types of transportation systems in a city or country are typical complementary systems. Taking railway and airline systems in China as an example, this paper proposes a network-based approach to model the vulnerability of complementary transportation systems, and based on this model, this paper further introduces a dynamic complementary strength metric, which can help decision makers design or select better complementary topologies from the vulnerability perspective. Also, based on a simple genetic algorithm, this paper analyzes whether critical components for single systems are still important when taking two systems as a whole for analysis. Results show that a protection strategy of hardening a few critical components is also good strategy for the combined system. In addition, the findings and several assumptions are further discussed to close the gap between theory and practice.

Efficient Approach to Compute Generalized Interdependent Effects between Infrastructure Systems

Most studies on infrastructure interdependencies only explore a subset of the possible damage states in which they operate. Typically, interdependence effects (IE) are measured in terms of common failure fractions (f) of components across systems. In this paper, an expanded damage space is explored where infrastructure systems can simultaneously experience different failure fractions f. This augmented space is then mapped onto a functionality space where IEs are evaluated as a function of the efficiencies of constitutive systems: a desirable feature because efficiency is typically measured and recorded in practice whereas f is not. These new IEs conditioned on efficiency levels are termed generalized interdependent effects (GIE), which can be predicted or detected in a computationally efficient fashion. Prediction is a function of initial efficiencies before damage propagation and can be used in prefailure analyses, whereas detection is a function of final joint efficiencies for postfailure analyses. To i...

Critical location identification and vulnerability analysis of interdependent infrastructure systems under spatially localized attacks

Infrastructure systems are usually spatially distributed in a wide area and are subject to many types of hazards. For each type of hazards, modeling their direct impact on infrastructure components and analyzing their induced system-level vulnerability are important for identifying mitigation strategies. This paper mainly studies spatially localized attacks that a set of infrastructure components located within or crossing a circle shaped spatially localized area is subject to damage while other components do not directly fail. For this type of attacks, taking interdependent power and gas systems in Harris County, Texas, USA as an example, this paper proposes an approach to exactly identify critical locations in interdependent infrastructure systems and make pertinent vulnerability analysis. Results show that (a) infrastructure interdependencies and attack radius largely affect the position of critical locations; (b) spatially localized attacks cause less vulnerability than equivalent random failures; (c) in most values of attack radius critical locations identified by considering only node failures do not change when considering both node and edge failures in the attack area; (d) for many values of attack radius critical locations identified by topology-based model are also critical from the flow-based perspective.