Resilience analysis of maritime transportation systems based on importance measures

Highlights

• The MTS network model based on the main ports and routes is established.

• State of the post-disaster MTS is analyzed and the residual resilience is proposed.

• Some residual resilience importance methods for the post-disaster MTS are given.

• 23 cities’ sea routes are used to demonstrate the applicability of the proposed method.

Abstract

In maritime transportation system (MTS), ports and ocean routes are essential for establishing and maintaining effective international trade routes. However, the ability of the ports to send and receive goods can be easily destroyed by political and natural interferences. This will cause a significant negative socio-economic impact such as port operation suspension and route disruption. Effectively implementing resilience management in MTS can therefore improve its ability to handle interruptions and minimizing losses. Based on the post-disaster analysis, this paper proposes a new method to optimize the residual resilience management of ports and routes in MTS and proposes an optimal resilience model. The residual resilience is then applied to some importance measures. The Copeland method is used to comprehensively rank the importance of ports and routes. The restoration priority of interrupted ports and routes of different importance measures for the purpose of minimizing residual resilience is also studied. Sea routes consisting of 23 cities are used to demonstrate the applicability of the proposed method.

Introduction

Under the trend of economic globalization, the international trade has been thriving and requiring long and complex supply chains. Maritime transportation system (MTS) is an important pillar of the international supply chain [1]. In a long and complex supply chain system, MTS is more likely to be disrupted by man-made and natural disasters. For example, in 2004, the coast of Indonesia's Sumatra Island was hit by a large earthquake, and the tsunami severely affected the global supply chain. The 2008 snow disaster in China caused some ports along the Yangtze river to be closed. Consequently, a large number of cargo ships in the Shanghai port were unable to berth and sail normally, and the cargo throughput of Shenzhen and Guangzhou ports dropped significantly. In 2011, the Tōhoku earthquake and tsunami in Japan resulted in the destruction of many ports, which costed Japan more than $3.4 billion in maritime trade losses. The port disruption in Indonesia and the hurricane in Australia in 2017 had a tremendous impact on the Asian coal market. In 2019, a report by Nanyang Technological University and Cambridge University showed that if 15 ports in 5 Asian countries (China, Japan, South Korea, Singapore and Malaysia) are directly paralyzed by cyber-attacks, it could cause economic losses of up to US$110 billion. The interruption of maritime transportation caused by natural disasters, maritime accidents, war factors and global public health has seriously affected the global economy and security issues. Wan et al. [1] develop a novel model to assess the risk factors of maritime supply chains by incorporating a fuzzy belief rule approach with Bayesian networks. Wan et al. [2] identify the major risk factors influencing the safety and security of maritime container supply chains to aid the effective management of the associated risks. Since disasters are extremely destructive and unpredictable, it is impossible to protect MTS by eliminating the occurrence of disasters. The best solution is to start from the perspective of recovery and restore the system to the greatest extent in the shortest time. Therefore, this paper combines resilience and different importance measures to study the optimal recovery sequence of failed ports and routes based on the minimum residual resilience. This method can restore system performance to the greatest extent in the shortest time, thereby reducing economic losses and better guiding the recovery process of the MTS after a disaster.

In terms of the resilience management of MTS, Verschuur et al. [3] research the extent of the disruption and the potential resilience of the port and maritime network. Liu et al. [4] develop a novel framework with supporting models to identify and analyze the relevant vulnerabilities in the maritime supply chains. Asadabadi and Miller-Hooks [5] propose models for assessing and improving resiliency and reliability of port networks. Wan et al. [6] propose a systematic review on transportation resilience with emphasis on its characteristics, and research methods applied in different transportation systems. Wu et al. [7] present a framework for the resilience assessment of maritime shipping networks by quantifying its recovery performance under natural hazards. Adjetey-Bahun et al. [8] propose a simulation-based model for quantifying resilience in mass railway transportation systems by quantifying passenger delay and passenger load as the system's performance indicators. Cimellaro et al. [9] evaluate disaster resilience based on analytical functions related to the variation of functionality. Zhang et al. [10] explore resilience measures in network systems from different perspectives and analyze the characteristics of nodes and edges during failures, the matrices of node resilience and edge resilience. Cai et al. [11] propose a dynamic Bayesian network to predict the resilience value of an engineering system. Chen et al. [12] establish a model of measuring supply chain resilience based on the cost composition of the supply chain operating in the interrupted environment. Bao et al. [13] propose a tri-level model explicitly integrating the decision making on recovery strategies of disrupted facilities with the decision making on protecting facilities from intentional attacks. Xing and Levitin [14] study the resilience of linear consecutively connected systems with connection elements under corrective maintenance. Feng et al. [15] present some general methodologies for resilience design under internal deterioration and external shocks and apply them into offshore wind farm. Hossain et al. [16] express the basic factors that can enhance the resilience of the port system as different resilience capacities and quantify the resilience of the port infrastructure by applying a Bayesian network. Almutairi et al. [17] develop a framework that finds the joint influences of disruptive scenarios and groups of stakeholders to disrupt the importance ranking of initiatives. Hu et al. [18] propose a framework to analyze marine liquefied natural gas offloading systems' dynamic resilience considering weather-related hazards based on Infrastructure Resilience-Oriented Modelling Language. To comprehend the interdependency between inland port infrastructure and its surrounding supply chain, Hossain et al. [19] outline three thorough interdependency types: geographic, service provision to identify the factors related to port disruption and its supply chain performance. Shafieezadeh and Burden [20] propose a probabilistic framework for scenario-based resilience assessment of infrastructure systems. Praetorius et al. [21] explore everyday operations of the Vessel Traffic Service system to gain insights in how it contributes to safe and efficient traffic movements. Pitilakis et al. [22] introduce an engineering risk-based methodology for stress testing critical infrastructures. Park and Lee [23] employ an event-based learning approach to analyze the operational discrepancy of container workflows in operational processes and estimate the operational manageability based on the management components from the analytical outcomes. Berle et al. [24] present a structured Formal Vulnerability Assessment methodology, seeking to transfer the safety-oriented Formal Safety Assessment framework into the domain of maritime supply chain vulnerability. He et al. [25] propose an approach on network modeling and robustness assessment for multimodal freight transport networks. Nguyena et al. [26] propose a risk analysis model featuring a quantification of the uncertainty.

The current research on resilience mainly focuses on complex systems. It is believed that resilience is determined by the degree and speed of performance recovery after system components fail. However, the recovery strategy after system component failure is also considered the key to managing resilience. This paper proposes using importance measures into resilience management. The purpose is to study the recovery sequence of failed components in the system, so that the system can quickly recover to its best state.

In terms of the resilience importance measures, Xu et al. [27] propose a new resilience-based component importance measure for networks. Fang et al. [28] propose the optimal repair time and the resilience reduction worth to measure the criticality of the components of a network system. Si et al. [29] provide the equations of various extended importance measures. Dui et al. [30] propose an extended joint integrated importance measure effectively to guide the selection of preventive maintenance components, aiming to maximize gains of the system performance. Dui et al. [31] study the Birnbaum importance measure, integrated importance measure, and the mean absolute deviation with respect to the changes in optimal system structure throughout the system's lifetime. Wu et al. [32] introduce an importance measure for selecting components for preventive maintenance. Almoghathawi and Barker [33] propose component importance measures to analyze the variations of a network recovery. Miziula and Navarro [34] extend the Birnbaum importance measure for the case of a system with dependent components to obtain relevant properties such as connections and comparisons with other measures proposed and studied recently. Barker et al. [35] provide two resilience-based component importance measures to measure component importance.

It can be seen from the above literature review that some work has been done to quantify the resilience of MTS from different perspectives. The existing literature lacks a resilience measure for solving the following problems: If the MTS suffers from disasters, multiple ports and routes are prone to fail at the same time. In the case of limited resources, how can one determine the recovery sequence of the port and routes so that the MTS can be repaired quickly in the shortest time?

This paper investigates the resilience of MTS from a dynamic perspective. Based on the minimum residual resilience optimization model, the residual resilience is applied to the optimal repair time (OPT) residual resilience importance measure, Birnbaum residual resilience importance measure, risk achievement worth (RAW) residual resilience importance measure and risk reduction worth (RRW) residual resilience importance measure, respectively. The recovery sequences of failed ports and routes under different residual resilience importance measures are comprehensively considered to determine the best recovery sequence of failed ports and routes. This method is a dynamic process, each time a failed node is restored, the traffic in the entire network is redistributed with maximum demand. This method can better adjust the traffic volume of different ports and routes, and make the entire network reach the optimal state of traffic volume. It can further enrich the literature in the field of quantitative assessment of maritime resilience.

This rest of this paper is structured as following. Section 2 first introduces the international routes. The MTS network model based on ports and routes is established. Next, the state of the post-disaster MTS is analyzed, and a concept of residual resilience is proposed. Section 3 proposes some residual resilience importance measures for the post-disaster MTS to evaluate the recovery priority of the interrupted ports and routes with the minimum residual resilience. Section 4 applies a numerical example of sea routes to verify the applicability of the proposed methods. Section 5 concludes the paper and proposes the future work.

Notations