Survival from disaster : interdependencies management in critical infrastructure networks

Today's critical infrastructure networks are becoming increasingly interdependent. The complexity of these interdependencies has created a new dimension of vulnerability, making the whole system very fragile under unexpected events. To make the whole system more resilient during disasters, the...

Full description

Bibliographic Details
Main Author: Mao, DeTao
Format: Others
Language:English
Published: University of British Columbia 2009
Online Access:http://hdl.handle.net/2429/12268
Description
Summary:Today's critical infrastructure networks are becoming increasingly interdependent. The complexity of these interdependencies has created a new dimension of vulnerability, making the whole system very fragile under unexpected events. To make the whole system more resilient during disasters, the Joint Infrastructure Interdependencies Research Program (JIIRP), aimed at developing methods for reducing its vulnerabilities. As part of the research work in JIIRP, this thesis mainly consists of three sections: (I) Interdependencies Control Strategy (ICS): Since to prevent all vulnerabilities would be intractable, an interdependencies control strategy, which can help to maintain the survival of the critical service(s) is proposed in Chapter 2. A generalized adjacency matrix (GAM) is proposed to represent the physical interdependencies among infrastructure networks. By computation of GAM, decision making for ICS can be made more effective. Moreover, measures for improving survivability of the system are proposed. ICS application to a case study at the UBC campus is detailed in Chapter 3, its effectiveness during the response stage and the recovery stage of the emergency management cycle are demonstrated. (II) Identification of Cascading Pathways for Mitigating Snow-Caused Power Outages: Since most of the present de-icing and anti-icing methods are not fully developed for industrial applications, building a power network that can tolerate any snow storm would be infeasible. In Chapter 4, based on investigation of Vancouver's power outage in November 2006, a dependency network has been built to represent how the cascading pathways unfolded during this disaster. The effects of a changed climate, the causalities and their consequences are illustrated by this model. Through analyses of the dependency network, we propose a systematic strategy to mitigate the impacts of a snow storm to power systems. (III) Contributions to I2Sim: I2Sim is a simulator which was developed to simulate disasters and to develop strategies for dealing with emergencies (Appendix A). The author's contributions to I2Sim are: (a) the modeling and implementation of methods to represent complex cells with multi-input and multi-output (Appendix B); (b)integration of the cluster demon into one single machine; (c) development of a library of functions on GAM operations(Appendix C). === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate