Security enhancement and consequence mitigation strategies for water infrastructure against physical destruction
Abstract
Protecting water infrastructure systems from intentional attacks and mitigating attack consequences are currently top priorities for the federal, state, and local governments in the United States. Up to now, most vulnerability assessment, resilience enhancement, and consequence mitigation approaches in water industry have been subjective and qualitative. This work complements existing research by proposing a set of quantitative methods, combining optimization techniques with practical simulation tools, for resilience assessment, security enhancement and consequence mitigation. The first half of this research develops a proactive method for allocating a security budget to a water supply network to maximize the network's resilience to physical attack. The method integrates max-min linear programming, hydraulic simulation, and genetic algorithms for constraint generation. The objective is to find a security allocation that maximizes an attacker's marginal cost of inflicting undesirable consequences through destruction of network components. The author illustrates the method using two example networks, including a large network with 685 nodes, and investigates its allocation effectiveness and computational characteristics. The second half of this research develops reactive methods that mitigate the consequences of water shortage resulting from destruction of facilities in water networks. These methods integrate search techniques such as branch-and-bound and genetic algorithms with a hydraulic solver in order to check demand feasibilities across a residual water network. The objective is to identify a feasible customer demand pattern that minimizes the consequences of water shortage in the downgraded network. The author presents a mathematical model of the problem addressed along with an exact solution methodology and several heuristics. The author also applies these to three water networks ranging in size from 12 to 685 nodes and compares the solution quality and computational efficiency.
Degree
Ph.D.
Advisors
Richard, Purdue University.
Subject Area
Industrial engineering|Civil engineering|Operations research
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.