Interdependent infrastructures and multi-mode attacks and failures: improving the security of urban water systems and fire response

摘要

This dissertation examines the interdependence between urban water distributionsystems and urban fire response. The focus on interdependent critical infrastructures isdriven by concern for security of water systems and the effects on related infrastructuresif water distribution systems are damaged by terrorist attack or natural disaster.A model of interdependent infrastructures (principally water distribution systemsand fire response) is developed called the Model of Urban Fire Spread (MUFS). Themodel includes the capacity to simulate firefighting water demands in a communitywater system hydraulic model, building-to-building urban fire spread, and suppressionactivities. MUFS is an improvement over previous similar models because it allowssimulation of urban fires at the level of individual buildings and it permits simulation ofinterdependent infrastructures working in concert.MUFS is used to simulate a series of multi-mode attacks and failures (MMAFs) ?events which disable the water distribution system and simultaneously ignite an urbanfire. The consequences of MMAF scenarios are analyzed to determine the most serious modes of infrastructure failure and urban fire ignition. Various methods to determineworst-case configurations of urban fire ignition points are also examined.These MMAF scenarios are used to inform the design of potential mitigationmeasures to decrease the consequences of the urban fire. The effectiveness of mitigationmethods is determined using the MUFS simulation tool. Novel metrics are developed toquantify the effectiveness of the mitigation methods from the time-series development oftheir consequences. A cost-benefit analysis of the various mitigation measures isconducted to provide additional insight into the methods? effectiveness and better informthe decision-making process of selecting mitigation methods.Planned future work includes further refinement of the representation of firepropagation and suppression in MUFS and investigation of historical MMAF events tovalidate simulation predictions. Future efforts will continue development of appropriateoptimization methods for determining worst-case MMAF scenarios.This work should be of interest to water utility managers and emergencyplanners, who can adapt the methodology to analyze their communities? vulnerability toMMAFs and design mitigation techniques to meet their unique needs, as well as toresearchers interested in infrastructure modeling and disaster simulation.