Advancing quantitative risk analysis for critical water infrastructure

Critical infrastructure systems play a vital role in the supply of lifeline services to businesses and the wider public. It is of paramount importance for national security, public health, and economic prosperity that these critical structures function properly. Unfortunately, with respect to drinking water infrastructures in the US, much of the pipeline assets are nearing the end of their useful life and utilities are challenged with maintaining these systems with limited budgets and information. Risk analysis is a useful decision making tool which can allow managers to better identify weaknesses, and aid better investment decisions regarding maintenance, inspection, and repair. The current practice for risk analysis and management of critical water systems falls short of the approaches preferred by risk researchers. The aim of this thesis is to advance to practice and theory. This involves the evaluation of existing methods as well as the incorporation of modern analytical tools to fundamentally advance the state of practice. This thesis first critically analyzes a popular risk analysis standard (J100-10) to establish the knowledge gap between practice and theory in the water domain. Two quantitative methodologies are then explored: machine learning and mathematical optimization. The research here demonstrates how they can be integrated into a broader risk framework and used to improve assessments for water systems. The work presented in this dissertation represents a significant contribution to the field of infrastructure risk and reliability analysis. While the domain application is specific to drinking water systems, the techniques can be applied for other types of networked infrastructures.