Sponsor
Portland State University. Department of Civil & Environmental Engineering
First Advisor
David Yang
Term of Graduation
Summer 2026
Date of Publication
8-10-2026
Document Type
Dissertation
Degree Name
Doctor of Philosophy (Ph.D.) in Civil & Environmental Engineering
Department
Civil and Environmental Engineering
Language
English
Subjects
Asset Importance Factor, Bayesian Updating, Infrastructure Resilience, Infrastructure Systems, Retrofit Prioritization, Risk Management
Physical Description
1 online resource (xi, 165 pages)
Abstract
Critical infrastructure systems such as transportation networks serve as the lifelines of modern society, supporting economic activity and ensuring access to critical services. Their disruption under extreme events can produce consequences extending far beyond physical damage to individual assets. In transportation networks for instance, failures of critical assets such as bridges and tunnels can reduce accessibility of communities, limit freight capabilities for commerce, and increase travel distance and travel time of road users. These consequences associated with the system functionality loss are tied not only to the physical vulnerability and failure of individual assets, but also to the system configuration and the corresponding interactions among failed assets. As a result, functionality risk of infrastructure systems, defined herein as the expected functionality loss, cannot be quantified or managed effectively by evaluating the composing assets independently or in isolation; rather, the joint failure patterns of all assets (i.e., the system states) must be considered and analyzed. This mandate for functionality risk assessment brings in three major challenges: (a) the exponential growth of possible system states, (b) the potential influence of low-probability, high-consequence events (i.e., "grey-swan" events), and (c) the difficulty in translating system-level risk into asset-level decisions. To address these challenges, this dissertation establishes a mathematically rigorous, computationally scalable, and operationally actionable methodology for functionality risk assessment and risk-based management for critical infrastructure systems under scenario and probabilistic hazards.
Specifically, the methodology is underpinned by an advanced sampling approach for functionality risk assessment. By introducing an auxiliary Bayesian updating problem, this approach leverages the transitional Markov chain Monte Carlo (TMCMC) algorithm to strategically guide the samples of system states to regions that contribute strongly to the functionality risk. As a result, the approach eliminates the need for exhaustive enumeration of all system states, while effectively capturing the contribution from "grey-swan" events compared to conventional sampling techniques (e.g., the crude Monte Carlo simulation). These advantageous features of the new approach allow for scalable and accurate assessment of the functionality risk of infrastructure systems.
The dissertation further investigates an important by-product of the proposed TMCMC-based approach, i.e., the fictitious "posterior" failure probability of an asset. It is shown mathematically that this "posterior" probability is proportional to the product of the actual failure probability of an asset and the conditional system functionality risk given the deterministic failure of that asset (and the probabilistic failures of all other assets). Leveraging this knowledge, an asset importance factor is created using the "posterior" failure probability and "posterior" reliability of an asset post-TMCMC sampling. It is demonstrated that the proposed asset importance factor is proportional to the derivative of the functionality risk with respect to the actual asset failure probability. Utilizing this connection between the proposed importance factor and the risk gradient, risk minimization under budgetary constraints can be approximated as a classical knapsack problem, where a near-optimal set of retrofit projects is selected using simple heuristics based on benefit-to-cost ratios. Benchmark studies presented in this dissertation indicate that project prioritization based on the proposed importance factor outperformed the three comparison measures across the evaluated scenarios and closely matches the exact optimization solutions across varying budget levels, retrofit effectiveness, and travel demand levels.
Finally, the overall methodology was applied to quantify the post-earthquake hospital accessibility risk in the Portland Metro Area under a Magnitude 9 Cascadia Subduction Zone (CSZ) scenario earthquake. The network model contains 1,756 nodes including 614 population nodes and 18 hospital nodes, 2,639 links representing primary and emergency response roadways designated by the metro, as well as 664 bridges and overpasses located on 461 links. Using the HAZUS fragility models for bridges and the proposed TMCMC approach, the expected reduction in hospital accessibility is estimated at 11.6%. The top-ten bridges that should receive high priority for retrofitting actions are also identified and mapped. Furthermore, the effect of fragility model simplification is investigated further by considering a homogenized fragility model derived from the bridge-specific HAZUS fragility curves. It was shown that using a homogenized fragility model may underestimate the accessibility risk by approximately 31%. Additionally, the homogenized fragility model can significantly alter the ranking of bridge retrofit priorities, especially for the identification of high-priority bridges.
Overall, by addressing the major challenges described above, this dissertation offers a novel, comprehensive, and coherent pathway to the accurate assessment of system functionality risks and the practical implementation of budget-constrained risk mitigation. Although the final application is centered on earthquake hazards, the methodology is general with respect to hazard types and functionality measures and can be further expanded to a wide range of infrastructure systems and extreme events.
Rights
© 2026 Anteneh Zewdu Deriba
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).
Persistent Identifier
https://archives.pdx.edu/ds/psu/45151
Recommended Citation
Deriba, Anteneh Zewdu, "Functionality Risk Assessment and Risk-based Management for Critical Infrastructure Systems" (2026). Dissertations and Theses. Paper 7217.