End-to-End Seismic Risk Analysis Framework for the Identification of Infrastructure Network Retrofits

Lifeline infrastructure network functionality is key for community operation and
safety, especially after natural disasters, such as earthquakes. Thus, this dissertation discusses seismic risk analysis of distributed lifeline infrastructure networks and
mitigation strategies informed by this risk analysis to limit the impacts of future
earthquakes. First, this thesis proposes an end-to-end framework that encompasses
seismic hazard characterization, network performance estimation, and network component retrofit selection. This thesis proceeds to analyze and propose methods within
each of these modules to maximize fidelity while maintaining computational palatability for application to large complex networks. A comparative analysis of techniques
for the simulation of ground motion amplitude fields is performed to determine the
optimal method. The results designate Circulant Embedding to be a computationally superior method and identify trade-offs in terms of computational efficiency and
applicability with other techniques. Then, risk informed heuristics for network component retrofit selection are proposed and evaluated against existing methods. The
findings demonstrate that the proposed risk informed metrics combined with existing
heuristics—specifically, the Tempered Pipe Participation Factor combined with the
Risk Achievement Worth—yield superior retrofit selection, indicating the value of the
integration of risk analysis to pipe importance estimation. Finally, statistical learning
techniques are applied to the estimation of network performance, and their potential
to be a computationally efficient supplement or replacement of physics based models is evaluated. This work identifies a strategy that may adequately estimate network
performance at a fraction of the computational expense compared to full physical
models. Ultimately, this dissertation contributes to the large body of research regarding the reliability of lifeline infrastructure networks exposed to natural hazards.
The framework established here provides a foundation for risk informed mitigation
strategies for real life complex networks in pursuit of resilient communities.

Wu, J and Baker, JW . (2020). End-to-End Seismic Risk Analysis Framework for the Identification of
Infrastructure Network Retrofits. Blume Earthquake Engineering Center Technical Report 204. Stanford Digital Repository. Available at: https://purl.stanford.edu/pd168xb2523