Impact of CyberShake on Risk Assessments for Distributed Infrastructure Systems

Yajie Lee, Christine A. Goulet, ZhengHui Hu, Kevin R. Milner, & Scott Callaghan

Submitted September 11, 2022, SCEC Contribution #12487, 2022 SCEC Annual Meeting Poster #274

Empirical ground motion models (GMMs) are typically used to quantify spatially correlated ground motion hazard in probabilistic seismic risk analysis (PSRA) for spatially distributed systems. They represent “average” source, path attenuation, and site response characteristics of global earthquakes, and are associated with large variability components that reflect a variety of crustal structures and conditions. The ground motions in a specific region are often different from those from the global average and are expected to exhibit lower variability and region-specific spatial correlation. Such differences can lead to biased and wider than necessary distributions of the risk metrics.
The SCEC CyberShake platform was designed as the first physics-based simulations that better capture the local source, path, and site effects for seismic hazard assessment than traditional GMMs. CyberShake simulates over 500,000 earthquake ruptures in southern California and propagates waves through 3D velocity models. The approach is consistent with PSRA in which we use CyberShake results as the hazard component.
For this work, we built on a recent PSRA study of the underground water pipeline network for the City of Los Angeles, where the system-level performance (measured by expected number of pipeline repairs) was established as a function of exceedance probability based on a large set of event simulations from the UCERF3 earthquake source model. The study utilized the NGA-West2 GMMs to construct random ground motions with a level of spatial correlation that was consistent with empirical models. By repeating our study using the events simulations from CyberShake 15.12, we explored differences from simulations relative to empirical. Our results show that using the same set of earthquake ruptures and pipe fragility models, CyberShake simulations result in significantly lower estimates of risks than those computed from the empirical approach. More specifically, at a 500-year return period, the expected number of pipeline repairs computed from CyberShake is about 26% lower than that from empirical GMMs. At a 2,500-year return period, the reduction reaches 41%. While a reduction was expected, more work is needed to fully understand the sources of these differences. Future work will quantify the impacts on PSRA of spatially distributed systems from median ground motion attenuation, variability, and spatial correlation modeled by the two alternative approaches.

Key Words
CyberShake, Probabilistic Risk Assessments, Distributed System, Infrastructure,

Citation
Lee, Y., Goulet, C. A., Hu, Z., Milner, K. R., & Callaghan, S. (2022, 09). Impact of CyberShake on Risk Assessments for Distributed Infrastructure Systems. Poster Presentation at 2022 SCEC Annual Meeting.

Related Projects & Working Groups
Earthquake Engineering Implementation Interface (EEII)