← Back to Sessions

Importance of Site-Specific Ground Motion Data for Critical Facilities

Description: 

For rigid structures of critical facilities situated on hard rock, accurate characterization of ground motions at high frequencies is critical. Empirical ground motion models (GMMs) typically lack enough observational data on hard rock sites, and predictions from the GMMs may carry substantial uncertainties. To adjust the seismic hazard for hard rock site condition, the current state-of-practice is to perform a V_S-kappa0 (kappa0 is the near-surface high frequency attenuation parameter) host-to-target correction to develop hard rock correction factors. However, in absence of site-specific ground motion (SSGM) data, the modeling parameters and their corresponding epistemic uncertainties are not well quantified, which leads to unreliable hard rock correction factors and hard rock adjusted seismic hazard results.

We have carried out an alternative approach to characterize the hard rock site response (HRSR) in probabilistic seismic hazard analysis (PSHA) using SSGM data. The approach relies on estimation of non-ergodic HRSR at a given site using SSGM recorded on broadband seismographs. Using non-ergodic site response, we effectively remove the site-to-site variability from PSHA while accounting for epistemic uncertainty of HRSRs which are well-quantified with sufficient recorded SSGM data. This approach can be implemented even for regions with low-to-moderate seismicity rate, where earthquakes with M > 3 recorded at distances > 50 km are common. Since 2020, we have instrumented multiple BC Hydro's critical dam facilities in British Columbia, Canada, to characterize the non-ergodic HRSR. Depending on the rate of seismicity and the level of noise at a given facility, we were able to collect sufficient SSGM data within 1-3 years. In addition, having multiple seismic stations at selected facilities has enabled us to explore the degree of topographic effects on HRSR. The non-ergodic HRSR can differ significantly from generic HRSR specifically at high frequencies, and the observed differences are likely due to factors that are not considered in typical ground response analysis (e.g., 2D, 3D effects).

Session: Assessing Seismic Hazard for Critical Facilities and Infrastructure – Insights and Challenges - I

Type: Oral

Date: 5/3/2024

Presentation Time: 02:45 PM (local time)

Presenting Author: Behzad

Student Presenter: No

Invited Presentation: 

Authors