System reliability analysis for seismic site classification based on sequential Gaussian co-simulation: A case study in Shiraz, Iran

Abstract

Seismic site classification plays a significant role in the estimation of site response and is carried out using average shear wave velocity through the top 30 m (VS30) soil layers. Meanwhile, at some sites with a number of boreholes, identification of the VS30 acquired from different boreholes does not lead to the classification of the site into one single category, and this makes the seismic classification of the site troublesome. This paper proposes a practical approach for obtaining the reliability index of the seismic site classification when not all VS30 calculated from the boreholes converge to that of a unique seismic class. For this purpose, the geostatistical method is combined with the system reliability approach. A real site with three boreholes, two of which fall into the ‘D' and the third into the ‘C' category based on the National Earthquake Hazards Reduction Program (NEHRP), is considered. The site is divided into 66 hypothetical soil columns. The shear wave velocity (VS) is estimated in depths of each column using the Sequential Gaussian Co-Simulation (SGCS) method based on sampled data determined nearby. Ten thousand realizations of VS with spatial distributions are generated via the SGCS method and are validated with the original data. In order for reliability analysis of the proposed classification approach, the reliability indices of the components are merged based on their corresponding correlation through the Sequential Compounding Method (SCM). The proposed method is proven to be consistent, fast, and the results are shown to be reliable.