Abstract

When performing a regional seismic risk assessment, one is interested in estimating the damage, losses, and, in general, the seismic performance of several structures spatially distributed within a region. In this type of probabilistic risk analysis, it is common to consider the spatial correlation between the ground motion intensities affecting different locations. However, the correlation of damage experienced by different structures conditioned on their levels of intensities, herein referred to as structure-to-structure damage correlation, is usually neglected. This study evaluates the effect of the structure-to-structure damage correlation and proposes a simplified model for incorporating it into regional seismic risk assessment through the use of copula functions. In order to illustrate the influence of the structure-to-structure damage correlation on the regional seismic risk, a case study in San Carlos, California, is analyzed considering and neglecting the structure-to-structure damage correlation. Results show that neglecting the structure-to-structure damage correlation may lead to significant underestimations of large-consequence risks.