Seismic loss estimation of wood-frame houses in south-western British Columbia

This study extends a seismic loss model of Goda and Hong for multiple buildings by incorporating an updated seismic hazard model for western Canada, a spatial correlation model of seismic effects for subduction earthquakes, and probabilistic models of the maximum inter-story drift ratio of wood-frame houses with different shear-wall configurations, and applies to 1415 existing wood-frame houses in south-western British Columbia. The extended model is likely to produce more realistic and accurate seismic loss estimates. A series of sensitivity analyses is carried out to identify the most influential model components and to provide a range of seismic loss estimates that are based on different but valid assumptions/models. The analysis results highlight the significance of using the updated seismic hazard model. For the upper tail of probability distribution of aggregate seismic loss, the selection of an adequate spatial correlation model is of great importance. Another key model component is the determination of the ultimate seismic capacity, which inevitably involves uncertainty and judgment. These results are useful for directing efforts in future studies to improve the model.