Modeling damage accumulation during ground-motion sequences for portfolio seismic loss assessments

Conventional earthquake risk assessments use fragility and vulnerability models that are based on seismic demands from individual (mainshock) ground motions, and implicitly assume that a structure is intact before an earthquake hits. This study develops a suite of more realistic state-dependent seismic fragility and vulnerability models for a wide range of building taxonomies, leveraging state-of-the-art methods to account for dynamic damage accumulation in structures due to multiple earthquake events (i.e., ground-motion sequences). Models are developed for 561 building classes (i.e., structural types) from the Global Earthquake Model's global database of fragility and vulnerability models. Four 2010–2012 Canterbury sequence earthquakes are then used to demonstrate an application of the developed models within a portfolio loss assessment, capturing the time-dependent nature of damage and loss in the vulnerability calculations. The results of this application indicate that accounting for damage accumulation across a series of events can significantly increase expected loss ratios compared to a conventional mainshock-only portfolio risk analysis. This work can help analysts to develop and apply state-dependent fragility and vulnerability models for quantifying the potential impact of damage accumulation in portfolio-scale seismic loss assessments.