Simplified seismic loss functions for suspended ceilings and drywall partitions

Abstract

Post-disaster reconnaissance reports frequently list non-structural components (NSCs) as a major source of financial loss in earthquakes. Moreover, minimizing their damage is also of vital significance to the uninterrupted functionality of a building. For efficient decision making, it is important to be able to estimate the cost and downtime associated with the repair of the damage likely to be caused at different hazard levels used in seismic design. Generalized loss functions for two important NSCs commonly used in New Zealand, namely suspended ceilings and drywall partitions are developed in this study. The methodology to develop the loss functions, in the form of engineering demand parameter vs. expected loss due to the considered components, is based on the existing framework for the storey level loss estimation. Nevertheless, exhaustive construction/field data are employed to make these loss functions more generic. In order to estimate financial losses resulting from the failure of suspended ceilings, generalized ceiling fragility functions are developed and combined with the cost functions, which give the loss associated with typical ceilings at various peak acceleration demands. Similarly, probabilities of different damage states in drywall partitions are combined with their associated repair/replacement costs to find the cumulative distribution of the expected loss due to partitions at various drift levels, which is then normalized in terms of the total building cost. Efficiencies of the developed loss functions are investigated through detailed loss assessment of case study reinforced concrete (RC) buildings. It is observed that the difference between the expected losses for ceilings, predicted by the developed generic loss function, and the losses obtained from the detailed loss estimation method is within 5%. Similarly, the developed generic loss function for partitions is able to estimate the partition losses within 2% of that from the detailed loss assessment. The results confirm the accuracy of the proposed generic seismic loss functions.