Cost estimation models for seismic retrofit of masonry school buildings

Abstract

Seismic retrofits can significantly reduce the vulnerability of masonry buildings. Multiple seismic retrofit alternatives are often available for a given building. Estimating the cost of each candidate action is essential to selecting and implementing the appropriate seismic retrofit initiatives. This research harnesses the capabilities of various regression models for cost estimation. A dataset from 167 retrofit projects for masonry school buildings was used to develop models. Three main retrofit actions were taken into consideration: shotcrete, steel belt and fiber-reinforced polymer. Various regression methods, including multiple linear regression, ridge regression, lasso regression, and elastic net regression, were applied to develop models that estimate the cost of each of the three retrofit actions based on a set of explanatory variables. Next, the models underwent a reduction process to simplify and increase their prediction precision. In most models, the reduction process eliminated multiple terms and enhanced the model prediction performance. This article identifies the height of the building as the most influential parameter governing retrofit costs. It examines the effect of proxy variables (e.g. lateral area of walls and added lateral strength) on the final retrofit cost by adding them as explanatory variables.