Predicting the service life of buildings and components

Abstract

The prediction of the service life of buildings and engineered structures is complex. Such predictions are usually carried out on a component-by-component basis, but must incorporate variations in component design, material type, usage and environment. Extensive databases of life prediction are required to cover the vast number of components/environments/usage combinations that exist in buildings or engineered structures. This study investigated the diverse methods used to model the life of components in structures, including data mining of maintenance and service records, expert opinion surveys (Delphi surveys), multiscale process modelling, and sensing. These methods were analysed in terms of the issues identified above and their applicability to meet the differing uses of life prediction data. Detailed case studies are presented and the strengths and weaknesses of the methods are analysed with respect to the final uses of life prediction.