A probabilistic performance evaluation for buildings and constructed assets

Abstract

The conservation state of buildings is of increasing interest due to the need to renovate aging building stock and provide a safe and healthy place for end users. Numerous uncertain factors have an impact on building condition, including environmental agents, building age, type of assets and maintenance. Previous studies have focused on identifying these factors, but the relationships among them remain unclear. This paper proposes a Bayesian network (BN) approach to develop a model for assessing a building’s condition. The BN model is based on an extensive review and evaluation of degradation causal factors supported by an analysis of 1974 defects and 5373 maintenance requests in forty buildings. The model was verified by sensitivity analysis, and the proposed approach was tested on an existing building. The model could be used as a supporting tool to identify renovation strategies that can enhance the conservation state of buildings and constructed assets.