A concept for developing construction element cost models for German residential building projects

Abstract

Early cost estimate of construction projects plays a vital role hence it forms a foundation for budgeting, planning, monitoring and even for litigation aims. In German practice, the first stage cost estimation is performed at the pre-design stage where no design details exist. Thus, the practical relevance of any model to be developed that requires further design information, would be questionable. The German norms describe construction cost elements as the subcategories for cost of construction works. Therefore, the accuracy of the first stage estimate can be improved by modelling construction cost elements where more detailed information incorporates. In this context, it is aimed to demonstrate the development of a predictive model for the most relevant cost element: external walls. Historical data was randomly sampled from Building Cost Information Centre (Baukosteninformationszentrum) database. The most significant predictors for cost of external walls were determined by backwards regression method. Multiple linear regression methodology is employed for the final parameter estimation. The results reveal that substantial variation in the cost of external walls can be explained by the proposed model which leads to high prediction accuracy. The development of the model is anticipated to influence further studies that address cost modelling in Germany.