Preliminary engineering cost estimation model for bridge projects

This paper addresses one of the costs of new bridges, i.e., the cost of doing the preliminary engineering (PE). This particular aspect of cost has largely been neglected because of the difficulty in obtaining appropriate data. Typically, PE costs are determined as a percentage of construction costs, disregarding other project-specific parameters. Bridge projects let by the North Carolina (NC) DOT between 2001 and 2009 were reviewed. From an analysis of these 461 projects, the writers developed statistical models linking variation in PE costs with distinctive project parameters. The primary contribution this paper makes to the body of knowledge is the finding that PE cost estimates for bridge projects are commonly and significantly underestimated. The writers found that bridge projects exhibited a mean PE cost ratio of 28%. This result is significantly greater than the percentage used in practice. An investigation of data sources related to bridge projects, a description of regression techniques applied to predict the PE cost of such projects, and an assessment of the predictive performance of models are addressed in this paper. New knowledge contributions include identification of key factors affecting the PE costs of bridges, illustration of how those key factors enabled us to predict PE costs of future projects, and discussion of the data collection problems that hindered model development. The writers' PE cost analyses and predictive modeling effort relate to infrastructure funding issues. All of the parties involved in infrastructure construction, from legislators to contractors, must understand costs and have confidence in the cost estimates developed to ensure efficient and effective funding of infrastructure projects. This paper on bridge PE costs addresses one component of infrastructure funding.