New approach to developing conceptual cost estimates for highway projects

Developing a reliable project cost estimate is a challenge for any state highway agency (SHA), especially at the conceptual stage. A conceptual estimate is defined in this paper as the estimate prepared at the point at which only 30% of the design is complete. This paper describes a statistical approach to producing a reliable conceptual cost estimate when few project design details have been finalized and many assumptions still form the basis of the estimate. This approach used an analysis similar to the program evaluation and review technique (PERT), which is more commonly used in project scheduling, to assign certainty factors to cost estimates. The approach uses a combination of historical bid data for major roadway items whose quantities can be estimated early in the development process and historical percentages for other major components of the project, called allowance and contingency factors. The paper focuses on (1)the methodology developed to analyze the historical bid data; (2)the analysis of 14 highway corridors with 77 projects whose as-bid construction costs were more than $830million; and (3)a cross-validation of the approach used to validate the accuracy of the predictive model. By using a PERT-type technique, construction costs were accurately predicted at the conceptual stage within 20%. However, approximately 85% of the corridor costs were accurately predicted within 15% of the actual cost. The proposed methodology provides a structured and consistent estimating approach that can be used by any SHA that needs to develop total project delivery estimates at the conceptual design stage.