01590730

Component

The preliminary cost estimate heavily influences the fate of a transportation project, yet it can be up to an order of magnitude off the final bid amount. Inaccurate cost predictions by a state department of transportation (DOT) can lead to either cost overruns or underruns in a project. DOTs usually set standards for the accuracy of plan estimate, and TxDOT’s goal is to limit the difference between plan estimate and the low bidder’s bid within 5% of each project. Currently, average winning bid prices are widely used as a benchmark to check engineer’s unit item prices leading to biased low prices. In this paper, the team proposed a quantile regression model to analyze historical bids to improve the accuracy of unit item price estimate and the overall plan estimate of a project. The empirical results suggest that the engineer’s estimate and winning bid lie around 40th percentile and 33rd percentile respectively for the majority of items. Validation results show that the proposed model provides better unit price estimate as well as total project cost estimate than engineer’s estimate. This methodology is being tested for implementation by the Texas Department of Transportation (TxDOT).

This paper was sponsored by TRB committee AFH10 Standing Committee on Construction Management.

01584066

16-6306

English

15p

2016

Transportation Research Board 95th Annual Meeting

Digital/other

Figures; References; Tables

Bids; Construction projects; Cost estimating; Costs; Engineers; Estimates; Regression analysis

Texas Department of Transportation

Administration and Management; Construction; Highways; I10: Economics and Administration; I50: Construction and Supervision of Construction

Transportation Research Board Annual Meeting 2016 Paper #16-6306

TRIS, TRB, ATRI

Jan 12 2016 6:48PM