Using Parametric Modeling to Estimate Highway Construction Contract Time

Abstract

Federal regulations require that state transportation agencies have written procedures for setting the construction contract time for highway projects. Since the institution of those regulations, state agencies have used a variety of methods to estimate and set contract time. FHWA recommends the use of methods based on production rate and activity precedence and logic in this process. For years, the Kentucky Transportation Cabinet used a system based around these requirements and recommendations only to find that the system produced estimates varying widely from actual construction times. The use of parametric modeling and historic contract times produced a system that more accurately estimates contract times and with an associated Microsoft Excel–based tool is more user friendly. This approach used project cost and bid item quantity data in multivariate regression–based modeling to develop a set of equations that provide the Kentucky Transportation Cabinet with contract time estimates for projects in a matter of minutes by using data readily available at later design stages.

Understanding the Effect of Bid Quantities, Project Characteristics, and Project Locations on the Duration of Road Transportation Construction Projects during Early Stages

“…The present paper, on the other hand, explores a model that does not control for size and location, but is still able to achieve more accurate results. The increased accuracy can be attributed to the use of charge days to measure project duration versus the use of calendar days as used in previous research [3]). Another interesting finding is that the present research found that Terrain has no significant effect on project durations, even when some researchers have previously found that Terrain is a driver of production rates (5), which can be tied to project durations.…”

“…As a result, this research’s effort is focused on determining which are the most significant variables when trying to determine project durations to ease the estimation of project durations. A better understanding of such relationships will help facilitate a more accurate estimate of contract times, which is a key step in determining contract duration ( 3 ). The current research aims to expand on those models created by Zhai et al ( 3 ) by incorporating new project characteristics into parametric models that have not been considered in prior research.…”

“…A better understanding of such relationships will help facilitate a more accurate estimate of contract times, which is a key step in determining contract duration ( 3 ). The current research aims to expand on those models created by Zhai et al ( 3 ) by incorporating new project characteristics into parametric models that have not been considered in prior research. The independent variables included in the current study are: construction bid quantities, location, project type, annual average daily traffic (AADT), project size (2003 USD), terrain type, and project condition (new versus old).…”

“…Having a good understanding about what influences project durations is an important part of developing accurate schedules. Accurate schedules can help mitigate the problems described by Zhai et al ( 3 ). First, short schedules might discourage smaller contractors, affect bid prices, and potentially reduce the quality of work produced.…”

“…In an effort to better understand the drivers of project durations, this research analyzes historical data from Colorado Department of Transportation (CDOT) using a multiple linear regression (MLR) model. Similar to the research conducted by Zhai et al ( 3 ), the current model uses bid quantities and Engineers’ Estimates as predictor variables and also incorporates terrain as a factor, making it possible to determine whether one predicting tool can be used for several states or if individual tools per state are required.…”