An investigation of factors that significantly affect pavement maintenance costs in Texas

Abstract

In the past twenty years, the focus of the highways industry has shifted from the construction of new highways to the maintenance of those which already exist. At the same time, the maintenance budgets of highway agencies are being significantly reduced. Consequently, it is important to ensure that maintenance funds are allocated in a consistent and efficient manner. One method of allocating funds involves using maintenance cost models to estimate the amount of money necessary to adequately maintain the pavement surface in the future. The objective of this research was to identify factors that significantly affect pavement maintenance costs. To do this, a total of seven pavement maintenance expenditure models were developed for highways in the state of Texas; four were developed for two-lane U.S. highways (asphalt, concrete, resurfaced, and thin asphalt pavements) and three were developed for interstate highways (asphalt, concrete, and resurfaced pavements). Many factors were considered including average daily traffic, equivalent single axle loads, distress score, ride score, annual rainfall, annual freeze/thaw cycles, functional classification, as well as the district that maintains the pavement. For each pavement type and highway classification, several factors were found that sufficiently affect pavement maintenance expenditures. The most common factor that was included in each model was the Texas Department of Transportation district that maintains the pavement. Although all of the considered factors were found to be significant for different pavement types, equivalent single axle loads and average daily traffic were found to be significant in almost all models. After the models were evaluated, they were applied to the pavement sections in four TXDOT districts, and the resulting predicted expenditures were compared to actual expenditures. It was found that although the models are significant, they are not accurate enough to be used in the maintenance budget allocation process. One of the primary causes of the inadequacy is due to shortcomings in the data used to calibrate the models.