Incorporating Variability into Pavement Performance, Life-Cycle Cost Analysis, and Performance-Based Specification Pay Factors

This paper describes a recent research study that examined how changes in design life affected the pavement life-cycle cost and ultimately how the reduction in or addition to life-cycle cost attributed to superior or inferior in-service performance could be used as a basis for establishing a pay factor for a performance-based specification. Previous models were developed with data from the Canadian Long-Term Pavement Performance Program, which indicated that overlay thickness, total prior cracking, annual freezing index, annual days with precipitation, and accumulated equivalent single-axle loads (ESALs) after 8 years affected the slope of pavement deterioration for asphalt overlay pavements. One of these models, as well as data from the U.S. Long-Term Pavement Performance test sites, is used to determine the service life of asphalt overlay pavements. This paper examines how the variability associated with overlay thickness, total prior cracking, and accumulated ESALs after 8 years affects the service life of asphalt overlay pavements. Furthermore, this paper considers the variability associated with the discount rate and incorporates all associated variability into the life-cycle cost analysis (LCCA). LCCA is performed by using Monte Carlo techniques. On the basis of a recent study, distributions for service life and life-cycle costs are developed by using both normal and lognormal distributions for overlay thickness. With the LCCA values for typical design lives, a sensitivity analysis is subsequently performed to evaluate the impact of 10%, 20%, and 30% differences in the in-service performance as compared to the design life. These LCCA differences are then used as a basis for establishing pay factors. Overall the paper attempts to relate design to in-service performance life-cycle cost and the ultimate use of pay factors.