Analysis of pavement overlay and replacement performance using random parameters hazard-based duration models

The effectiveness of pavement overlays and pavement replacements in terms of their impact on pavement life is not well understood. This is complicated further by data collection limitations and by the effect that physical deterioration, load volumes, weather, geology, and other factors may have on their effectiveness. Understanding the survivability of overlays and replacements has the potential to provide improved resource allocation and more effective use of state funds. In this paper, pavement overlays and replacements are assessed for their effectiveness on pavement life for urban roads. Using data from Indiana, seemingly unrelated regression equations first are estimated to predict the pavement performance over time. Using these forecasts and historical thresholds, the service life of the pavement is determined and random parameter duration models are estimated to identify influential factors affecting pavement service life. The model-estimation results provide some new insights into the interrelationships among pavement rehabilitation, pavement condition, pavement service life, road functional class, traffic loads and trucks, weather and soil condition, and rehabilitation expenditures.