Uncertainty quantification for pavement life-cycle stages

Life-cycle assessment (LCA), a common sustainability metric, is usually adopted to quantify the environmental consequences of a product. It has been shown that rolling resistance (RR), a major component of pavement LCA use stage, has significant impact on transportation-related energy consumption. Pavement related RR mainly includes pavement structure, surface roughness (or smoothness) and texture. This research aims at addressing current challenges in pavement LCA use stage. A robust framework is proposed to evaluate RR via developing models for pavement roughness- and structural-induced RR. A roughness–speed impact (RSI) model was developed to quantify the energy and environmental impacts due to RR. The model uses vehicle-specific power as part of the pavement–vehicle interaction (PVI) analysis. According to the model, one unit change of IRI (1 m/km) results in 3% and 2% fuel consumption, respectively, at high and low speeds (105 and 56 km/h) for passenger cars. In addition to the RSI model, the study proposes a practical approach to assess the vehicle excess fuel consumption (EFC) due to pavement deflection. The developed relationship relies on the fundamental energy-deformation principles obtained by conducting nonlinear regression analysis on 3-D finite element (FE) simulations. The proposed model is formulated using a quadratic form of maximum pavement deflection. Factors affect EFC includes, truck loading and speed and pavement temperature. It was found that the estimated EFC for a heavy truck could be as low as 0.03% for a half loaded truck at a temperature of 0 °C a speed of 115 km/h and as high as 6.5% for a fully loaded truck at a temperature of 40 °C and a speed of 8 km/h. This could be increased for low volume road pavement structure. At a speed of 100 km/h, a typical HS20-44 truck could consume an additional 0.5% fuel due to structural rolling resistance (SRR). Uncertainty of pavement roughness has significant impact on the energy and emission output of the pavement-vehicle system depending on the precision level of the model used, input variabilities, and prior knowledge of the model parameters. When quantified uncertainties, successfully utilized in this study, are implemented, LCA parameters prediction would be improved. The introduced RR models may be used as part of the decision-making for short-term energy and emission reduction policies