Pervious Concrete Pavement Performance Modeling Using the Bayesian Statistical Technique

Abstract

Because pervious concrete pavement (PCP) has a porous structure and can percolate water to an underground layer, it has been proposed as a stormwater best management practice (BMP), an environmentally friendly product, and sustainable paving materials. This porosity makes PCP susceptible to freeze-thaw damage in cold climates. Therefore, PCP has not been widely applied and investigated in such a climate. Long-term performance data are rarely available, and no performance model has been developed for PCP to date. The main objective of this research is to integrate expert knowledge (using the Markov-chain process) and experimental data (PCP field investigations) to build a performance model for PCP through incorporation of the Bayesian technique. The combination of these sources of data is an efficient and effective approach to build a performance model for a new type of pavement, such as PCP, which has not had a long-term performance database. As a result, a robust linear performance model is developed and applied to predict the service life of PCP. The service life of PCP is estimated to be approximately nine years using the developed performance model. In general, the expert knowledge leads to more conservative results rather than experimental data.