Influence of field collected clogging material on the coefficient of permeability in porous asphalt mixture

Urban hydrology undergoes significant changes with the increasing prevalence of impervious surfaces during urbanization. Stormwater runoff from these impervious surfaces introduces pollution into surrounding waterways. A common objective in urban areas is the management of stormwater runoff and the improvement of water quality near cities, often achieved through the use of porous asphalt mixtures (PAM). However, the hydrological performance of permeable pavements tends to degrade over time, primarily due to the gradual accumulation of materials in the field that clog the surface. The extent of surface clogging becomes a critical factor in evaluating the effectiveness of permeable pavements and determining the absorbed permeability. The improvement rates of clogging are influenced by factors such as the location, site characteristics, and frequency of rain events. In this study, four different types of mixtures were employed, and samples were created and exposed to simulated stormwater events. The goal was to elucidate the relationships between the physical components of the system and the development of surface clogging. Models were developed using an artificial neural network (ANN) tool to analyze and predict clogging progression rates. The models incorporated variables such as thickness, material per cycle, infiltration time (seconds), and voids filling material. The ANN-based model demonstrated a high level of accuracy, with a 97% success rate in predicting the level of clogging along a stretch of permeable pavement. In summary, the developed model proves to be a reliable tool for predicting the degree of clogging in permeable pavements. By adhering to PAM requirements and considering the clogged percentage, the model can anticipate the precise cumulative coefficient of permeability value, contributing to effective management strategies for urban storm water runoff.