Temperature effects on the infiltration rate through an infiltration basin BMP

Infiltration Best Management Practices (BMPs) are becoming more readily acceptable as a means of reducing postdevelopment runoff volumes and peak flow rates to pre-construction levels, while simultaneously increasing recharge. However, the design, construction, and operation of infiltration basins to this point have not been standardized due to a lack of understanding of the infiltration processes that occur in these structures. Sizing infiltration BMPs to hold and store a predetermined volume of runoff, typically called the Water Quality Volume, has become a widely accepted practice. This method of sizing BMPs does not account for the infiltration that is occurring in the BMP during the storm event; which could result in significantly oversized BMPs. The objective of this study was to develop a methodology to simulate varying infiltration rates observed from a large scale rock infiltration basin BMP. The results should aid in improved design of such structures. This methodology is required to predict the performance of these sites using single event and continuous flow models. The study site is a Pervious Concrete Infiltration Basin BMP built in 2002 in a common area at Villanova University. The system consists of three infiltration beds filled with coarse aggregate, lined with geotextile filter fabric, overlain with pervious concrete and underlain by undisturbed silty sand. The BMP is extensively instrumented to facilitate water quantity and quality research. The infiltration performance of the site is the focus of the study. Recorded data indicates a wide variation of linear infiltration rates for smaller storm events. A model was developed using the Green-Ampt formula to characterize the infiltration occurring in the basin for small storm events characterized by an accumulated depth of water in the infiltration bed of less than 10 cm. The effectiveness and accuracy of the model were measured by comparing the model outputs with observed bed water elevation data recorded from instrumentation on site. Results show that for bed depths of < 10 cm, hydraulic conductivity is the most sensitive parameter, and that the storm event measured infiltration rate is substantially less then the measured saturated hydraulic conductivity of the soil. The governing factor affecting hydraulic conductivity, and subsequently, infiltration rate is temperature; with higher rates occurring during warmer periods, affecting the infiltration rate by as much as 56%.