Green infrastructure practices treat stormwater runoff by capturing, infiltrating, and treating water at the site level. Many practices have underdrains that allow water to exit the green infrastructure practice into the stormwater system. Real-time controls of the underdrain may be able to improve hydraulic performance of these systems; however, it is unclear how simple control rules can best be applied to meet competing objectives of green stormwater infrastructure or account for unique site conditions. This study seeks to address this challenge by modeling various control algorithms and site conditions of a permeable pavement with a real-time control system applied to the underdrain. The model was validated at a site where a real-time control system was installed at the end of a 6-inch underdrain in a permeable pavement located in Milwaukee, WI. Using this model, the study tested the performance of various control algorithms which used water level in the system to open and close a valve at the end of the underdrain. In addition, a sensitivity analysis was conducted to evaluate the impact the site characteristics (blockage of the underdrain, infiltration rate, drainage area, groundwater table level, etc.) had on the performance of the system. Results indicate that real-time controls improve the volume of captured water that is infiltrated and help to balance the competing objectives of maximizing infiltration and preventing overflow. However, plugging the underdrain completely performs best if maximizing infiltration is the lone objective. In addition, the sensitivity analysis indicated that real-time controls did not significantly increase infiltration in cases where exfiltration rates into the subgrade soil (uniform sand/gravel, K = 1.0E-1) is high enough to soak in most of the incoming runoff. This study shows that real-time controls may have potential to adapt to various design parameters and desired outcomes to improve the performance of green infrastructure.
