Abstract
The capacity of extensive green roofs (EGRs) in runoff reduction and pollutants control significantly decreased with increasing rainfall intensity. This is attributed to most of EGRs are the lower-drainage method and use a thinner substrate layer because of the structural load-bearing limits of buildings. Existing solutions depend on increasing the rainwater storage by thickening the substrate layer of EGRs, or replacing the substrate materials. We propose to enhance the water retention capacity of EGRs by adjusting vertical drainage positions and moving the lower-drainage at the bottom to upper-drainage over the substrate layer. To test the runoff control performance of EGRs with the upper-drainage method, we designed three EGR models with two different (upper- and lower-) drainage methods based on the common EGR's structure in practice and carried out total 48 experiments under different rainfall intensities with an artificial rainfall simulator. The experimental design also included adjusting structural factors of the EGR to further explore the influences of roof slope, status of vegetation growth (height and canopy cover ratio), and model size. The average runoff retention rate for the upper-drainage EGR was 61.82 %, which was much higher than the 23.94 % for the lower-drainage EGR. The concentrations of TN and TP in runoff with the upper-drainage were 2.83 and 0.18 mg/L, which were much lower than the 8.47 and 2.10 mg/L for the lower-drainage EGR. EGRs with upper-drainage overall performed better in runoff control than EGRs with lower-drainage. Vegetation height and model size did not significantly affect the rainwater control performance for different drainage methods, while the change of slope affected the rainwater interception of the EGRs with upper-drainage method.