Optimization design of key parameters for bioretention cells with mixed filter media via HYDRUS-1D model and regression analysis

The structural design of bioretention cells (BRCs) is important for stormwater flood control and pollutant removal. How to determine the key parameters and their values that affect the operation effect of facilities is a key problem to be solved. Synthetic storm experiments were monitored and analyzed to determine these regulation effects. The BRC with filler of BSM (Bioretention soil media) +10% fly ash showed excellent water reduction rate and event mean concentration (EMC) removal rate. BSM was the mixture of sand, soil and wood chips, and the ratio was 65:30:5 by mass. Based on the experimental data, the sensitivity of model parameters in Hydrus-1D was defined by using the Morris classification screening method. The key hydrological and water quality parameters were calibrated and validated with a determination coefficient R-2 above 0.6 and a Nash-Sutcliffe efficiency coefficient NSE above 0.5. The key structure parameters in BRCs were optimized by using the calibrated HYDRUS-1D model and the response surface methodology (RSM). Results show that the BRC with the filler of BSM + 10% fly ash has the smallest optimized thickness of the filler layer, while the BRC with the filler of BSM + 10% green zeolite is the largest, and the difference between the two media combination is 16-26 cm. The design return period and the confluence ratio affect the design water volume jointly. Under the three-year design return period, the confluence ratios of five kinds of filling combinations (BSM + 10% WTR, BSM + 10% fly ash, soil, BSM, BSM + 10% green zeolite) are recommended to be <= 20:1, 20:1, 15:1, 15:1 and 10:1 respectively