Zamarripa, Sonia C. (2017-08). Optimization of Bioretention Hydrological Performance and Nitrate Reduction in Urban Stormwater Runoff. Master's Thesis.
Thesis
A bioretention cell is a low-impact development practice that reduces urban stormwater runoff and improves water quality. The bioretention cell located at the Texas A&M AgriLife Research and Extension Center in Dallas, Texas was used for this study. The bioretention cell collects water from an adjacent parking lot and provides detention and filtration through vegetation, engineered media, a gravel layer, and an internal water storage zone. Due to equipment malfunction, out of 33 events from 2013 to 2015, only 22 events were used for the hydrological analysis, 19 for the nitrate analysis, and 10 for both hydrological and nitrate analysis. These evaluations showed that the bioretention cell reduced runoff by an average of 76%. The nitrate analysis yielded a statistically significant nitrate mean reduction of 26% with a p-value of 0.001. For the 10 events that had both qualitative and quantitative data, the nitrate reduction was first considered using a concentration reduction equation; the result was 39% nitrate concentration reduction. When considering the impact of volume reduction, the mean mass reduction equation resulted in 81% reduction. The variation between the nitrate concentration reduction (39%) and mean mass reduction (81%) showed that a majority of the reduction in nitrate was attributed to the reduction in runoff volume, indicating that by optimizing volume reduction, nitrate reduction would also improve. Using a 1-D Hydrus model, the study of hydrological reduction performance for a bioretention cell in Dallas, Texas, resulted in a few optimization potentials. Only three events had complete datasets with sufficient runoff volume reduction to compare the actual measured outflow against the simulated outflow. Two storm events (September 2, 2013 and September 28, 2013) were used for calibration purposes, and one event (September 21, 2013) was used to validate the model. A sensitivity analysis using Hydrus resulted in the following discoveries about the soil properties. Higher water volume attenuations can be achieved with smaller residual water content, larger saturated water content, larger inverse of the bubbling pressure (empirical value), and larger pore distributions (empirical value). The Hydrus sensitivity analysis also showed that increasing exfiltration into the native soil would result in higher reductions of runoff volume due to dryer media coinciding with smaller outflows. During the calibration, the initial water content values of the soil were estimated by trial and error to obtain the correct outflow volume. The two calibrated events were studied to discover that the initial water content conditions could be obtained since both storms required 55 cm of pressure to produce outflow and the pressure remained at 55 cm immediately after the storm. The results of the validated storm showed that Hydrus can simulate the water flowing through a bioretention cell very well, yielding an NSE value of 0.82 and an R^2 value of 0.92 when compared to the measured outflow data and also that the initial water content conditions can be determined without field data. A limitation of using Hydrus-1D is the inability to have two lower boundary conditions to simulate flow into the native soil. This limitation restricts the use of the model to single event simulations instead of longer continuous event simulations.
A bioretention cell is a low-impact development practice that reduces urban stormwater runoff and improves water quality. The bioretention cell located at the Texas A&M AgriLife Research and Extension Center in Dallas, Texas was used for this study. The bioretention cell collects water from an adjacent parking lot and provides detention and filtration through vegetation, engineered media, a gravel layer, and an internal water storage zone. Due to equipment malfunction, out of 33 events from 2013 to 2015, only 22 events were used for the hydrological analysis, 19 for the nitrate analysis, and 10 for both hydrological and nitrate analysis. These evaluations showed that the bioretention cell reduced runoff by an average of 76%. The nitrate analysis yielded a statistically significant nitrate mean reduction of 26% with a p-value of 0.001. For the 10 events that had both qualitative and quantitative data, the nitrate reduction was first considered using a concentration reduction equation; the result was 39% nitrate concentration reduction. When considering the impact of volume reduction, the mean mass reduction equation resulted in 81% reduction. The variation between the nitrate concentration reduction (39%) and mean mass reduction (81%) showed that a majority of the reduction in nitrate was attributed to the reduction in runoff volume, indicating that by optimizing volume reduction, nitrate reduction would also improve.
Using a 1-D Hydrus model, the study of hydrological reduction performance for a bioretention cell in Dallas, Texas, resulted in a few optimization potentials. Only three events had complete datasets with sufficient runoff volume reduction to compare the actual measured outflow against the simulated outflow. Two storm events (September 2, 2013 and September 28, 2013) were used for calibration purposes, and one event (September 21, 2013) was used to validate the model. A sensitivity analysis using Hydrus resulted in the following discoveries about the soil properties. Higher water volume attenuations can be achieved with smaller residual water content, larger saturated water content, larger inverse of the bubbling pressure (empirical value), and larger pore distributions (empirical value). The Hydrus sensitivity analysis also showed that increasing exfiltration into the native soil would result in higher reductions of runoff volume due to dryer media coinciding with smaller outflows. During the calibration, the initial water content values of the soil were estimated by trial and error to obtain the correct outflow volume. The two calibrated events were studied to discover that the initial water content conditions could be obtained since both storms required 55 cm of pressure to produce outflow and the pressure remained at 55 cm immediately after the storm. The results of the validated storm showed that Hydrus can simulate the water flowing through a bioretention cell very well, yielding an NSE value of 0.82 and an R^2 value of 0.92 when compared to the measured outflow data and also that the initial water content conditions can be determined without field data. A limitation of using Hydrus-1D is the inability to have two lower boundary conditions to simulate flow into the native soil. This limitation restricts the use of the model to single event simulations instead of longer continuous event simulations.