Journal of Sustainable Water in the Built Environment

Abstract

In an attempt to reduce the impacts of stormwater, stakeholders are exploring ways to improve the performance of existing stormwater control measures adjacent to highways, including vegetated filter strips (VFSs). This study assesses the hydrologic performance of a VFS amended with a vegetated compost blanket (VCB) through evaluation of dynamic flow modification, event volume storage, and cumulative performance. Over 2.25 years, 278 rainfall events were observed at a highway median in Maryland for two VCBs of 7.6-cm depth and 30-m width along the highway, one at 3-m length and one at 6-m length. Supplemental greenhouse mesocosm experiments provided supporting information on stormwater storage and slope impacts. VCBs were found to significantly reduce both stormwater flow and volume. Peak flows were significantly reduced by 39% for the 3-m and 72% for the 6-m VCB in comparison to highway runoff. At the highest flowrates, both VCBs were unable to significantly reduce flow, and instead acted as conveyance. Total influent stormwater volume capture over the entire study period was found to be 44% and 55% for the 3-m and 6-m VCBs, respectively. A useful design estimate for representative storage capacity, using saturated moisture content and wilting point of the media, was determined through use of a hockey-stick fit for inflow-outflow curves. Improvements in VFS soil hydraulic properties and vegetative cover over the course of the experiment were observed, indicating additional mechanisms for hydrologic improvement.

Practical Applications

To reduce and manage excess stormwater, engineers have designed green systems called stormwater control measures to absorb and filter stormwater. One of these systems is a vegetated filter strip (VFS), a thin layer of soil on the sides of highways often planted with grasses. To test whether a layer of compost placed on top would significantly absorb more stormwater, two large vegetated compost blankets (VCBs) were studied on a highway median in Maryland, and small test setups that simulate a highway median topped with a VCB were studied in a greenhouse. It was found that the VCBs lowered the highest peak flows during storms by 39%–72%, with longer flow path VCBs having a better performance. The systems were also able to store 44%–55% of the total volume of stormwater over a 2.25-year study period. An equation was created to estimate the volume of stormwater a VFS with VCB should be able to store; the equation uses the compost and soil volumes and standard compost and soil properties. The VCBs laid over a VFS in this study were effective in reducing the amount of highway stormwater more than a VFS alone.

Nitrate (${\mathrm{NO}}_3^{-}$) pollution from onsite wastewater treatment systems (OWTS) is a concern in coastal communities that rely on groundwater for drinking water because of health risks associated with high levels of ${\mathrm{NO}}_3^{-}$ and the potential for eutrophication in areas where ground and surface water are connected. We examined the relationship between OWTS density and groundwater ${\mathrm{NO}}_3^{-}$ concentration in glacial soils with different parent material in a coastal watershed in the town of Charlestown, Rhode Island (USA). The area is underlain by glacial till and fluvial deposits, with groundwater as the only source of potable water and OWTS as the only means of treatment. The density of OWTS/ha was not significantly different between glacial fluvial ($\text{median}=1.0$; range, 0.04–4.8; $n=207$) and till ($\text{median}=1.3$; range, 0.012–3.81; $n=160$) soil parent material. Nitrate levels (mg $\mathrm{N}/\mathrm{L}$) in shallow well samples taken from 2013 to 2022 were significantly higher in till ($\text{median}=3.7$; range 0–15; $n=160$) than in glacial fluvial ($\text{median}=2.7$, range 0–9.7; $n=207$) parent material. Groundwater ${\mathrm{NO}}_3^{-}$ levels increased linearly with density of OWTS, and the slope (mg N · $\mathrm{ha}/\mathrm{L}$ · OWTS) and intercept (mg $\mathrm{N}/\mathrm{L}$) values for the regression were significantly higher for till (0.99; 2.28) than glacial fluvial (0.66; 1.95) parent material. Regression intercept values $>0.5\mathrm{mg}\mathrm{N}/\mathrm{L}$, corresponding to a density of $0\mathrm{OWTS}/\mathrm{ha}$, point to widespread mixing of contaminated groundwater. Cumulative probability analysis showed that the proportion of samples with ${\mathrm{NO}}_3^{-}$ levels corresponding to Extreme pollution risk ($>5\mathrm{mg}\mathrm{N}/\mathrm{L}$) was much higher in till than in glacial fluvial parent material at the same OWTS densities. Fewer than 10% of the ${\mathrm{NO}}_3^{-}$ values were in the Low risk category ($<0.5\mathrm{mg}\mathrm{N}/\mathrm{L}$) regardless of OWTS density or parent material. Our results suggest that OWTS density and soil parent material should be part of the criteria for water management and land use decisions to protect public and environmental health.

Nitrogen pollution in stormwater runoff is among the most difficult to mitigate via bioretention, particularly for dissolved species like ${\mathrm{NO}}_3^{-}\text{-}{\mathrm{NO}}_2^{-}({\mathrm{NO}}_{\mathrm{x}})$. Bioretention soil mix (BSM) amended with compost often leaches nitrogen, making the system a nitrogen source rather than a sink. Several design modifications have been proposed to enhance nitrogen removal pathways, like microbial denitrification, by creating anoxic zones, but studies report varied removal rates. This study evaluated nitrogen removal in bioretention systems equipped with orifice-outlet controls. While these controls primarily reduce outflow rates, it is expected that nitrogen removal would increase due to extended hydraulic residence time. The study consisted of six field-scale bioretention mesocosms with two outlet configurations (orifice/standard) and three BSM types (NEW, mature, and alternative). Six synthetic storms measured total Kjeldahl nitrogen (TKN) and ${\mathrm{NO}}_{\mathrm{x}}$ removal performance. Continuous flow monitoring characterized the orifice-outlet performance in response to natural storm events, and salt-pulse tracer testing measured the mean residence time. Additionally, soil samples were characterized for TN, TOC, pH, $\mathrm{C}:\mathrm{N}$, ${\mathrm{NH}}_4^{+}$, and ${\mathrm{NO}}_3^{-}$ and analyzed with qPCR for $nirK$ and $nirS$ gene abundance, which are indicative of denitrification activity. Orifice-controlled outlets significantly increased residence times and decreased ${\mathrm{NO}}_{\mathrm{x}}$ effluent concentrations compared to standard outlets. However, they also increased TKN effluent concentrations, resulting in no net benefit from a TN reduction perspective. Nitrogen leaching was most significant in the new BSM, but this leaching gradually reduced even within this study of six storms. The alternative BSM had the lowest effluent concentrations and the best removal performance, but plant growth was severely limited. $NirK$ genes were only detected in the alternative BSM; however, $nirK$ was present in low abundances compared to other published studies, suggesting denitrification played a minor role in nitrogen removal from these mesocosms.