Evaluation of Roadside Filter Strips, Dry Swales, Wet Swales, and Porous Friction Course for Stormwater Treatment

Abstract

Due to NPDES regulations, the North Carolina Department of Transportation (NC DOT) is required to treat stormwater from NC DOT facilities throughout North Carolina. There are hundreds of miles of existing right-of-way swales and filter strips across North Carolina. Relatively few roadside swales and filter strips have been tested for water quantity and quality control. Also, no studies exist on swales with wetland characteristics. This paper presents an assessment of dry swale, wetland swale, and filter strip performance along an interstate highway in North Carolina. Four existing right-of-way linear swales along I-40 were monitored to determine their hydrologic and water quality effectiveness. Two different treatments were examined: one dry swale and one which was allowed to establish wetland vegetation and hydrology. This experimental design was replicated once. Also addressed was the impact of the vegetated filter strip between the shoulder and the edge of the swale. Samples have been collected from 17–21 events (depending on the site) and analyzed for TKN, NO2–3-N, TN, NH4-N, Organic-N, TP, and TSS. It should be noted that this section of highway had a porous friction course (PFC) applied, which had an impact on swale and filter strip performance. Mean effluent TN concentrations were lower for the swales with wetland characteristics than the non-wetland swales. No such difference was observed for TP effluent concentrations. TP concentrations measured at the edge-of-highway were low (mean < 0.11 mg/L) at all four sites, resulting in poor reduction ofTPEMCs by the swales and roadside filter strips. Due to the presence of a porous friction course on the highway, mean TSS concentrations from the roadway were below 32 mg/L at all four sites. Effluent concentrations of TSS from both the swales and filter strips were higher than edge-of-pavement concentrations. The swales and filter strips did not perform well using traditional concentration reduction metrics for TP and TSS; this was mainly due to the lower sediment-bound pollutant concentrations derived from the porous friction course overlay.