The Performance of Bioretention Areas in North Carolina: A Study of Water Quality, Water Quantity, and Soil Media

Show full item record

Title: The Performance of Bioretention Areas in North Carolina: A Study of Water Quality, Water Quantity, and Soil Media
Author: Sharkey, Lucas John
Advisors: William F. Hunt III, Committee Chair
Robert O. Evans, Committee Member
Deanna L. Osmond, Committee Member
Abstract: Flooding, stream bank erosion, and closures of fisheries have led to increased regulation and the use of Best Management Practices (BMPs) for stormwater management. Stormwater BMPs, such as bioretention areas, are popular among developers because they fulfill both landscape and water quality needs; however, questions on design persist. A total of four bioretention areas were studied in central North Carolina investigating phosphorus (P) and nitrogen (N) removal, soil media, volume reduction, and the effectiveness of including an internal storage (IS) zone. A laboratory study investigated the affect of soil test-P, or P-Index, on phosphate (PO4) outflow concentrations from typical bioretention soil media. Two cells in Greensboro, NC, were continuously monitored for hydrology and water quality from July, 2003, to September, 2004. Both cells were 5% of their 0.202 ha (0.5 ac.) watersheds and contained nominally 1.2 m (48 in.) of fill media underlain by a 15 cm (6 in.) drainage layer of crushed stone. One cell (G-1) contained an IS zone of 0.45 m (18 in.), and the other cell (G-2) was left conventionally drained. The sites were studied to determine the affect of an IS zone on hydrology, concentration reduction, and load reduction. Although lower outflow frequency was found for G-1, and a greater delay of peak flow was experienced, no significant difference (p > 0.05) in outflow was prevalent. G-1 and G-2 reduced total inflow volume by 51% and 48% respectively. The cells treated approximately 60% of all inflow. Because both cells dramatically increased outflow concentrations, load reductions were highly dependent on water volume reduction. G-2, containing a high P-Index fill soil between 86 and 100, increased total phosphorus (TP) load by 39%; whereas, G-1, containing a medium P-Index fill soil between 25 and 50, reduced TP load by 16%. Total phosphorus concentrations leaving G-1 were significantly lower (p = 0.002) than G-2. The TP outflow concentrations were correlated to the soil test-P, and P-Index, of the fill soil. Median nitrate (NO3) concentrations were reduced 78% by G-1 and 35% by G-2; however, total kjeldahl nitrogen (TKN) and ammonium (NH4) concentrations were increased dramatically by both cells. TN was not significantly different (p = 0.38) between the two cells, leading to similar TN load reductions of 43% and 38%, removing 1.86 kg (4.1 lbs) and 1.77 kg (3.9 lbs) for the conventional and IS cells respectively.. Two additional cells constructed with conventional drainage in Louisburg, NC, were monitored for hydrology and water quality from May, 2004, to December, 2004. One cell (L-2) was lined with plastic and the other (L-1) was left un-lined. Both were approximately 4.5% of their watersheds and filled with identical, non-agricultural soils containing a low P-Index (1 to 2), to a media depth of 0.75 m (30 in.). Mean TP and TN concentration reduction for 14 storms measured were 33% and 30% respectively for L-1, and 47% and negative 60% respectively for L-2. Furthermore, L-2 reduced NO3 concentration by 60%; whereas, for L-1 it was only reduced by 14%. The differences between the cells were attributed to L-2 going anaerobic, increasing N reduction, and a low inflow TP concentration entering L-2 producing a misleading increase in TP concentration. Respectively, L-1 and L-2 reduced TN load by 65% and 58%; whereas, TP load was reduced by 69% and 10%. The high percent reduction by L-1 was attributable to the low P content within the fill soil; whereas, L-2 received a lower TP concentration and was saturated, therefore unable to reduce TP at a high rate. Because L-2 was lined, evapotranspiration (ET) could be estimated. For L-2, 19% of the total inflow volume was lost to ET. Given the cells were placed side-by-side, the ET rate from L-2 was used to estimate groundwater recharge for L-1, which amounted to 8% of total inflow to L-1 leaving to groundwater. Volume reduction was significantly different (p = 0.004) between the two cells in Louisburg, confirming that a considerable portion of water was lost to groundwater even though it was placed within a tight clay in-situ soil [saturated hydraulic conductivity 0.1 — 1.2 cm/hr (0.04 - 0.47 in.⁄hr)]. L-1 treated 77% of inflow; whereas, L-2 treated 89% of inflow. The laboratory study compared three soil textures, (A) approximately 90% sand, (B) approximately 85% sand, and (C) approximately 75% sand. Four replicates of each were raised to four P-Index levels ranging between 3 and 120. A concentration of 0.18 mg PO4-P⁄L was applied weekly for ten weeks at a rate of 0.5 L per week to all 48 replicates and was compared to outflow concentrations. Determined by the laboratory study, P-Index had a significant (p = 0.001) effect on outflow concentrations, which increased as P-Index was increased for all three soil textures. Soil texture also had a significant (p = 0.001) effect on outflow concentrations, which increased as percent sand increased. Given the findings from the laboratory and field studies, a low to medium P-Index fill soil (less than 25) is suggested for use in bioretention areas if no concentration increase in phosphorus, and more specifically PO4, is desired. Soil textures ranging between 76% sand and 84% sand are suggested to retain the greatest mass of phosphorus, and to provide a hydraulic conductivity which induces nitrogen removal.
Date: 2006-06-07
Degree: MS
Discipline: Biological and Agricultural Engineering

Files in this item

Files Size Format View
etd.pdf 4.879Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record