Abstract
Concerns regarding non-point source pollution have been steadily increasing over recent decades around the world. Sudden discharges of stormwater runoff from paved surfaces, which cannot soak into the ground, lead to flooding, and water pollution into aquatic ecosystems. Green infrastructure (GI) systems have been employed as an environmentally sustainable alternative to treat stormwater in urban areas since the late 1990s in the hope of mitigating impervious surface hydrology effects. GI could protect aquatic ecosystems from stormwater quantity and quality pollution due to land development and human activities. There have been many short-term studies that document the performance of individual GI systems and their ability to protect waterways from the detrimental effects of urban stormwater. However, few studies exist comparing the long-term performance of different types of GI designs in the same location, and the removal mechanism of salt by GI vegetation is poorly understood. The potential of different vegetation types for mitigating deicing salt has not been documented significantly, and little is known about the effects of maintenance on GI performance.
This dissertation uses the flow and composite flow-weighted sampling to compare water quality improvements of four typical GI practices (bioretention, grass channel, compost-amended grass channel, and bioswale) along a road in Lorton, Fairfax County, Virginia. The three objectives of this dissertation include: (1) compare long-term monitoring of the water-quality performance of four GI practices receiving similar influent pollutant loadings; (2) explore the potential of different vegetation types to reduce deicing salt released from a bioretention by transpiration; (3) evaluate the economic and environmental maintenance costs and benefits of GI based on water-quality performance.
For the first part of this dissertation, approximately 60 storm events have been monitored and sampled since 2018, and 24 relatively complete storms are selected for comparison and evaluation. The performance of these four monitored GI designs ranges significantly. Grass channel performed best on both runoff and pollutant load reductions, serving a relatively small contributing drainage area, while the bioretention, which has the second highest volume and mass load reduction for pollutants, serves the biggest contributing drainage area, and its performance is considerably more consistent than most of the other types of GI systems monitored.
The second part of this dissertation investigates the attenuated transport of salts by vegetation in the bioretention basin (BR). BR is a typical GI system wherein stormwater runoff is routed to a soil basin planted with vegetation and has been shown to reduce deicing salt loads in surface runoff, but the removal mechanism of salt is poorly understood. This work explores the potential of different vegetation types to reduce deicing salt released from a BR by transpiration. Six engineered soil media columns were built in a laboratory greenhouse to simulate a 1012 m2 BR basin along Lorton Road, Fairfax County, Virginia. The effect of vegetation types (Blue Wild Indigo and Broadleaf Cattail) and influent salt concentration on flow volume and salt mass reduction were quantified for multiple storm events. For all storm events, inflow concentrations, and vegetation types, Cl− load reduction ranged from 26.1% to 33.5%, Na+ load reduction ranged from 38.2% to 47.4%, and volume reductions ranged from 11.4% to 41.9%. Different inflow salt concentrations yielded different removal rates of deicing salt, and for a given column, salt removal decreased over sequential storm events. For each influent salt concentration, columns planted with Broadleaf Cattail (BC) performed better for volume and salt mass reductions than columns planted with Blue Wild Indigo (BWI), which in turn performed better than the controls.
The third part of this dissertation addresses the effects of long-term maintenance on GI performance. Seven maintenance events with a forebay restoration are monitored from 2018 to 2022 to evaluate the efficiency of maintenance activities. Stormwater quality performance among these four GI practices before and after the seven in-field maintenance activities was assessed according to pollutant load on dissolved organic carbon (DOC), total dissolved nitrogen (TDN), total suspended solids (TSS), and runoff reductions for the 14 storm events over four years. The average runoff reductions over all monitored storms were 74%, 85%,63%, and 68% for bioretention, grass channel, compost-amended grass channel, and bioswale, respectively. Over the seven maintenance events, the mean runoff reduction of all monitored GI designs improved by 3% after maintenance, DOC mass load reduction increased by 41%, TDN mass load reduction improved by 25%, and TSS mass load reduction increased by 2%. All the pollutant and runoff reduction for swales improved after spring maintenance events, and all monitored GI systems performed better after spring maintenance events compared with their performance after fall maintenance activities, which is potentially due more to the vegetation growth than the maintenance work.
This dissertation will help researchers, engineers, and policymakers better understand how GI practices are functioning and identify ways to optimize their effectiveness. It can also help determine the capacity of different types of GI systems to store and treat stormwater, track changes in water quality over time, and evaluate the effectiveness of their maintenance activities.
