Constructed Wetlands Storm Water Quality Treatment

EA developed a cost effective pollutant removal strategy which also converted a portion of the 30-m (100-ft) critical area buffer to a forested wetland filter strip. The strategy consisted of a wetland perimeter swale and berm system which collects surface water runoff and, through a series of level spreaders, directs the decanting storm water through a 30-m (100-ft) wide perimeter filter strip planted with wetland trees and shrubs. In order to establish the swale, berm and buffer area, it was necessary to remove areas of impervious paving in the buffer. Portions of the buffer area were made-land. Soil samples were collected and tested. The use of crusher run and limestone dust as a fill material resulted in relatively high soil pH values with a high buffering capacity. The removal of paved areas, inappropriate fill texture, and high soil pH values necessitated the import of topsoil on portions of the site. Plant species were selected on the basis of the soil test results and a set of functional attributes relating to wildlife habitat values. Evergreen species were selected to provide winter cover, species producing hard and soft mast were included, and other species were selected for their limb dropping and cavity forming propensities. Through the selection of eight canopy and six shrub species, this planting is expected to provide a variety of habitat for urban wildlife and water dependant species from the adjacent waterway. During the construction of the project, site supervision and monitoring during a heavy precipitation event demonstrated that the design is effective. The design has resulted in phosphorus removal in excess of the 10% required by the City Critical Area Program. In addition, birds have been observed flying into the canopy of the trees even before the planting was completed. This project was completed mid-October 1994 and water quality monitoring efforts proposed in conducted in 1997 indicate sediment and nutrient removal efficiencies higher than predicted by the Critical Area loading model.