Sediment removal from urban runoff using seep berms and vegetative filtration

Abstract

Previous field demonstration projects in metro-Atlanta have shown that seep berms, which are elongated sedimentation basins at the outlet of a disturbed land area, can provide high suspended sediment trap efficiencies with respect to coarse sediments on construction sites having drainage areas greater than five acres. Previous literature has shown that vegetative filter strips are efficient traps for fine suspended sediment in stormwater runoff. A combination of a seep berm and vegetative filter in series was studied in this thesis as an erosion control measure with quantification of its flow resistance and sediment removal efficiency. First, a field demonstration project was implemented to evaluate seep berms as a viable erosion control measure through a side-by-side comparison with the more commonly-used silt fences on construction sites with drainage areas less than five acres in metro Atlanta. High suspended sediment trap efficiencies were recorded for the seep berm on two separate sites, and the seep berm was shown to be superior to silt fences with respect to sediment control in the site runoff. Then a vegetative filter was studied in the laboratory in a specially-built flume for that purpose. The relationship between vegetative drag coefficient and various parameters reflecting flow conditions and vegetation density in steady, uniform open channel flow was studied in the flume. Both rigid, emergent vegetation and submerged, flexible vegetation were studied at two different plant densities. The application of porous media flow concepts to open channel flow through vegetation resulted in a collapse of data for vegetative drag coefficient for the various vegetation types and densities into a single relationship when plotted against vegetative stem Reynolds number. Point velocity and turbulence intensity profiles at different locations in the vegetative filter were recorded with an acoustic Doppler velocimeter to observe the turbulence structure of the flow and its effects on vegetative drag and settling of sediment. A sediment slurry consisting of a suspension of fine sand was fed into the flume, and an automated sampler was used to measure suspended sediment concentrations along the vegetative filter length for a series of discharges from which sediment flux and trap efficiency could be determined. Experimental data for trap efficiency were plotted against a dimensionless settling efficiency for each type of vegetation and density. These relationships, along with the one developed for the coefficient of drag, were applied in a numerical design technique that allows designers to determine the flow depth, velocity and trap efficiency of a vegetative filter of known dimensions for a given flow rate, sediment grain size distribution, slope, and vegetation density. In a typical design example, the combined trap efficiency proved that a seep berm followed by a vegetative filter can be a very effective erosion control measure.