Strategic Flow Deflection: A Cost-Effective Solution for Stormwater Phosphorus Control

The Environmental Protection Agency (EPA), Region 1 is finalizing a new phase of the National Pollutant Discharge Elimination System (NPDES) Permit for separated stormwater systems owned by municipalities discharging into the Charles River, Massachusetts. The annual total phosphorus (TP) loading from these systems must be greatly reduced per the Total Maximum Daily Load (TMDL) Report for Nutrients in the Lower Charles River Basin by EPA and the Massachusetts Department of Environmental Protection (MassDEP) (MassDEP and USEPA, 2007). These requirements will pose a huge technical and financial challenge to the affected communities, especially in highly urbanized areas.In order to develop a TP control scheme that would meet the new TMDL limits, an analysis was performed to characterize the particulate portion of stormwater phosphorus by particle size. Surface runoff and pipe flow samples were collected in a recently separated area in East Cambridge during two storm events. Samples were sieved or filtered using 250,106, 45, and 25-μm sieves and 10- and 0.45μm filters. The screened or filtered sub-samples were then analyzed for total and dissolved phosphorus as well as total solids. Statistical analyses confirmed that, during these storms, approximately 75% of the TP in pipe flow was attached to solid particles between 25 and 45 microns, while the dissolved portion accounted for less than 10% on average.Subsequently, hydrographs generated by synthetic storms in a typical rainfall year were modeled and fluid shear stress profiles were calculated at strategic locations within the catchment. Annual stormwater phosphorus loadings were estimated based on the fluid shear stress necessary to suspend and convey particles of different size. Storms were divided in three phases with regard to phosphorus loading: dissolved-only phase, first-flush phase and non-first flush phase. The dissolved phase occurred in the early and late stages of the storm when the fluid shear stress remained below the 25-micron suspension threshold. The first-flush phase started when this suspension threshold was reached and ended when a higher shear stress value that would ensure all 25 to 45 micron particles had been transported was attained. The non-first flush phase followed the first-flush phase and ended when shear stress fell below the 25-micron suspension threshold again. Average TP concentrations were assigned to each phase based on sampling results and literature values.The calculated TP loadings and distributions were used to design a strategic, passive flow deflection system which maximized the TP removal with minimum flow deflection. With this system, deflected flows can either be conveyed to an adjacent sanitary/combined sewer or to an off-line stormwater treatment system. The deflection system consists of a receiving chamber with a passive, flow-regulating valve at the outlet allowing deflection of small, highly polluted flows in the beginning of the storm (first flush). When larger, cleaner flows reach the deflection point (non-first flush phase), the valve completely shuts and flows follow its normal path towards the outfall.A detailed comparison of available phosphorus removal technologies indicated that flow deflection is the only feasible solution in this highly urbanized area in order to meet the TP requirements in the short to mid-term timeframe. A 20-year life-cost analysis identified this alternative as the least expensive, with the least operation and maintenance requirements, and with the highest probability to meet the new phosphorus loading target. This technology is easily transferable and is an effective, inexpensive way to help meet the TMDL requirements in urban communities.