Assessing the Effectiveness of Proprietary Stormwater Treatment Devices Matt A. Wilson,1 John S. Gulliver, Omid Mohseni, and Raymond M. Hozalski Dept. of Civil Engineering, University of Minnesota, Minneapolis, MN 55455-0116 1 Corresponding author: PH (612) 624-4613; FAX (612) 624-4398; email: email@example.com Abstract Proprietary underground devices have become a popular means of stormwater treatment in dense urban areas due to tight space constraints. Several technologies are employed to remove sediment and other debris from stormwater runoff prior to discharge into lakes, rivers, and streams, including hydrodynamic separation and filtration. Field monitoring studies have been undertaken which have shown these devices, especially the hydrodynamic separation systems, to be capable of removing a significant amount of solids, or incapable of removing solids contained in surface runoff. The results of field monitoring studies depend upon accurate sampling of the runoff, both upstream and downstream of the device. Obtaining representative samples of suspended sediment at each location is a challenge. This research explores two objectives: 1) the feasibility of controlled field testing, and 2) the performance of each device when subjected to field testing under a variety of treatment discharges, and with a wide range of sediment sizes. Each device is carefully cleaned prior to the testing and then a bulk solids analysis is performed on sediment captured by the device so as to eliminate errors due to sampling of influent and effluent water. After data analysis is complete, a revised sizing criterion is proposed which will improve overall performance and sizing of such devices. The resulting approach, refined through field experiments, will be incorporated into an assessment (monitoring and field testing) protocol that will be used by consultants, local governments, and state agencies to assist in selecting, designing, and evaluating stormwater treatment technologies for public infrastructure improvement projects. Introduction The 1987 Amendments to the Clean Water Act implemented a two-phase program to regulate stormwater discharge. Phase I focused on large municipalities, construction sites and industrial facilities, while Phase II expands these policies to include smaller municipalities, construction sites, and industrial facilities. Stormwater Pollution Prevention Programs will be implemented on facilities owned and/or operated by the state, city, town, county, flood control, watershed district or other similar entity, and can be assisted by the assessment of the water quality performance of stormwater Best Management Practices, such as detention ponds, bioretention, swales, infiltration basins, filter strips, and others. Proprietary underground sedimentation devices provide an alternate approach for removal of trash and sediment from stormwater runoff. They have become a popular means of treatment in dense urban areas due to tight space constraints, or as a pre-treatment to other devices such as ponds and infiltration basins to reduce maintenance costs. Several technologies have entered the marketplace to remove sediment and other debris from stormwater runoff. A number of field monitoring studies have been undertaken to quantify the pollutant removal performance of underground stormwater treatment structures (Waschbusch 1999, Strynchuk et al 2000, England 2001, Yu and Stopinksi 2001, Bonestroo, Rosene, Anderlik and Associates, Inc. 2002 and 2003, ETV 2005a and 2005b, Roseen et al 2005, Fassman 2006). Much of the available research differs in both experimental and evaluation criteria, which contributes to the variation in reported removal data. The task is complicated by the fact that many types of proprietary structures exist, and that underground structures have been installed in a wide variety of watersheds, i.e. with diverse land uses, climate, and geology. Additionally, the fact that monitoring relies on sampling can be problematic for relatively coarse material like sands, unless sampling is performed isokinetically (Othmer and Berger 2002) and throughout the cross-section. As a result, obtaining a representative, unbiased sample at both the inlet and outlet portions of the treatment systems is a challenge (Andoh and Saul 2003). The effect is performance data that shows little consistency between studies, even for a given proprietary underground structure. Research Objectives and Experimental Approach This research investigated the practicality of controlled field testing, as an alternative to field monitoring, as a means for the assessment of underground structures for stormwater treatment. The concept of testing eliminates the dependence on the uncertainties inherent with sampling for performance evaluations. Instead, field testing makes use of simulated runoff, rather than actual storm events, utilizing water and sand that is artificially supplied to a clean device. At the completion of a test, personnel enter the device and remove the sediment retained during the test, allowing for a bulk solids analysis on a known quantity of sand both delivered and retained. In addition to providing a more certain performance assessment, the approach should enable comparison of results for a particular device across different watersheds, climates, & geology (i.e. different pollutant loading), influent flowrates, and size of treatment unit. Some research suggests that scaling factors could be applied to individual parameters influencing performance, i.e. flowrate, particle size, device size, etc., to help correct for these differences (Guo 2005); however, a simpler means of providing this comparison can be accomplished by plotting the removal efficiency as the dependent variable versus the appropriate dimensionless parameter, as explained in the following paragraphs. Prospective sites from throughout the Twin Cities metropolitan area were identified, screened, and evaluated for field testing potential based on a variety of characteristics: 1) location out of vehicle traffic lanes for safety and traffic handling concerns; 2) proximity to a fire hydrant for use as a water source; 3) maximum treatment rate of the BMP device due to finite maximum discharges from hydrants; and 4) device allowing for human access to treatment chamber sump for cleanout activities. The system to be tested also needed to provide a suitable location within the storm drain system for flowrate measurement using a pre-calibrated weir and pressure transducer. Proper level measurement required an avoidance of tailwater effects and the induction of a hydraulic jump a suitable distance upstream of the weir so that flow was subcritical for level measurement. Appropriate permits were obtained from governing agencies. Four sites were chosen for field testing. Prior to the commencement of testing activities, each site required several preparation procedures: a circular weir and pressure transducer were installed within the storm drain system for flowrate measurement; the device was dewatered and cleaned with the assistance of vacuum trucks provided by each city; and a piping system was customized for the delivery of hydrant water used as influent test water. Additionally, sand was sieved into three size fractions for use in each simulated runoff event, with median sizes 107 µm (ranging from 89 µm to 125 µm), 303 µm (ranging from 251 µm to 355 µm), and 545 µm (ranging from 501 µm to 589 µm). The procedure for field testing an underground structure for stormwater treatment includes numerous steps: 1) establishing a safe work zone, following confined space entry regulations; 2) installing an inflatable plug(s) upstream of the BMP device to seal off nuisance and/or extraneous flows present in the storm drain system; 3) dewatering the device with sump pumps and removing solids with a wet/dry vacuum cleaner; 4) establishing an appropriate flowrate through the system using real time level measurements from a pressure transducer and datalogger, and conditioning the flow with a gate valve on the hydrant; 5) introducing 10-15 kg of pre-sieved sand [equal parts of 107 µm, 303 µm, and 545 µm sands] to the influent hydrant water at 200 mg/L using a pre-calibrated sediment feeder; 6) following a settling period, dewatering the device with one or more sump pumps, and collecting the retained solids with a wet/dry vacuum cleaner; and finally 7) oven-drying, sieving, and weighing each mass fraction of solids retained. This data (7), together with the appropriate dimensionless parameter, can be plotted as a performance curve specific to a type of underground structure. Each test produces three data points since three discrete sand size ranges were utilized. Each device was tested under four flowrate conditions in triplicate, so the performance curve for each device is populated by 36 data points. Based upon previous experience with settling of suspended sediment in lakes (Dhamotharan, et al. 1981), the Peclet number (Pe = Vh/D) and a dimensionless time, were shown to have a major influence upon sediment deposition ratio, where Vs is particle settling velocity, h is settling depth, and D is the turbulent diffusion coefficient. Pe is defined as the ratio of advection to diffusion, where advective, settling forces are opposed by turbulent diffusion in the system tending to keep a sand particle in suspension. D scales with a velocity times a length scale, where the length scale is the shortest dimension of the flow. In most cases, the length scales of underground structures (diameter d and settling depth h) are similar, so settling depth h is chosen as the characteristic length for the turbulent diffusion coefficient. Thus D ~ Uh, and by continuity D ~ Qh/A. If the vertical cross sectional area A (the direction of mixing) is taken to be dh, then D ~ Q/d and Pe = Vshd/Q. The selection of settling depth h as the characteristic length for diffusion is somewhat arbitrary, but allows for both of an underground structure’s orthogonal length scales (diameter d and settling distance h) to contribute to Pe, leading to a more descriptive dimensionless parameter. The Pe derivation above compares favorably to the dimensionless time variable, T = trVs/h, where tr = V/Q is the residence time and V is the tank volume, that also influences sediment deposition ratio. Thus, T = d2hVs/Qh = d2Vs/Q. Results At high Pe, i.e. large particles and therefore high settling velocities, Vs, coupled with low flowrates, Q, a stormwater treatment device can be expected to be successful removing particles from an influent. Conversely, at low Pe, i.e. small particles and therefore low settling velocities, Vs, coupled with high flowrates, Q, a device can be expected to remove particles from influent with less success. This has been upheld in the results obtained, illustrated by the sample performance curve depicted in Figure 1. Conclusion Understanding how devices perform under varying flowrates, sediment sizes, and size of treatment chamber is important and helpful for consultants, local governments, and state agencies when selecting, designing, and evaluating stormwater treatment technologies for public infrastructure improvement projects. However, the effectiveness of proprietary underground stormwater treatment devices depends upon the settling velocity of influent solids, i.e., solid size and density, in addition to the size and design of the device. To predict performance and to size devices, a suspended solid size distribution of typical runoff from the watershed is needed. We have developed a robust and accurate testing protocol for underground devices that determines performance based upon the solid size distribution and density of the influent, in addition to the water discharge and temperature. It has been successfully verified on four devices in field tests, and other devices in laboratory tests. The next, more challenging, goal is to develop a simple method of determining this size distribution of solids in stormwater runoff. Acknowledgment This project was funded by the Minnesota Local Road Research Board and the Metropolitan Council. References Andoh, R.Y.G. and A.J. Saul (2003). The use of hydrodynamic vortex separators and screening systems to improve water quality. Water Science and Technology, vol 47, no. 4, pp 175-183. Bonestroo, Rosene, Anderlik and Associates, Inc. (2002). Walker Avenue V2B1 Performance. March 19. Bonestroo, Rosene, Anderlik and Associates, Inc. (2003). Walker Avenue V2B1 2001-2002 Performance Assessment. April 15. Dhamotharan, S., et al (1981). Unsteady One-Dimensional Settling of Suspended Sediment. Water Resources Research Vol 17, No 4, p 1125-1132. England, G. (2001). Success stories of Brevard County, Florida stormwater utility. Journal of Water Resources Planning and Management, vol 127, no. 3, May/June. Environmental Technology Verification (ETV). (2005a). ETV Report: BaySaver Separation System, model 10k. 05/21/WQPC-WWF, EPA/600/R-05/113, September. Environmental Technology Verification (ETV). (2005b). ETV Report: Vortechs System, model 1000. 05/24/WQPC-WWF, EPA/600/R-05/140, September. Fassman, E.A. (2006). Improving effectiveness and evaluation techniques of stormwater best management practices. Journal of Environmental Science And Health Part. 41: 1247-1256. Guo, Q. (2005). Development of adjustment and scaling factors for measured suspended solids removal performance of stormwater hydrodynamic treatment devices. Proceedings of the World Water and Environmental Resources Congress 2005. Anchorage, AL, May 15-19. Nnadi, F.N., et al. (2005). Side-by-side evaluation of stormwater proprietary BMPs. Proceedings of the World Water and Environmental Resources Congress 2005. Anchorage, Alaska, May 15-19. Othmer, E.F. and B.J. Berger. (2002). Future Monitoring Strategies with Lessons Learned on Collecting Representative Samples. Storm Water Program, CSUS Office of Water Programs. Roseen, R.M., et al. (2005). Normalized technology verification of structural BMPs, Low Impact Development (LID) designs, and manufactured BMPs. Proceedings of the Watershed Management Symposium. Williamsbug, VA. Schwarz, T. and S. Wells. (1999). Continuous deflective separation of stormwater particulates. Advances in Filtration and Separation Technology.Vol 12, pp 219-226. Strynchuk, J., et al. (2000). Continuous deflective separation (CDS) unit for sediment control in Brevard County, Florida. Proceedings of the Watershed Management Symposium. Fort Collins, CO. Waschbusch, R.J. (1999). Evaluation of the effectiveness of an urban stormwater treatment unit in Madison, Wisconsin, 1996-97. U.S.G.S. Water-Resources Investigations Report 99-4195. Middleton, Wisconsin. Yu, S.L. and M.D.Stopinski. (2001). Testing of ultra-urban stormwater best management practices. VTRC 01-R7, Virginia Transportation Research Council: Charlottesville, VA, 1-43. Figures SAMPLE PERFORMANCE CURVE 100% 90% 80% Removal Efficiency 70% 60% 50% 40% 30% 20% 10% 0% 0.0 1.0 2.0 3.0 4.0 5.0 6.0 Pe = L1*L2*Vs/Q Figure 1. Typical performance curve for a proprietary underground structure. Twelve tests were performed on each device (four different flowrates in triplicate), and each test assessed performance for three sand sizes, for a total of 36 data points. Better removal has been shown for high Pe, and conversely less effective removal for low Pe.