Turbidity in Water Supply Reservoir due to Resuspension

1. Water Quality New York City Department of Environmental Protection Bureau of Water Supply Resuspension as a Source of Turbidity in a Water Supply ReservoirEmmet M. Owens, Rakesh K. Gelda, Steven W. EfflerUpstate Freshwater Institute, Syracuse NYDonald C. PiersonNew York City Dept. of Environmental Protection, Kingston NYWatershed Science & Technical ConferenceNew York Water Environment Assoc.West Point, NYSeptember 2009

2. Schoharie Reservoir Diverts water from Schoharie Cr. (Mohawk R. basin) into Shandaken Tunnel, Esopus Cr., Ashokan Res., and Catskill Aqueduct

3. Dam Schoharie Reservoir • long, narrow shape; steep bottom slopes • deep, thermally stratified • short residence time; function is primarily diversion • large watershed; 9/17/99 (Hurricane Floyd) reservoir rose 9.8 m (32 ft.) in 24 hrs • episodes of elevated turbidity driven by runoff events, exacerbated by reservoir drawdown 30 m Manor Kill 0 0.5 1 Scale in km 15 m Intake to Shandaken Tunnel Bear Kill Schoharie Creek

4. Modeling Goals • understand factors leading to historical turbidity events • contribute to design features of potential structural turbidity control alternatives • allow evaluation of turbidity control alternatives: structural and operational

5. Monitoring Program • stream inflows: USGS • reservoir outflows & operations: NYCDEP • local meteorology: NYCDEP • routine temperature and turbidity monitoring: tributaries, water column and withdrawal: NYCDEP • event-based monitoring: Schoharie Creek (robotic); water column (robotic and manual gridding): UFI (Sept. 2002 – Dec. 2005).

6. Historic Reservoir Drawdown • Median annual drawdown = 17 m (56 ft.) • 2002-05 monitoring period: 2 full reservoir years, 2 with significant drawdown

7. Schoharie Reservoir Turbidity Model • state variable is turbidity Tn (an optical property) • while there is no conservation principle for Tn, it is treated as if it is mass (good empirical evidence for doing so) • turbidity model considers following processes: turbidity loading, deposition, transport, export, and resuspension

8. Model Framework: CE-QUAL-W2 (W2) • two-dimensional approach assumes that temperature and turbidity are uniform over width of the basin • hydrothermal component of model previously applied by UFI • model enhanced by UFI to simulate turbidity and resuspension (W2Tn)

9. Early Model Testing • assumed that stream loading is the only source of particles and turbidity • resulted in underprediction of observed Tn in water column and withdrawal during certain runoff events • underprediction was greater during periods of reservoir drawdown

10. Resuspension Relationship Focus on field measurements to validate model predictions of shear stress

11. Two Sources of MotionCausing Resuspension • Stream Inflow – high current velocity near mouth of Schoharie Creek during runoff events • Waves – oscillatory motion associated with wind-driven surface waves

12. Resuspension Due to Stream Inflow A Full Reservoir, Low Streamflow: Large A, Small Q Small V (Deposition) Streamflow (Q) Velocity (V) = Area (A) Reservoir cross sections near creek mouth under two conditions Drawdown A Drawdown, High Streamflow: Small A, Large Q Large V (Resuspension)

13. Shear Stress relationship t = rgV2/CB2 g = acceleration of gravity

14. Resuspension Zones • Resuspension in inflow region • due to Schoharie Cr runoff events Intake Schoharie Cr. Bear Kill

15. T-RDI 1200 KHz Workhorse Monitor ADCP HydrodynamicMonitoringE.A. Cowen, Cornell Univ. Nortek Vector ADV Temperature loggers Aug. – Sept. 2004

16. Observed and Predicted Bed Stress

17. Wave-Induced Resuspension • fetch < 1500 m; wave heights < 30 cm (small) • due to small waves, wave-induced bed stress and resuspension occur where depth < 1 m (narrow strip along lee shore) • effect of drawdown: sediments that are in a depositional environment at full reservoir are exposed to resuspension during drawdown

18. Resuspension Zones • Resuspension in inflow region • due to Schoharie Cr runoff events 2. Wave Resuspension at Shoreline (SW, W, NW winds dominant) Intake Schoharie Cr. Bear Kill

19. October 2001 Severe Drawdown (19 meters; 62 ft.) Gatehouse and Intake Structure

20. Surface Wave Model • Donelan/GLERL model used to simulate waves and associated bottom motion and bed stress • Previously applied to coastal ocean, large estuaries, Great Lakes; first application to small lake or reservoir1 • measurements of wave height and period made with submerged pressure sensors were used to validate the model 1Owens, E.M. 2009. Observation and simulation of surface waves in two water supply reservoirs. Jour. of Hydraulic Engr. 135(8): 663-670.

21. Surface WaveModelValidationOct.-Nov. 2002

22. Drawdown Conditions – 2002 Example simulations follow

23. Model Performance: 13-16 Oct 2002 Green: inflow resuspension Red: no resuspension White: all resuspension

24. Probability that Withdrawal (Tunnel) Turbidity is less than X (days Tn > 10 NTU) Sept 2002 - Dec 2005

25. Conclusions • tributary input is generally the dominant source of turbidity to Schoharie Reservoir • resuspension near creek mouth caused by runoff events can be an important contributing source, particularly during drawdown • wave-driven resuspension is source to surface waters, and is a minor contributing source of turbidity • turbidity model accurately represents these two resuspension processes