Presentation on theme: "An Evaluation of BMP Removal Efficiency in the Kensico Reservoir Watershed New York City Department of Environmental Protection 16-17 September 2008 Prepared."— Presentation transcript:
An Evaluation of BMP Removal Efficiency in the Kensico Reservoir Watershed New York City Department of Environmental Protection 16-17 September 2008 Prepared by Prepared for 2008 NYWEA Watershed Science and Technical Conference and
1 Presenters: Paul Muessig, Michael Powell, and Jon Trombino EA Science and Technology 3 Washington Center Newburgh, New York John Canning New York City Department of Environmental Protection Division of Drinking Water Quality Sutton Park Office 465 Columbus Avenue Valhalla, New York
2 Introduction ● Acknowledgements – NYCDEP staff ● 45 BMP structures in Kensico Reservoir watershed to improve and maintain water quality ● Monitoring Plan and Quality Assurance Project Plan ● 5 BMPs selected for intensive monitoring ● 4 on perennial tributaries to Kensico Reservoir ● NYCDEP Kensico Laboratory processed all samples (turbidity, total suspended solids [TSS], total phosphorus, and fecal coliform)
3 Tasks ● Monitoring a minimum of 10 storm events at 5 selected BMPs, 2000-2007 ● Precipitation, hydrologic, and water quality data compiled to Microsoft ACCESS ● Formatted and submitted to EPA/ASCE International BMP Database ● Evaluate removal efficiency of BMPs
25 Field Methods ● ISCO 6700 Autosampler – timed and flow- weighted grab samples over duration of storm event at inlet and outlet weirs ● Continuous stage and discharge recording ● Precipitation record at Kensico Reservoir ● Hydrology staff selected 10 samples for analysis—4 each on rising and falling limb of hydrograph and 2 around peak
26 Statistical Methods ● Paired comparisons between influent and effluent concentrations ● Efficiency ratios ● Comparison of influent and effluent probability curves
27 Paired Comparisons of Influent and Effluent Event Mean Concentrations BMPN Method of Comparison Significant Reduction? BMP 12- Retention Pond TSS15t-testYes Turbidity17t-testYes Total phosphorus 15t-testYes Fecal coliform 16Wilcox Signed Rank Yes BMP 13- Wetland Meadow TSS13t-test on log- transforms No Turbidity13t-testYes Total phosphorus 13t-testYes Fecal coliform 13t-test on log- transforms Yes BMPN Method of Comparison Significant Reduction? BMP 37 Retention Pond TSS18Wilcox Signed Rank Yes Turbidity18Wilcox Signed Rank Yes Total phosphorus 17t-test on log- transforms Yes Fecal coliform 17t-testYes BMP 57 Sand Filter TSS10t-testYes Turbidity10t-test on log- transforms Yes Total phosphorus 8t-testYes Fecal coliform 10t-test on log- transforms No
29 BMP 12 Retention Pond Effluent Probability Plot for TSS Cumulative Frequency Concentration (mg/L) Good removal over influent concentrations >30 mg/L. Estimated irreducible concentration is 2 mg/L.
30 BMP 37 – Retention Pond Effluent Probability Plot for TSS Cumulative Frequency Concentration (mg/L) Good removal with influent >10 mg/L. Irreducible Concentration = 10 mg/L
31 BMP 57 - Sand Filter Effluent Probability Plot for TSS Cumulative Frequency Concentration (mg/L) Good removal over range of influent concentrations.
32 BMP 12 - Retention Pond Effluent Probability Plot for Fecal Coliform Cumulative Frequency Concentration (cfu/100mL) For small storm events, BMP is actually a contributor. Some removal with influent concentrations >1,000 cfu/100mL.
33 BMP 13 – Wet Meadow Effluent Probability Plot for Fecal Coliform Cumulative Frequency Concentration (cfu/100mL) Consistent removal ( 50%) over range of influent concentrations.
34 BMP 37 – Retention Pond Effluent Probability Plot for Fecal Coliform Cumulative Frequency Concentration (cfu/100mL) Removal with influent >20,000 cfu/100mL.
35 Conclusions ● TSS Consistent 70-80% removal (with exception of BMP 13 – Wet Meadow) Removal of turbidity and total phosphorus is similar to TSS ● Fecal Coliform Sand filter is ineffective Retention ponds and wetland meadow remove 50% at high influent concentrations Residual in retention ponds and wetland meadow may contribute at low influent concentrations