Variations on a Theme, Groundwater Sensitivity

Transcription

1 Variations on a Theme, Groundwater Sensitivity Amy Ritter, Waterborne Environmental, Inc. Mark Cheplick, Waterborne Environmental, Inc. Isha Khanijo, Waterborne Environmental, Inc. American Chemical Society, Philadelphia August 25 th,

2 Overview PWC (GW) conceptual model Background Sensitivity of the interaction of e-fate parameters and PWC GW model assumptions DT50 interactions Application type Layer thickness (dispersion) Volatilization PRZM5 Freundlich implementation (non-linear sorption) and a comparison to FOCUS WINPRZM (based on PELMO/PEARL implementation) Application timing (rainfall timing) Conclusions 2

3 USEPA Drinking Water Assessment Current EFED model in PWC (Version 1.52, PRZM5) Six standard scenarios run for Tier-1 evaluation WI Corn Sands, Delmarva, GA Coastal, NC Coastal, FL Potatoes, and FL citrus Post breakthrough average and maximum concentration from 30 year run (repeated applications and 30-year climate) used for chronic and acute assessment, respectively Previous EFED models for GW Sci-Grow (Tier I screening model) PRZM-GW (version 1.07, PRZM ) 3

4 PRZM-GW/PWC Conceptual Model A drinking water well beneath an agricultural field which draws water from a shallow unconfined, high water-table aquifer with a well screen. Source: 4

5 PWC (GW) Conceptual Model Processes include water flow, chemical degradation and sorption, rain, irrigation and crop specific factors including transpiration, pesticide interception and management practices. Soil temperature simulated Slower aerobic soil degradation with depth up to 1m Only Hydrolysis below 1 m Compartment size 50cm from 1m 8m 0-1 m compartmentalization and degradation factors 5 Source:

6 Background Pesticide with < 1 day aerobic soil half-life and no hydrolysis was run using EPA s groundwater model with six standard scenarios with maximum label use rate WI sand was worst case scenario High groundwater concentrations were predicted, when in field studies chemical was not seen below 12 inches. Sensitivity runs were done for WI corn scenario Figure out what's causing high concentrations To see impact of modeling assumptions used in conceptual model 6

7 WI Corn Scenario (Selected for GW Sensitivity) Located in Central Sands region in South Central Wisconsin Corn for grain was the second largest crop by area for the state of Wisconsin Vulnerable groundwater that may be used by private wells Texture for the Coloma series is Loamy sand, and is classified as hydrologic soil group B (60 to 80% sand), organic carbon from.07% to 0. 46% (surface) Meteorological File La Crosse, WI (W14920) Curve number = 10 Crop emergence on May 1 and harvest on October 20 Core depth = 1000 cm (10 m) 7

8 Five Case Studies for Sensitivity Case Study 1 Subsurface Degradation Parameterization Case Study 2 Application Type Case Study 3 Compartment Thickness Case Study 4 Degradation including Volatilization and Non-linear Sorption Case Study 5 Rainfall on Application Day Variety of Chemical Properties used in Case Studies Parameter Type of Value Value Koc (L/kg) Small 10 Hydrolysis (days) Stable, Medium 0.0, Soil Half-life (days) Fast, Medium, Long 1.0, 10.0, Volatilization On/Off Hk= Freundlich, 1/n On/Off 0.9 Example Application Timing 1 pre-emergent application of 1 kg/ha on April 15 8

9 Case Study 1 Subsurface Degradation Parameterization Degradation properties: 1-day aerobic soil degradation Stable hydrolysis 9

10 Impact of Soil DT 50 Simulation in the Model (WI Corn Scenario) 2m 1m (1) (2) (3) (4) (5) 9m Water Table 1 m Well Screen Modeled Concentration (1) PWC (linearly declining degradation in the top 1m, no hydrolysis) (2) Constant degradation in top 1m (3) Linearly declining degradation in the top 2m (4) Decay rate for cm extended for next 1m (5) Linearly declining degradation in the top 1m, 1000-d half life simulated below 1m 10 Figure not to scale

11 Impact of Soil DT 50 Degradation in the Model Normalized concentrations (ppb) compared for different methods of simulating soil DT50 in PWC model Acute Chronic Standard WI Corn Scenario Constant degradation in top 1m Linearly declining degradation in the top 2m Decay rate for cm extended for next 1m 1000-d half life simulated below 1m Soil Half-life = 1 day, Hydrolysis = 0 (stable), Koc =10 L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 11

12 Concentration, ppb Impact of Soil DT 50 Degradation in the Model Standard WI Corn Scenario Constant degradation in top 1m Linearly declining degradation in the top 2m Decay rate for cm extended for next 1m 1000-d half life simulated below 1m 0.0 1/1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Soil Half-life = 1 day, Hydrolysis = 0 (stable), Koc =10 L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 12

13 Case Study 2 Application Type Degradation properties: 1-day aerobic soil degradation Stable hydrolysis 13

14 Impact of Application in the Model (WI Corn Scenario) Incorporation depth (1) (2) (3) (4) (5) 9m Water Table 1 m Well Screen Modeled Concentration (1) Linearly decreasing with depth (2) Uniform with depth (3) T-band (4) Applied at depth (5) Linearly increasing with depth 14 Figure not to scale

15 Concentration, ppb Impact of Application Type in the Model Concentrations compared for different methods of application in PWC model Linearly decreasing with depth (4 cm) Uniform with depth (4 cm) Acute Chronic T-band (4 cm) Applied at depth (4 cm) Linearly increasing with depth (4 cm) Soil Half-life = 1 day, Hydrolysis = 0 (stable), Koc = 10 L/kg All runs done with: WI Sand Scenario; Applied at a rate of 1 kg/ha at 2 weeks before emergence. 15

16 Case Study 3 Compartment Thickness (Dispersion) Chemical 1 Degradation properties: 1-day aerobic soil degradation Stable hydrolysis Chemical 2 Degradation properties: 1000-day aerobic soil degradation 100-day hydrolysis 16

17 Impact of Compartment Thickness in the Model (WI Corn Scenario) 9m 6 horizons in 1m 8m (1) (2) (3) 50 cm 20 cm 1) 10 1 cm 2) cm 3) cm 4) cm 5) cm 6) cm Water Table 50 cm 1) 10 1 cm 2) cm 3) cm 4) cm 5) cm 6) cm 1 m Well Screen Modeled Concentration (1) Standard Scenario (2) Standard WI Corn Scenario, 800 cm compartments as 20 cm (3) Standard WI Corn Scenario, 20 to 100 compartments as 10 cm Figure not to scale 17

18 Concentraion, ppb Impact of Dispersion in the Model Comparison for different compartment thicknesses in PWC model Standard WI Corn Scenario Standard WI Corn Scenario, 800 cm compartments as 20 cm Standard WI Corn Scenario, 20 to 100 cm compartments as 10 cm /1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Soil Half-life = 1 day, Hydrolysis = 0 (stable), Koc = L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 18

19 Normalized Concentration, ppb Impact of Dispersion in the Model Comparison for different compartment thicknesses in PWC model Acute Chronic Standard WI Corn Scenario Standard WI Corn Scenario, 800 cm compartments as 20 cm Standard WI Corn Scenario, 20 to 100 cm compartments as 10 cm Soil Half-life = 1 day, Hydrolysis = 0 (stable), Koc = 10 L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 19

20 Concentraion, ppb Impact of Dispersion in the Model Comparison for different compartment thicknesses in PWC model Standard WI Corn Scenario Standard WI Corn Scenario, 800 cm compartments as 20 cm Standard WI Corn Scenario, 20 to 100 cm compartments as 10 cm /1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Soil Half-life = 1000 day, Hydrolysis = 100, Koc = 10 L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 20

21 Normalized Concentration, ppb Impact of Dispersion in the Model Comparison for different compartment thicknesses in PWC model Acute Chronic Standard WI Corn Scenario Standard WI Corn Scenario, 800 cm compartments as 20 cm Standard WI Corn Scenario, 20 to 100 cm compartments as 10 cm Soil Half-life = 1000 day, Hydrolysis = 100, Koc = 10 L/kg All runs done with: WI Corn Scenario; CAM 1 application (linearly decreasing with depth). Applied at a rate of 1 kg/ha at 2 weeks before emergence. 21

22 Case Study 4 Degradation including Volatilization and Non-linear Sorption Chemical 1 Degradation properties: 1-day aerobic soil degradation Stable hydrolysis Chemical 2 Degradation properties: 1000-day aerobic soil degradation 100-day hydrolysis Chemical 3 Degradation properties: 10-day aerobic soil degradation Stable hydrolysis Chemical 4 Degradation properties: 1-day aerobic soil degradation 100-day hydrolysis 22

23 Concentration, ppb Concentration, ppb Impact of Volatilization Acute % % % 93% % % % 93% H=0, S=1d H=100d, S=1000d H=0, S=10d H=100, S=1d CAM 8, Vol = off CAM 8,Vol = on CAM 1,Vol = off CAM 1,Vol = on Chronic % % % 88% 51% % % 90% H=0, S=1d H=100d, S=1000d H=0, S=10d H=100, S=1d CAM 8, Vol = off CAM 8,Vol = on CAM 1,Vol = off CAM 1,Vol = on 23 All runs done with: WI Corn Scenario; Applied at a rate of 1 kg/ha at 2 weeks before emergence.

24 Concentration, ppb Concentration, ppb Impact of Freundlich (Nonlinear Sorption) 50 Acute H=0, S=1d H=100d, S=1000d H=0, S=10d PWC,CAM 8, F = off PWC,CAM 8,F = on PWC,CAM 1,F= off PWC,CAM 25 1,F = on winp,cam 8, F = off winp,cam 8,F = on winp,cam 1,F= off winp,cam 1,F = on 20 Chronic H=0, S=1d H=100d, S=1000d H=0, S=10d PWC,CAM 8, F = off PWC,CAM 8,F = on PWC,CAM 1,F= off PWC,CAM 1,F = on winp,cam 8, F = off winp,cam 8,F = on winp,cam 1,F= off winp,cam 1,F = on All runs done with: WI Corn Scenario; Applied at a rate of 1 kg/ha at 2 weeks before emergence. 24

25 Case Study 5 Rainfall on Application Day Chemical 1 Degradation properties: 1-day aerobic soil degradation Stable hydrolysis Chemical 2 Degradation properties: 1000-day aerobic soil degradation 100-day hydrolysis Chemical 3 Degradation properties: 10-day aerobic soil degradation Stable hydrolysis 25

26 4/1/1973 4/2/1973 4/3/1973 4/4/1973 4/5/1973 4/6/1973 4/7/1973 4/8/1973 4/9/1973 4/10/1973 4/11/1973 4/12/1973 4/13/1973 4/14/1973 4/15/1973 4/16/1973 4/17/1973 4/18/1973 4/19/1973 4/20/1973 4/21/1973 4/22/1973 4/23/1973 4/24/1973 4/25/1973 4/26/1973 4/27/1973 4/28/1973 4/29/1973 4/30/1973 Rainfall (cm) Impact of Heavy Rain (> 2 inch) on Application Day w14920.dvf (WI Corn) w14920.dvf with >2" rain moved from DOA to 4/22 in 1973 w14920.dvf with >2" rain moved from DOA to 4/17 in /15/1973, 4/22/1973, 5.23 cm 5.23 cm 4/22/1973, 5.23 cm

27 Concentration, ppb Impact of Heavy Rain on Application Day w14920.dvf (WI Corn) with 5.23 cm on DOA (4/15/73) w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/17/73 w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/22/ /1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Parameter Value Hydrolysis (days) 0 Soil Half-life (days) 1 Application type CAM 1 Weather Acute (ppb) Chronic (ppb) w14920.dvf cm rain 4/ cm rain 4/

28 Concentration, ppb Impact of Heavy Rain on Application Day w14920.dvf (WI Corn) with 5.23 cm on DOA (4/15/73) w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/17/73 w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/22/ /1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Parameter Value Hydrolysis (days) 100 Soil Half-life (days) 1000 Application type CAM 1 Weather Acute (ppb) Chronic (ppb) w14920.dvf cm rain 4/ cm rain 4/

29 Concentration, ppb Impact of Heavy Rain on Application Day w14920.dvf (WI Corn) with 5.23 cm on DOA (4/15/73) w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/17/73 w14920.dvf with 5.23 cm rain moved from 4/15/73 to 4/22/ /1/1961 6/24/ /15/1971 6/6/ /27/1982 5/19/1988 Parameter Value Hydrolysis (days) 0 Soil Half-life (days) 10 Application type CAM 1 Weather Acute (ppb) Chronic (ppb) w14920.dvf cm rain 4/ cm rain 4/

30 Impact of Heavy Rain on Application Day Normalized Concentration (ppb) Normalized Concentration (ppb) Acute Standard scenario weather 5.23 cm rain 4/ cm rain 4/22 H=0, S=1d H=100d, S=1000d H=0, S=10d Chronic Standard scenario weather cm rain 4/ cm rain 4/22 H=0, S=1d H=100d, S=1000d H=0, S=10d 30

31 Conclusions Of the 5 cases, the degradation factor profile in the 0-1m zone and hydrolysis only from 1m to GW is the most sensitive. This approach for Tier 2 may be appropriate but given the high sensitivity on model predictions further investigation may be warranted to whether the result is not overly conservative for some chemicals Application method is largely non-sensitive for PRZM5 GW - depth of incorporation can increase the concentrations. Numerical dispersion in PRZM5 can have different effects based on properties of chemical for both acute and chronic concentrations by spreading out the center of mass. This spreading out changes the chemical mass interacting with the degradation factors in the 0-1 m zone. 31

32 Conclusions (cont) Volatilization lowers concentrations for three reasons. First from mass lost directly from system due to the volatilization. Secondly, by spreading the center of mass throughout profile. Finally, from upward movement of gas phase which then repartitions into solution phase and subjected to the faster degradation of upper profile PRZM5 implementation of NLS appears to generate answers that are quite different than the FOCUS model implementation of NLS (PEARL, PELMO, WINPRZM, MACRO). It is especially apparent in CAM 1 applications (linearly decreasing with depth) with short aerobic soil half-lives. Modeling application on heavy rainfall day on has a multi-fold impact on groundwater concentrations of low persistence compounds compared to persistent compounds. 32

33 Thank you! Any questions? American Chemical Society, Philadelphia August 25 th,