Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
By Todd S. Miller, Edward F. Bugliosi, Kari K. Hetcher-Aguila, and David A. Eckhardt
Prepared in cooperation with the Oswego County Health Department
Scientific Investigations Report 2007–5169
U.S. Department of the Interior U.S. Geological Survey U.S. Department of the Interior DIRK KEMPTHORNE, Secretary U.S. Geological Survey Mark D. Myers, Director
U.S. Geological Survey, Reston, Virginia: 2007
For more information on the USGS--the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment: World Wide Web: http://www.usgs.gov Telephone: 1-888-ASK-USGS
Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted materials contained within this report.
Suggested citation: Miller, T.S., Bugliosi, E.F., Hetcher-Aguila, K.K., Eckhardt, D.A., 2007, Hydrogeology of two areas of the Tug Hill glacial-drift aquifer, Oswego County, New York: U.S. Geological Survey Scientific Investigations Report 2007–5169, 42 p, online only. iii
Contents
Abstract...... 1 Introduction...... 2 Purpose and Scope...... 2 Methods ...... 4 Test Wells and Water-Level Measurements...... 4 Water-Quality Samples...... 4 Regional Hydrogeologic Setting...... 5 Hydrogeology of the Pulaski Water-Supply Area ...... 5 Ground Water...... 7 Water Quality ...... 7 Physical Properties ...... 7 Inorganic Ions and Nutrients...... 12 Bacteria...... 13 Geochemical Age Dating ...... 13 Dissolved Gases...... 13 Chlorofluorocarbons...... 13 Sulfur Hexafluoride ...... 13 Tritium ...... 15 Hydrogeology of the Sandy Creek/Lacona Water-Supply Area...... 15 Hydrology...... 18 Water Quality ...... 20 Physical Properties ...... 20 Inorganic Ions and Nutrients...... 20 Geochemical Age Dating...... 20 Dissolved Gases...... 24 Chlorofluorocarbons...... 24 Sulfur Hexafluoride ...... 24 Simulation of Ground-Water Flow ...... 26 Model Design and Model Grid...... 26 Boundary Conditions...... 26 No-Flow Boundaries...... 26 Specified-Flux Boundaries...... 26 Head-Dependent Flux Boundaries...... 29 Model Calibration...... 29 Model Sensitivity...... 30 Horizontal Hydraulic Conductivity...... 32 Vertical Hydraulic Conductivity...... 32 Recharge...... 32 Simulation of Ground-Water Withdrawals and Determination of Contributing Areas...... 34 iv
Comparison of Geochemical Age Dating and Ground-Water Traveltimes Computed from Numerical Modeling...... 37 Apparent Ground-Water Ages ...... 37 Flow-Model Traveltimes and Chlorofluorocarbon Age Comparisons...... 37 Summary...... 39 Acknowledgments...... 41 Selected References...... 41
Figures
1–2. Maps showing— 1. Location of Sandy Creek/Lacona and Pulaski study areas, physiographic provinces, and Tug Hill glacial aquifer...... 3 2. Surficial geology and location of production wells for the Village of Pulaski in Richland, N.Y...... 6 3. Diagram of generalized geologic section A-A’ through part of Tug Hill glacial-drift aquifer at Richland, N.Y...... 8 4. Map showing water table in the vicinity of the Pulaski well fields during fall of 2003...... 9 5–6. Graphs showing— 5. Temperature of water in Village of Pulaski dug well OW 493, piezometer OW 501, and adjacent stream from July 2003 through November 2004...... 12 6. Tritium decay curves for Ottawa, Canada and value of tritium in wells OW 497, OW 499, and OW 500 sampled on May 7, 2004 near Pulaski, N.Y...... 15 7. Map showing surficial and bedrock geology, model boundary, and production- and test-well locations, and location of line of geologic section in the Sandy Creek/Lacona, N.Y. study area...... 16 8. Diagram of generalized hydrogeologic section B-B’ near Sandy Creek/ Lacona, N.Y...... 17 9. Map showing streamflow measurements, ground-water-flow directions, surface-water seepage area, and production wells in the southern part of the Sandy Creek/Lacona aquifer, N.Y...... 19 10. Diagram of wells and apparent ages of ground water determined by concentrations of chlorofluorocarbons in the villages of Sandy Creek and Lacona’s well field near Lacona, N.Y...... 21 11. Map showing three-dimensional-flow model cells, simulated cell types, and location of conceptual section C-C’ for the Sandy Creek/Lacona, N.Y., study area...... 27 12. Diagram of conceptual generalized, cross-sectional representation of ground-water-flow model showing cell types and directions of recharge and discharge to the aquifer in the Sandy Creek and Lacona, N.Y. area ...... 28 13–14. Maps showing— 13. Hydraulic conductivities for different aquifer deposits and model layers used in ground-water-flow model calibration...... 31 14. Water-table contours based on the calibrated, three-dimensional-flow-model output in the Sandy Creek/Lacona, N.Y. study area...... 33 15. Graph showing results of sensitivity analyses of hydraulic head in the sand and gravel aquifer in the Sandy Creek/Lacona, N.Y. study area...... 34 v
16a–b. Maps showing— a. Simulated areas contributing to production wells pumping at 100 gallons per minute, derived from particle-tracking results based on calibrated, three- dimensional-flow-model output in the Sandy Creek/Lacona, N.Y. study area...... 35 b. Simulated areas contributing to the southernmost production wells pumping at 200 gallons per minute, derived from particle-tracking results based on calibrated, three-dimensional-flow-model output in the Sandy Creek/Lacona, N.Y. study area...... 36 17. Conceptual diagram showing possible sources of older ground water recharging the sand and gravel aquifer near Lacona, N.Y...... 38
Tables
1. Inorganic and physical constituents of water samples from selected wells in the Pulaski, N.Y. study area, 2003–04...... 10 2. Concentrations of tritium, chlorofluorocarbon (CFC), and air-temperature correction data for wells in the Pulaski, N.Y. study area, 2003-04 ...... 14 3. Inorganic and physical constituents of ground water from wells in the villages of Sandy Creek and Lacona, N.Y. study area...... 22 4. Tritium, chlorofluorocarbon (CFC), and air-temperature correction data for wells in the Sandy Creek and Lacona, N.Y. study area...... 25 5a. Difference between measured and simulated heads at 11 selected non-pumped wells in the Sandy Creek/Lacona, N.Y. model study area ...... 30 5b. Difference between measured and simulated streamflow in the Sandy Creek/ Lacona, N.Y. model study area...... 30 6. Steady-state ground-water budget for the ground-water-flow model of the Tug Hill glacial-drift for average annual recharge conditions...... 32 vi
Conversion Factors, Datum, and Water-Quality Units
Multiply By To obtain Length inch (in.) 2.54 centimeter (cm) inch (in.) 25.4 millimeter (mm) foot (ft) 0.3048 meter (m) mile (mi) 1.609 kilometer (km) Area square foot (ft2) 0.09290 square meter (m2) square mile (mi2) 259.0 hectare (ha) square mile (mi2) 2.590 square kilometer (km2) Volume gallon (gal) 3.785 liter (L) gallon (gal) 0.003785 cubic meter (m3) gallon (gal) 3.785 cubic decimeter (dm3) million gallons (Mgal) 3,785 cubic meter (m3) cubic foot (ft3) 28.32 cubic decimeter (dm3) cubic foot (ft3) 0.02832 cubic meter (m3) Flow rate acre-foot per day (acre-ft/d) 0.01427 cubic meter per second (m3/s) acre-foot per year (acre-ft/yr) 1,233 cubic meter per year (m3/yr) acre-foot per year (acre-ft/yr) 0.001233 cubic hectometer per year (hm3/yr) foot per second (ft/s) 0.3048 meter per second (m/s) foot per day (ft/d) 0.3048 meter per day (m/d) cubic foot per second (ft3/s) 0.02832 cubic meter per second (m3/s) cubic foot per day (ft3/d) 0.02832 cubic meter per day (m3/d) gallon per day (gal/d) 0.003785 cubic meter per day (m3/d) gallon per minute (gal/min) 0.06309 liter per second (L/s) million gallons per day (Mgal/d) 0.04381 cubic meter per second (m3/s) million gallons per day per square 1,461 cubic meter per day per square mile [(Mgal/d)/mi2] kilometer [(m3/d)/km2] inch per year (in/yr) 25.4 millimeter per year (mm/yr) mile per hour (mi/h) 1.609 kilometer per hour (km/h) Hydraulic conductivity foot per day (ft/d) 0.3048 meter per day (m/d) Transmissivity* foot squared per day (ft2/d) 0.09290 meter squared per day (m2/d)
Other Abbreviations Used in this Report foot per foot ft/ft milliliter mL vii
Temperature in degrees Celsius (°C) may be converted to degrees Fahrenheit (°F) as follows: °F=(1.8×°C)+32 Temperature in degrees Fahrenheit (°F) may be converted to degrees Celsius (°C) as follows: °C=(°F-32)/1.8 Vertical coordinate information is referenced to the National Geodetic Vertical Datum of 1929 (NGVD 1929). Horizontal coordinate information is referenced to the North American Datum of 1983 (NAD 83). Elevation, as used in this report, refers to distance above the vertical datum. *Transmissivity: The standard unit for transmissivity is cubic foot per day per square foot times foot of aquifer thickness [(ft3/d)/ft2]ft. In this report, the mathematically reduced form, foot squared per day (ft2/d), is used for convenience. Specific conductance is given in microsiemens per centimeter at 25 degrees Celsius (µS/cm at 25°C). Concentrations of chemical constituents in water are given either in milligrams per liter (mg/L) or micrograms per liter (µg/L). viii
This page has been left blank intentionally. Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
By Todd S. Miller, Edward F. Bugliosi, Kari K. Hetcher-Aguila, and David A. Eckhardt
Abstract salt in the ground water at the Village of Pulaski well field. Three of the four samples were analyzed for total coliform Two water-production systems, one for the Village of bacteria and Escherichia coli (E. coli), and only total coliform Pulaski and the other for the Villages of Sandy Creek and bacteria were detected in all three samples. E. coli was not Lacona in Oswego County, New York, withdraw water from detected in any samples. the Tug Hill glacial-drift aquifer, a regional sand and gravel The Villages of Sandy Creek and Lacona use about aquifer along the western flank of the Tug Hill Plateau, and 270,000 gallons of water per day for public consumption— provide the sole source of water for these villages. As a result alternating withdrawals from northern and southern well of concerns about contamination of the aquifer, two studies fields located in glaciolacustrine beach, glaciofluvial outwash, were conducted during 2001 to 2004, one for each water- and alluvial inwash deposits consisting mostly of silty sand production system, to refine the understanding of ground- and gravel. Four test wells were drilled, hydraulic heads water flow surrounding these water-production systems. were measured, and water samples collected between 2001 Also, these studies were conducted to determine the cause of and 2003 and analyzed for physical properties, inorganic the discrepancy between ground-water ages estimated from constituents, nutrients, bacteria, tritium, dissolved gases, previously constructed numerical ground-water-flow models and chlorofluorocarbons. for the Pulaski and Sandy Creek/Lacona well fields and the The aquifer in the Sandy Creek/Lacona area is highly apparent ground-water ages determined using concentrations susceptible to contamination because the aquifer (1) is of tritium and chlorofluorocarbons. unconfined, (2) is highly transmissive, (3) is thin (10 to 25 The Village of Pulaski withdrew 650,000 gallons per feet) and narrow (about 0.5 miles wide), and (4) has relatively day in 2000 from four shallow, large-diameter, dug wells short ground-water flowpaths from recharge to discharge finished in glaciolacustrine deposits consisting of sand with areas (including wells). Additionally, drainage ditches east some gravelly lenses 3 miles east of the village. Four 2-inch of the southern well field intercept westward-flowing ground diameter test wells were installed upgradient from each water, which then is routed to an area just upgradient from production well, hydraulic heads were measured, and water the production wells where some of the water seeps back into samples collected and analyzed for physical properties, the aquifer and is captured by these wells, effectively short- inorganic constituents, nutrients, bacteria, tritium, dissolved circuiting the ground-water-flow system. gases, and chlorofluorocarbons. Water-quality samples collected from three wells in Recharge to the Tug Hill glacial-drift aquifer is from the Sandy Creek/Lacona southern well field indicate that precipitation directly over the aquifer and from upland the ground water is a sodium-bicarbonate/sulfate type and sources in the eastern part of the recharge area, including of good quality; however, in one water sample, the sodium (1) unchannelized runoff from till and bedrock hills east of concentration of 63 milligrams per liter slightly exceeded the the aquifer, (2) seepage to the aquifer from streams that drain U.S. Environmental Protection Agency’s Secondary Maximum the Tug Hill Plateau, (3) ground-water inflow from the till and Contaminant Level of 60 milligrams per liter. Five samples bedrock on the adjoining Tug Hill Plateau. were collected from test and production wells near, and in, Water-quality data collected from four piezometers near the southern well field and analyzed for concentrations of the production wells in November 2003 indicated that the chlorofluorocarbons and tritium, to determine ground-water water is a calcium-bicarbonate type with iron concentrations ages that ranged from approximately 25 years old or less. that slightly exceeded the U.S. Environmental Protection Results from a three-dimensional, finite-difference, Agency’s Secondary Maximum Contaminant Level in three ground-water-flow model used to simulate ground-water of the four samples. The relatively small concentrations of flow in the Sandy Creek/Lacona study area indicated that major ions and nutrients in the samples indicate that there is (1) about 61 percent of recharge to the aquifer is from no contamination from septic-tank effluent and dissolved road precipitation directly on the aquifer, 38 percent of recharge is 2 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York from streamflow seeping into the aquifer, one percent is from 20–50 ft thick) travels from recharge areas to wells in less unchanneled runoff and ground-water inflow from uplands, than 25 years. However, ground-water traveltimes that were and (2) the majority (86 percent) of discharge from the aquifer estimated from numerical ground-water-flow models by the is from ground-water seepage to the headwater tributaries U.S. Geological Survey (USGS) (Zarriello, 1993) in the Sandy west (downgradient) from the well fields with the remaining Creek/Lacona area and by Malcolm Pirnie, Inc., (2000) in the 14 percent discharging to streams, manmade ditches, and the Village of Pulaski area, showed discrepancies compared to two production wells in the central part of the aquifer. Little traveltimes determined by chemical data (Komor, 2002). or no water pumped from the production well in the sand and Plausible explanations for the discrepancies between gravel aquifer appears to be derived from bedrock because the results of ground-water-flow modeling and geochemical (1) a till layer on top of bedrock impedes flow between the dating approaches are that (1) estimated values of hydraulic sand and gravel aquifer and the underlying bedrock, (2) the conductivity were too great or porosity values were too small, hydraulic-head gradient between the two is downward, not producing model-simulated ground-water traveltimes that upward, and (3) the water chemistry in each is different. were shorter than those determined by chemical age dating; Ground-water particle-tracking results indicate that (2) two-dimensional flow models that were used to depict the ground-water traveltimes are relatively short, generally taking system (Zarriello, 1993, and Malcolm Pirnie Inc., 2000), did less than 1.5 years for most of the water to flow from the not adequately address vertical components of ground-water eastern boundary of the aquifer to the southern well field. This flow, which when accounted for in a three-dimensional flow is in contrast to geochemical age dates that indicate that water model would tend to increase the length of flow paths and, withdrawn from the southern well field is 13 to 17 years old. therefore, increase traveltimes; (3) other geohydrologic units A plausible reason for the discrepancy in ground-water age not included in the numerical model, such as bedrock, might might be that much of the recharge is relatively older water contribute significant amounts of older water to the production from upland sources. Possible upland sources of the relatively wells; and (or) (4) geochemical age-dating methods were old water are (1) ground-water discharge from till and bedrock affected by chemical dispersion, trapped air in the aquifer, units that enters streams that drain the Tug Hill uplands or biogeochemical reactions, which can lead to erroneous during base-flow conditions, then lose water to the sand and ground-water ages. gravel where they flow over the aquifer, and (2) ground-water In 2001, the USGS, in cooperation with the Villages inflow from adjacent till and bedrock uplands. Similarly, of Sandy Creek and Lacona and the Oswego County Health this discrepancy in ground-water age has been noted in Department, began a 2-year study of the part of the Tug Hill another area of the state with a similar hydrogeologic setting glacial-drift aquifer that is tapped by wells for the Village of where older stream water from the adjacent till and bedrock Lacona in the Town of Sandy Creek, Oswego County, N.Y. In uplands was a source of recharge to the adjacent sand and 2003, the USGS, in cooperation with the Village of Pulaski gravel aquifer and contributed to the apparent age of the and the Oswego County Health Department, began a 2-year ground water. study of the part of the Tug Hill glacial-drift aquifer that is tapped by wells for the Village of Pulaski in the Hamlet of Richland, Oswego County, N.Y. Results of this study were compared with results of the numerical ground-water-flow Introduction model developed by Malcolm Pirnie, Inc. (2000) and the previous geochemical study by Komor (2002). Shallow ground water is readily susceptible to contamination from many sources, including leaking petroleum-storage tanks, fertilizers, septic systems, and Purpose and Scope road salt. A quantitative estimate of the rate and direction of ground-water flow facilitates the determination of the potential This report (1) presents data and interpretations that effects of contamination from land-surface sources, as well refine the understanding of the ground-water-flow system in as the development of reasonable “protection” strategies parts of the Tug Hill glacial-drift aquifer in Oswego County, to ensure the good quality of the ground water. The age of specifically near the production wells for the villages of Sandy ground water can be used to estimate recharge rates and Creek/Lacona and Pulaski, and (2) discusses the differences refine hydrologic models of ground-water-flow systems. The in estimates of ground-water recharge ages and ground-water- ages of ground water from 28 wells in Oswego County were flow rates estimated using numerical ground-water-flow evaluated in 1999 (Komor, 2002) by examining the ratios of modeling for the area near the Sandy Creek/Lacona well the stable isotopes Oxygen-18 (18O) and Deuterium (2H) and field and those estimated using geochemical data presented by using geochemical age-dating techniques that measure by Komor (2002) for these areas. The report presents and concentrations of chlorofluorocarbons (CFCs) and tritium. summarizes the chemical data used to determine the apparent These data indicate that at two well sites (Villages of Sandy age of ground water in these two areas and compares the Creek and Lacona, and the Village of Pulaski) (fig. 1) water results of chemical-age data to ground-water traveltimes withdrawn from a shallow, unconfined aquifer (typically derived from a numerical ground-water-flow model of the Introduction 3
76°10' 76°5'
Study Area 73° +45° 77° 44°+ JEFFERSON COUNTY OSWEGO COUNTY
N.Y. 73° 42°+ +42° 43°40' 80° P.A. Sandy Creek/ Lacona study area North Pond Village of Sandy Creek Village of Lacona
South Tug Hill Pond Plateau Lake Ontario Plain
Lake Ontario
81
43°35'
Salmon River Hamlet of Pulaski Richland study area Village of Pulaski
EXPLANATION PHYSIOGRAPHIC PROVINCES TUG HILL PLATEAU–Uplands of till overlying bedrock, yields little water LAKE ONTARIO PLAIN–Surficial deposits generally consist of drumlins (till) and interdrumlin areas of glaciolacustrine deposits consisting of fine sand, silt, and clay TUG HILL GLACIAL-DRIFT AQUIFER–Aquifer material generally consists of fine-to-medium lacustrine sand (mostly in the western part) and coarser beach sand and gravel (mostly in the eastern part) deposited at the shoreline of proglacial Lake Iroquois 0 1 2 MILES Boundary between Lake Ontario Plain and Tug Hill Plateau Physiographic Provinces 0 1 2 KILOMETERS 81 Interstate 81 Base from U.S. Geological Survey, 10-meter Digital Elevation Model, Surficial geology modified from Universal Transverse Mercator projection, Zone 18N, NAD 1927 Cadwell and Muller, 1986
Figure 1. Location map of Sandy Creek/Lacona and Pulaski study areas, physiographic provinces, and Tug Hill glacial aquifer. 4 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Sandy Creek/Lacona area. This report combines information study area, test well depths ranged from 19 to 43 ft below land gathered from two separate studies (one near Pulaski, N.Y. and surface. At all test well sites, elevations of measuring points the other near Sandy Creek/Lacona, N.Y.) because production on the tops of casings and land surface were leveled from a wells in both study areas tap the Tug Hill glacial aquifer, known elevation reference point referenced to the National which has similar hydrogeologic characteristics in both areas, Geodetic Vertical Datum of 1929 (NGVD 1929). All test wells and because the approaches to both studies are similar. were developed for a minimum of 30 minutes, and in certain The first part of this report (1) discusses the hydro- cases, up to 4 hours to minimize the turbidity of the water geology of part of the Tug Hill glacial-drift aquifer that is pumped from each well. tapped by wells for the Village of Pulaski, (2) presents the In the Pulaski study area, water temperature and water- results of ground-water age dating and chemical analyses level data loggers were installed in (1) one of the Village of from sa mples collected from test wells with interpretation, Pulaski’s dug wells, (2) in a stream approximately 20 ft south and (3) provides a water-table elevation map for the area from the well, and (3) in a piezometer between the stream and around the Village of Pulaski well field. The second part of the dug well to determine seasonal fluctuations of ground- this report (1) discusses the hydrogeology of part of the Tug water and surface-water temperatures and hydraulic heads Hill glacial-drift aquifer that is tapped by wells for the Village (hydraulic head is referred to as head in the rest of this report). of Lacona, (2) presents the results of ground-water age dating The temperature sensors were deployed within the lower part and chemical analyses from samples collected from test wells of the screened area of the piezometers and within the dug with interpretation, and (3) presents results of simulation of well, and in the screened interval open to the streamflow above the ground-water-flow system from a numerical ground-water- the bed of the stream. These data were used to determine flow model. possible infiltration of surface water to the production well. In the Pulaski study area, heads were measured in (1) the piezometers that were installed for this study, (2) one Methods domestic well and one commercial well in the Hamlet of Richland, (3) the Village of Pulaski production wells, and Water-quality samples were collected during winter 2003 (4) local streams to determine the local configuration of the and spring 2004 from four 2-in. diameter test wells installed water table and direction of ground-water flow around each upgradient from each production well for the Village of production well. In the Sandy Creek/Lacona study area, head Pulaski and were analyzed for physical properties, inorganic measurements were made in the test wells, and water-level constituents, nutrients, bacteria, sulfur hexafluoride (SF6), measurements were made in drainage ditches and local stream dissolved gases, and CFCs. During summer 2001 and spring channels to aid in construction of the numerical ground-water- 2002, water-quality samples were collected from four 2-in. flow model, determine vertical distribution of heads, and diameter test wells and from one production well near the calibrate the model. Villages of Sandy Creek and Lacona and were analyzed for physical properties, inorganic constituents, nutrients, bacteria, Water-Quality Samples SF6, dissolved gases, and three isotopes of CFCs [(CFC-11 (CFCl ), CFC-12 (CF Cl ), and CFC-113 (C F Cl )]. Apparent 3 2 2 2 3 3 At both the Pulaski and Sandy Creek/Lacona sites, water ground-water ages computed from tritium, SF , dissolved 6 samples were collected from piezometers with short-screen gases, and CFC analyses were compared with time-of- lengths to ensure that the sampled water was from a discrete travel analyses from a numerical ground-water-flow model depth in the aquifer rather than from production wells that constructed for this investigation and with previous ground- have long screens drawing water from large vertical sections water ages determined in a previous geochemical study by of the aquifer that don’t represent discrete ground-water-flow Komor (2002). zones within the aquifer. Before water samples were collected from the piezometers, a peristaltic pump was used to purge Test Wells and Water-Level Measurements at least 10 casing volumes of water. Then, a stainless-steel submersible pump with copper tubing was used to collect Test wells were installed in both study areas to collect water samples. Field measurements were made of pH, specific stratigraphic, water-level, and water-quality data. Eleven conductance, and water temperature. 1.5-in.-diameter piezometers with 2-ft-long screened drive In November 2003, water-quality samples were collected points were driven around the four Village of Pulaski from 2-in.-diameter piezometers directly upgradient from production wells, and four 2-in.-diameter test wells were each of the four production wells in the Pulaski well field. The installed using an auger rig in the Sandy Creek/Lacona study samples were analyzed for nutrients, common ions, inorganic area. The depths of the piezometers in the Pulaski study area constituents, bacterial content, three CFC isotopes, SF6, and ranged from 5 to 11 ft below land surface and the piezometers tritium. Additional samples for CFCs were collected in May were finished so that the bottoms of the screens were at 2004 from three piezometers for comparison of the results approximately the same elevations as the bottom of the with the November 2003 samples. Samples for tritium were corresponding production wells. In the Sandy Creek/Lacona collected in August 2004 to compare the tritium-based ages Hydrogeology of the Pulaski Water-Supply Area 5 with ages determined from CFC analyses. At the Sandy Creek/ alluvial and Pleistocene glaciofluvial sand and gravel deposits, Lacona southern well field, water samples were collected encompasses the eastern part of the county. The western part in November 2001 from a well and three nearby nested of the county lies in the Lake Ontario Plain (fig. 1), which piezometers. Another sample was collected from a USGS contains drumlins consisting of till and interdrumlin areas test well 2,200 ft upgradient and northeast from the Villages of glaciolacustrine deposits consisting of fine sand, silt, and of Sandy Creek and Lacona’s southern well field. The water clay. The beach sand and gravel deposited at the shoreline samples were analyzed for the same constituents as for the of proglacial Lake Iroquois generally forms the boundary Pulaski study area, except that bacteria were excluded. between the two physiographic provinces (fig. 1). Bedrock Tritium samples were analyzed at the USGS Isotope in the area consists of Late Ordovician and Early Silurian Geochemistry Laboratory in Menlo Park, Calif. CFC samples shale, siltstone, and sandstone units that dip gently to the were collected according to specific sampling methods to southwest—where fractured, the units form aquifers that minimize atmospheric exposure (U.S. Geological Survey, typically yield water to wells in amounts sufficient only to 2003), and were analyzed at the USGS Chlorofluorocarbon supply homeowners and small farms (Miller, 1982). Laboratory in Reston, Va. Age-dating of ground water using Precipitation that falls on the Tug Hill Plateau either these methods gives an approximate time since the water infiltrates into and recharges till deposits and bedrock units entered the saturated zone after recharge (when precipitation and (or) runs off into westward-flowing streams. Many enters the aquifer), essentially stopping the change of streams that drain the western flank of the Tug Hill Plateau are concentrations of these environmental tracers by cutting off sources of recharge to the Tug Hill glacial-drift aquifer (fig. 1) their interaction with the atmosphere. The ages are based on because these streams typically lose water as they flow across matching the concentrations of the tracers in ground water to the eastern part of the aquifer. Water lost from these streams is the historical atmospheric concentrations of the tracers, which a major source of recharge (in addition to precipitation) to the are well documented. aquifer (Miller and others, 1989). Dissolved gas data are used to define recharge temperatures for CFC dating, to reconstruct initial nitrate concentrations based on the amount of denitrification that Hydrogeology of the Pulaski has occurred, and to trace sources of ground-water recharge. In addition to providing gas concentrations, excess air Water-Supply Area concentrations are provided, and the recharge temperature and the amount of denitrification are estimated. Temperature In 2000, the Village of Pulaski obtained its water supply and pressure (excess air) of recharge water are important of about 650,000 gal/d (Malcolm Pirnie, Inc., 2000) from four components in determining the age of the water based on concrete-lined, dug wells with diameters of 8 to 10 ft, ranging concentrations of CFCs, and can be calculated from the ratios from 8 to 12 ft deep, in the Tug Hill glacial-drift aquifer in the of dissolved nitrogen and argon gas concentrations in the Hamlet of Richland (figs. 1 and 2). The wells are 3 mi east water along with the recharge elevation (Heaton, 1981; Heaton of the Village of Pulaski in the eastern part of the Hamlet of and Vogel, 1981; Stute and Schlosser, 1999). Details of CFC Richland (fig. 2). The wells are installed in glaciolacustrine dating are explained in Busenberg and Plummer (1992), deposits consisting of sand with some gravelly lenses and Plummer and others (1993), and Reilly and others (1994). intercept ground water flowing to the headwaters of Spring The samples for inorganic constituents, nutrients, and Brook. The aquifer materials in the Richland area were bacteria for both the Pulaski and Sandy Creek/Lacona areas deposited about 14,000–15,000 years ago when coarse-grained were collected by the USGS and Oswego County Health sediments were deposited along the shore of proglacial Lake Department and were analyzed by Upstate Laboratories in Iroquois (Ridge, 2003). Syracuse, N.Y., a private laboratory approved by the New Four types of unconsolidated deposits form the York State Department of Health. The analytical results of unconfined Tug Hill glacial-drift aquifer in the area of the inorganic, nutrient, and bacteriological sampling can provide Pulaski well field: (1) near-shore lacustrine fine-to-medium sand with some gravelly lenses, (2) beach sand and gravel, insight into potential sources of contamination within the (3) wind-blown sand (dunes), and (4) kame sand and gravel aquifer and can aid in estimation of the relative age of the (fig. 2). Nearshore sand deposits formed along both sides ground water and determination of ground-water-flow paths in of the barrier-beach ridge that trends north-south through the area. the Hamlet of Richland. Nearshore sand deposits formed in shallow standing water west of the shoreline, and backwash Regional Hydrogeologic Setting and washover deposits formed in a shallow lagoon that was present between the barrier beach to the west and kame Oswego County is in north-central New York and deposits that flanked the Tug Hill Plateau to the east (fig. 2). encompasses 953 mi2 on the southeastern shore of Lake Wind-blown sand was deposited on the leeward (east) side Ontario (fig. 1). The Tug Hill Plateau, a bedrock upland area of the barrier beach by the prevailing winds when the level mantled in most places with till and locally with Holocene of glacial Lake Iroquois began to decline, thus exposing the 6 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
76°03'30" 76°03' 76°02'30" Inland beach ridge
Area shown in figure 4 43°34'30"
Sand dunes
OW 497 OW 498 OW 496 Hamlet of OW495 Richland OW 494 Spring OW 499 Brook OW 493 A’ A OW 500 Collection cistern OW 491 OW 492 (B7) Water gravity fed OW 501 through pipe to distribution system 43°34' Sand Barrier- dunes beach ridge
Backshore and washover sand Nearshore deposits sand deposits
43°33'30" 0 500 1,000 FEET
0 125 250 METERS Base from U.S. Geological Survey Richland, NY Quadrangle, 1982, 1:24,000, Geology from Miller, 1980b Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION LACUSTRINE FINE TO COARSE SAND–Sand with some A A’ gravelly lenses that was deposited nearshore from proglacial LINE OF GEOLOGIC SECTION–Section shown in figure 3 Lake Iroquois beach and as backwash and washover behind barrier beach PRODUCTION WELL (VILLAGE OF PULASKI) AND OW 493 WELL NUMBER–Well number assigned by BEACH SAND AND GRAVEL–Ridge of well-sorted U.S. Geological Survey (USGS) coarse sand to cobble gravel deposited at the shoreline of proglacial Lake Iroquois INDUSTRIAL WELL (SCREENED) AND WELL NUMBER– OW 491 Well number assigned by USGS. (Number in parentheses WIND-BLOWN SAND–Dunes and blankets of fine to (B7) assigned by well owner) medium sand KAME SAND AND GRAVEL–Hummocky mounds DOMESTIC WELL AND WELL NUMBER–Well number of poorly sorted sand and gravel deposited by OW 492 assigned by USGS glacial meltwaters PIEZOMETER (DRILLED FOR THIS STUDY BY USGS) INDUSTRIAL PRODUCTION WELL FIELD OW 501 AND WELL NUMBER–Well number assigned by USGS.
Figure 2. Surficial geology and location of production wells for the Village of Pulaski in Richland, N.Y. (Section A–A’ shown in figure 3). Hydrogeology of the Pulaski Water-Supply Area 7 near-shore sand on the windward (west) side of the beach to form the headwaters of tributaries in the area, such as near wind erosion. Spring Brook (fig. 4). Increased ground-water discharge into The Pulaski water-production system works solely by headwater springs and saturation of the ground in the area of gravity. The water collected from the four production wells the well field near Spring Brook reported by the Pulaski water- drains by gravity through a pipe system and is routed to a supply operator in November 2003 (Gary Stevens, Water- main “collection cistern” to the west of the wells (fig. 2). Supply Operator, Village of Pulaski, oral commun., 2003) From the collection cistern, the water continues by gravity to may have been caused by decreased ground-water withdrawals the west where it is chlorinated at a treatment building, and from 2.5 Mgal/d to 0.75 Mgal/d by the paper company during from there it is routed to a storage tank on the east side of the that year (Dave Montique, oral commun., 2004). Decreased Village of Pulaski. withdrawals by the paper company would likely result in higher water levels and increased ground-water flow westward toward the Pulaski well field. Ground Water The Tug Hill glacial-drift aquifer is as much as 65 ft thick Water Quality at an industrial well field in the eastern part of the study area (well OW 491, figs. 2 and 3). However, deep subsurface data Results of the inorganic analyses from the four in the Pulaski supply-well area is not available to accurately piezometers near the Pulaski production wells sampled determine the thickness in the western part of the aquifer. on November 12 and 25, 2003 show that the ground water Because of the lack of data, the aquifer depth was estimated to directly adjacent to the wells is potable (table 1). The major be between 30 and 50 ft, on the basis of a minimum thickness ions in the water were calcium and bicarbonate, and the from the drilling record (0–28 ft of sand and gravel) of well water met State and Federal water-quality guidelines for all OW 492 (300 ft east of the production well field) (fig. 2). The constituents analyzed except iron, which slightly exceeded the depth to the bottom of the aquifer was estimated by extending U.S. Environmental Protection Agency (USEPA) Secondary the regional dip of the westward sloping bedrock that is found Maximum Contaminant Level (SMCL) of 0.3 mg/L in three of in surrounding areas of the Tug Hill glacial-drift aquifer to the four samples (table 1). The relatively low concentrations well OW 492. of the major ions and nutrients in the samples indicated that Heads were measured in the Village of Pulaski septic-tank effluent and road salt runoff contamination do production wells, piezometers installed by the USGS around not affect the ground-water quality of the Village of Pulaski those production wells, and several privately owned wells in production wells. the study area during fall 2003. Water-table contours were drawn to represent the elevation of the water table (fig. 4). Physical Properties In general, the water table slopes to the west, indicating that ground water generally flows from east to west. The ground- Field pH of all samples ranged from 7.8 to 8.0, and water gradient is about 0.01 ft/ft in areas away from pumped specific conductance ranged from 167 to 247 µS/cm at 25°C wells and streams, but near ground-water discharge areas in November 2003, and from 86 to 165 µS/cm in May 2004 (such as pumped wells and streams) the gradient and velocity (table 1). The lowest value of specific conductance (86 µS/cm) of ground water increases. was measured on May 7, 2004, in a sample from test well The Tug Hill glacial-drift aquifer in the vicinity of the OW 499. The lower specific conductance values for the May Pulaski production wells is recharged from precipitation 2004 samples from wells OW 497, OW 499, and OW 500 that falls directly on the land surface overlying the aquifer indicate that fresh water (likely from spring snowmelt) that and on the kame sand and gravel that extends southeastward recently recharged the aquifer had little time to dissolve any from the Hamlet of Richland, and also from upland sources minerals and increase the specific conductance. These results in the eastern part of the aquifer (Miller and others, 1989). indicate relatively rapid flow through the aquifer (indicative These upland sources include (1) unchannelized runoff of shallow or short flow paths in a relatively homogenous from till and bedrock hills that border the east side of the material), with negligible contact with the bedrock. aquifer, (2) seepage losses from streams that drain the Tug Water temperatures in the sampled wells ranged from 7.6 Hill Plateau and that flow over and lose water to the aquifer, to 11.7°C in November 2003 and from 7.4 to 8.3°C in May and (3) ground-water inflow from the till and bedrock on the 2004. Water temperatures recorded continuously at (1) the adjoining Tug Hill Plateau. Village of Pulaski dug well OW 493, (2) an adjacent unnamed Most of the ground-water inflow from the southeast is stream 20 ft south of the production well, and (3) piezometer likely intercepted by industrial production wells for a paper OW 501 (installed between the stream and the production company southeast of Richland (hatched and shaded area well) show that temperature in the stream and piezometer in lower right corner, fig. 2), but inflow not captured by varied over time, whereas the temperature in the production the company’s wells may flow westward and discharge to well remained fairly constant (fig. 5). This may indicate the Pulaski production wells and to numerous springs that that there is little hydraulic connection between the nearby 8 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
A A’ West East FEET 600 Inland beach ridge
Backshore OW 491 Barrier Sand dunes beach and (B7) ridge washover sand deposits Nearshore sand 550 deposits
Spring- fed OW 492 Pond OW 493 (PW S3) Stream 500
450 VERTICAL EXAGGERATION x20 0 500 1,000 FEET VERTICAL DATUM IS NGVD 1929 0 100 200 METERS
EXPLANATION BEACH SAND AND GRAVEL–Ridge of well-sorted, WATER TABLE–Shows altitude of thinly bedded, medium to coarse sand and fine water level during average annual recharge to medium gravel deposited at the shoreline of conditions. Arrow shows general direction proglacial Lake Iroquois of ground-water flow
WIND BLOWN SAND–Dunes and blankets of fine to medium sand OW 493 WATER-SUPPLY WELL AND WELL NUMBERS– (PW S3) Upper well number assigned by U.S. Geological Survey (USGS); number in parentheses assigned LACUSTRINE FINE TO COARSE SAND–Sand with by Village of Pulaski some gravelly lenses that was deposited just off shore from glacial Lake Iroquois or as backwash in a lagoon behind the barrier beach OW 491 INDUSTRIAL PRODUCTION WELL (SCREENED) (B7) AND WELL NUMBERS–Upper well number KAME SAND AND GRAVEL–Hummocky mounds assigned by USGS, number in parentheses of poorly sorted sand and gravel deposited by Water level assigned by well owner glacial meltwaters in well Well TILL–Poorly sorted sediments at the base, within, screen{ or on top of the glacier that were deposited on top of bedrock. Sediments consist of pebbles and cobbles embedded in a silty clay matrix. Poor OW 492 DOMESTIC WELL AND WELL NUMBER– permeability Well number assigned by USGS
SILTSTONE AND SHALE–Ordovician age
Figure 3. Generalized geologic section A–A’ through part of Tug Hill glacial-drift aquifer at Richland, N.Y. (Line of section shown in figure 2.) Hydrogeology of the Pulaski Water-Supply Area 9
76°03'23" 43°03' 43°34'25"
520
512 510
500 OW 496 494.43 x 494.40 x 491.79 493.84 x x x WELL FIELD 492.46
OW 495 488.88
490 Brook
495.74 WELL FIELD510 OW 494 x Spring x 496.73 x x 496.21 490.7 496.17
490 495.8 500 main collection cistern 503.1 WELL FIELD x 499.07 500 x 517 OW 493 499.33 500.02
510 520 0 250 500 FEET
0 50 100 METERS 43°34'04" Topographic base from U.S. Geological Survey Richland, N.Y., 1982, 1:24,000, topographic map, Universal Transverse mercator projection, Zone 18N, NAD 1927; Orthophoto base from New York State Office of Cyber & Critical Infrastructure Coordination, Spring 2004
EXPLANATION 510 WATER-TABLE CONTOUR–Shows approximate altitude of water table (fall of 2003). Contour interval 10 ft. Datum is NGVD 1929. Dashed where approximate GROUND-WATER-FLOW PATH–Arrow shows generalized direction of ground-water flow 496.17 x PIEZOMETER WATER-LEVEL MEASUREMENT SITE–Number is altitude of water-table surface. Datum is NGVD 1929 499.33 SURFACE-WATER-LEVEL MEASUREMENT SITE–Number is altitude of water surface. Datum is NGVD 1929 OW 493 PRODUCTION-WELL SITE WATER-LEVEL MEASUREMENT–Number is altitude of water surface (OW numbers are USGS well numbers referred to in report). Datum is NGVD 1929 512 DOMESTIC-WELL WATER-LEVEL MEASUREMENT SITE–Number is altitude of water-table surface. Datum is NGVD 1929 517 COMMERCIAL-WELL WATER-LEVEL MEASUREMENT SITE–Number is altitude of water-table surface. Datum is NGVD 1929
Figure 4. Water table in the vicinity of the Pulaski well fields during fall of 2003. 10 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York ------<.5 <.5 <.5 Total Total <.05 <.05 <.05 Ortho- <0.5 nitrogen Kjeldahl <0.05 phosphate
) -1 ------3 <.5 <.5 <.5 ------as 90 74 <0.5 116 107 (mg/L (as N) (00608) HCO Ammonia Bicarbonate ) 3
1 ------1.0 2.4 0.8 1.0 ------95 74 88 61 Total (as N) (00613) Nitrate alkalinity (mg/L as CaCO ------<.05 <.05 <.05 150 120 160 TDS 120 <0.05 Nitrite (00613) C) ------o ( 9.0 7.6 7.5 8.9 7.4 8.3 <.2 <.2 <.2 11.7 <0.2 (00950) Fluoride Temperature Temperature E. Coli, Escherichia coli . Units are milligrams per liter unless otherwise E. Coli, Escherichia C, degrees Celsius; --, not measured; <, less than; µS/cm, microsiemens C, degrees o 5 8 ------10 17 86 247 190 165 243 167 161 (µS/cm) Specific Chloride conductance conductance ; mL, milliliters; -1 3 ------pH 7.9 7.9 7.8 8.0 <.05 <.05 <.05 Total Total (units) <0.05 phosphorus , calcium carbonate; HCO 3 Date Date sampled sampled 05/07/04 11/25/03 05/07/04 11/25/03 05/07/04 05/07/04 11/12/03 11/12/03 11/25/03 11/25/03 05/07/04 05/07/04 11/25/03 11/25/03 9 9 6 6 9 9 9 9 Well depth Well Well depth Well (feet below (feet below land suface) land suface) type type Sand Sand Sand Sand Sand Sand Sand Sand Aquifer Aquifer Inorganic and physical constituents of water samples from selected wells in the Pulaski N. Y. study area, 2003–04. Inorganic and physical constituents of water samples from selected wells in the Pulaski N. Y.
well well USGS USGS county county number number Water samples were field-collected by USGS personnel; pH, specific conductance, and water temperatures were measured in field by USGS personnel; all constituent concentrations measured by Oswego by USGS personnel; all constituent concentrations measured Oswego temperatures were measured in field conductance, and water by USGS personnel; pH, specific samples were field-collected Water 1 do. do. OW 499 OW do. 499 OW do. OW 500 OW OW 500 OW OW 498 OW OW 498 OW do. do. OW 497 OW 497 OW Village of Pulaski Wells Village of Pulaski Wells Village Table 1. Table solids; CaCO TDS, total dissolved [USGS, U.S. Geological Survey; Samples to be analyzed for constituent concentrations were collected by Oswego County Deptartment of Health and analyzed by Upstate Laboratories, Inc. in Syracuse, N.Y. Locations are County Deptartment of Health and analyzed by Upstate Laboratories, Inc. in Syracuse, N.Y. Samples to be analyzed for constituent concentrations were collected by Oswego 2.] in figure shown per centimeter at 25 degrees Celsius; (00950), USGS National Water Information System (NWIS) parameter code; Celsius; (00950), USGS National Water per centimeter at 25 degrees by USGS personnel. temperature were measured in field conductance, and water properties were collected by USGS personnel. pH, specific samples to be tested for physical noted. Water County Department of Health. Hydrogeology of the Pulaski Water-Supply Area 11 ------2.3 4.1 2.3 1.5 E. Coli negative negative negative (colonies Potassium per 100 mL) .03 .02 3 9 ------0.02 <.02 24 100 mL) Manganese (colonies per Total coliform coliform Total ------<.2 <.2 <.2 3.5 5.9 4.3 4.9 <0.2 Bromide Magnesium ) 3 ------.44 -- .44 -- .18 -- Iron 0.46 CaCO filtered 100 Hardness (mg/L, as E. Coli, Escherichia coli . Units are milligrams per liter unless otherwise E. Coli, Escherichia C, degrees Celsius; --, not measured; <, less than; µS/cm, microsiemens C, degrees o 5.3 8.5 4.9 ------22 -- 32 -- 23 -- 27 12 Sodium Calcium ; mL, milliliters; -1 3 8 5 ------10 <5 2.7 3.0 2.3 2.5 Silica (00945) Sulfate , calcium carbonate; HCO 3 Date Date sampled sampled 11/25/03 05/07/04 11/25/03 05/07/04 11/12/03 05/07/04 11/12/03 05/07/04 11/25/03 11/25/03 11/25/03 05/07/04 11/25/03 05/07/04 9 6 9 9 9 6 9 9 Well depth Well Well depth Well (feet below (feet below land surface) land surface) type type Sand Sand Sand Sand Sand Sand Sand Sand Aquifer Aquifer Inorganic and physical constituents of water samples from selected wells in the Pulaski N.Y. study area, 2003–04.—Continued Inorganic and physical constituents of water samples from selected wells in the Pulaski N.Y.
well well USGS USGS county county do. do. 499 OW OW 500 OW do. 498 OW OW 497 OW number number do. do. 499 OW OW 500 OW do. 498 OW OW 497 OW Village of Pulaski Wells Village of Pulaski Wells Village Table 1. Table solids; CaCO TDS, total dissolved [USGS, U.S. Geological Survey; per centimeter at 25 degrees Celsius; (00950), USGS National Water Information System (NWIS) parameter code; Celsius; (00950), USGS National Water per centimeter at 25 degrees by USGS personnel. temperature were measured in field conductance, and water properties were collected by USGS personnel. pH, specific samples to be tested for physical noted. Water Locations are County Deptartment of Health and analyzed by Upstate Laboratories, Inc. in Syracuse, N.Y. Samples to be analyzed for constituent concentrations were collected by Oswego 2.] in figure shown 12 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York stream and production well and that water that flows into the Manganese was detected in three of the samples at production well likely comes from areas upgradient (east) concentrations ranging from <0.02 to 0.03 mg/L, and did not from the well, not from the stream (fig. 5). exceed the SMCL for manganese (0.05 mg/L). The relatively low concentrations of iron and manganese are typical for ground water in this area and are derived from the dissolution Inorganic Ions and Nutrients of these trace elements from minerals in the sand and Analyses for inorganic ions in samples collected on gravel deposits. November 12 and 25, 2003 (table 1) show that the dominant The predominant nutrient was nitrate as N, which cations were calcium and sodium. Calcium concentrations ranged in concentration from 0.8 to 2.4 mg/L; the USEPA ranged from 22 to 32 mg/L, and sodium concentrations ranged Maximum Contaminant Level (MCL) for nitrate in drinking from 4.9 to 12 mg/L, which is less than the USEPA Drinking water (10 mg/L as N) was not exceeded in any sample. Nitrite, Water Advisory of 60 mg/L. Magnesium and potassium ammonia, total organic nitrogen, orthophosphate, and total were detected at relatively low concentrations, from 3.5 to phosphorus were not detected in any sample. The relatively 5.9 mg/L and from 1.5 to 4.1 mg/L, respectively (table 1). low nutrient concentrations indicate minimal contamination by The dominant anions were chloride and sulfate in the sample fertilizer, manure, or septic-tank effluent in the source water analysis; chloride concentrations ranged from 5 to 17 mg/L, that flows to the Village of Pulaski production wells. and sulfate from less than 5 to 10 mg/L. Neither the SMCL for chloride (250 mg/L) nor that for sulfate (250 mg/L) was Bacteria exceeded. Bicarbonate was not analyzed, but estimated to be the major anion in the water samples from major-ion, Three of the four samples collected in November 2003 mass-balance calculations. were analyzed for total coliform bacteria and Escherichia Iron was detected in all four samples and concentrations coli (E. coli). Total coliform bacteria were detected in all ranged from 0.18 to 0.46 mg/L; the SMCL for iron three samples; colony counts ranged from 3 to 24 colonies (0.30 mg/L) was slightly exceeded in three of four samples. per 100 mL of water. E. coli were not detected in any sample
65
OW 493 Village of Pulaski production well (15 feet from stream) OW 501 piezometer (6 feet from stream) 60 Stream
55
50
45 TEMPERATURE, IN DEGREES FAHRENHEIT TEMPERATURE, 40
35 May 2003 Nov 2003 Mar 2004 July 2004 Nov 2004 DATE Figure 5. Temperature of water in Village of Pulaski dug well OW 493, piezometer OW 501, and adjacent stream from July 2003 through November 2004. (Locations of stream and wells shown on figure 2). Hydrogeology of the Pulaski Water-Supply Area 13
(table 1). Coliform bacteria are naturally present in the Register, National Archives and Records Administration, environment and by themselves do not indicate a health threat, 1990). Because production virtually ceased by 1996, but rather are used as an indication of the possible presence of measuring CFCs provides a means to date young water (less other potentially harmful bacteria such as E. coli, which come than 50 years from present) by measuring concentrations only from human and animal fecal waste. The USEPA MCL of CFC-11, CFC-12, and CFC-113 to identify ground for total coliform is zero; therefore, any detection of coliform water recharged since approximately 1941 (CFC-12), 1947 is considered a potential health threat. The Village of Pulaski (CFC-11), and 1955 (CFC-113). Ground-water dating with combines the water from all four production wells after it has CFC-11, CFC-12, and CFC-113 is possible because (1) the flowed by gravity toward the village. Once combined, the atmospheric mixing ratios of these compounds are known water is chlorinated to kill bacteria before being distributed to and (or) have been reconstructed over the past 50 years, the public. (2) the solubilities in water based on Henry’s Law are known, and (3) concentrations in air and relatively young water are relatively high and can be measured. Geochemical Age Dating Samples were collected from all four 2-in. piezometers in November 2003, but results from the analyses of these samples A numerical ground-water-flow model has been were questionable because of sampling field preparation and constructed by Malcolm Pirnie, Inc. (2000) to simulate stabilization errors. Three of the piezometers were resampled ground-water conditions in part of the Tug Hill glacial-drift in May 2004 to compare the values from the winter and aquifer near the Pulaski well field as part of a wellhead- spring samples. When the age estimates of all three tracers protection study. Results from the model indicate that the were not in agreement, the most reliable tracer (CFC-12) was time-of-travel for the longest ground-water-flow path, roughly used to determine the apparent age of the water (Plummer 7,000 ft from the easternmost edge of the aquifer to the Village and Busenburg, 1999). Similar ages were obtained from both of Pulaski production wells, is less than 1 year (Malcolm sample sets (table 2), though the May 2004 samples are more Pirnie, Inc. 2000). In contrast, concentrations of CFC and reliable because the sample bottles did not have substantial tritium in samples collected by the USGS to estimate the headspace. Apparent ages of the water samples based on CFC apparent recharge age of ground water in 28 community wells analyses range from approximately 25 years old to modern (in in Oswego County (Komor, 2002) indicate that at the Pulaski this study, the term modern means that the water is no older collection cistern, the estimated apparent age of ground water than 0 to 1 year from the date collected). is 24 years old. However, this result indicates a composite Older water (in this study, older water refers to water of different ages of water from each of the production older than mid-1970s), such as that collected from piezometer wells. In order to confirm the apparent age of ground water OW 500 (CFC ages estimated in the mid-1970s, table 2), determined by Komor (2002), samples were collected from indicates that the water is either from a deep flow system that 2-in. piezometers (OW 497, OW 498, OW 499, and OW 500) has relatively long flow paths within the aquifer or is water of upgradient from each production well and analyzed for mixed ages—some old and some young. Younger water, for concentrations of CFCs, SF , dissolved gases, and tritium 6 example from piezometer OW 499 (CFC ages estimated as (table 2). modern, table 2), suggests a source that has shorter ground- water-flow paths and quicker traveltimes, possibly from a Dissolved Gases relatively shallow flow system. The temperatures of recharge water calculated for the samples, and the gas concentrations they are based on, are Sulfur Hexafluoride given in table 2. The recharge temperatures were corrected for Sulfur hexafluoride (SF6) is a gas found in the atmosphere excess air and are considered reliable for the calculation of the that can be used to estimate the age of ground water that was CFC age. recharged from about 1970 to the present (Busenburg and
Plummer, 2000). Substantial production of SF6 began in the 1960s for use in high voltage electrical switches. SF can be Chlorofluorocarbons 6 especially helpful in determining the age of ground water CFCs are anthropogenic compounds developed in the that was recharged within the last 10-15 years as opposed 1930s and used in various industrial applications, including to using CFC concentrations, which have leveled off in very coolants, propellants, and solvents. Their production was recent precipitation, making it difficult to accurately determine discontinued by 37 nations in 1987 when they were found the age. Natural SF6 can be present in some crystalline and to play a part in the destruction of the ozone layer of the carbonate rocks, however, and can naturally increase the atmosphere (Elkins, 1999). In 1992, 90 countries agreed to concentrations of SF6 in the water, making the apparent age cut off production of CFCs, and by 1996, production of CFCs of the water too young. Each of the samples analyzed for SF6 ceased in the United States under Section 608 of the 1990 during this study produced unrealistic ages and were assumed revision of the Clean Air Act legislation (Office of the Federal to be contaminated, possibly by natural sources of SF6 in the 14 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York 3 3 to early 1980s 1990s Comments Mid 1970s Late 1970s Mid 1980s Early Modern Modern Late 1980s
2 ------(Tu) 20.66 20.97 20.75 Tritium concentration concentration
1 13.8 37.3 17.4 22.0 26.1 166.6 193.9 CFC-113 304.1 434.1 510.2 657.3 694.7 522.1 CFC-12 (in pptv) 1,205.2 Average calculated Average atmospheric mixing ratio 228.5 202.7 129.8 184.5 218.2 204.6 187.1 CFC-11 ------of 5.8 7.2 8.3 7.1 (°C) water Average Average recharge temperature 0 0 0 ------0.001 Methane 1 ------6.37 2.27 3.50 16.28 Carbon dioxide ------7.86 6.42 7.16 6.55 Oxygen ------Average concentration Average 0.754 0.788 0.713 0.761 Argon ------C, degrees Celsius; pptv, parts per trillion by volume; Tu, tritium units; CFC, chlorofluorocarbon; NP, not possible; --, measured; contamin., tritium units; CFC, chlorofluorocarbon; NP, Tu, parts per trillion by volume; Celsius; pptv, C, degrees o 19.168 19.944 22.911 21.326 Nitrogen Date sampled 11/25/03 05/07/04 11/12/03 11/25/03 05/07/04 05/07/04 11/25/03 9 6 6 9 9 9 9 Well depth Well (feet below (feet below land surface)
type Sand Sand Sand Sand Sand Sand Sand Aquifer Concentrations of tritium, chlorofluorocarbon (CFC), and air-temperature correction data for wells in the Pulaski, N.Y. study area, 2003-04. correction data for wells in the Pulaski, N.Y. Concentrations of tritium, chlorofluorocarbon (CFC), and air-temperature
Average of values for two samples. for two of values Average samples collected on September 1, 2004. Tritium that is less than 1 year old from the sampled date. is water Modern water well 1 2 3 USGS county number OW 499 OW do. OW 500 OW OW 498 OW do. contaminated. Units are milligrams per liter unless otherwise noted. Locations of wells are shown in figures 2 and 7.] in figures contaminated. Units are milligrams per liter unless otherwise noted. Locations of wells shown do. Table 2. Table mg/L, milligrams per liter; [USGS, U.S. Geological Survey; OW 497 OW Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 15 underlying bedrock. Therefore, no reliable information was Each sample from wells OW 497, OW 499, and OW 500 obtained from the SF6 analyses. collected for tritium analysis on September 1, 2004, contained approximately 21 Tritium Units (TU) (table 2). A comparison of this value to the decay lines in figure 6 shows that, because Tritium of multiple peaks in the tritium concentration in precipitation Tritium (a radioactive isotope of hydrogen – 3H) can be over the last 30 years, the decay line for 21 TU can intercept multiple possible years of recharge since 1970; thus, the used to determine the age of ground water relative to nuclear- tritium-derived date is only a general indication of the age of weapons testing during the 1960s because large amounts of these ground-water samples. The tritium concentrations in tritium were released into the stratosphere during this time samples from the three wells indicate that the water entered (Rozanski and others, 1991). Tritium in the stratosphere is the aquifer post-1970. incorporated into precipitation that falls worldwide, and the concentrations of tritium have been reconstructed for the years since the late 1950s. Therefore, tritium concentrations in ground water can be used to infer the age of the ground Hydrogeology of the Sandy Creek/ water (number of years since the precipitation entered the Lacona Water-Supply Area ground-water system as recharge, essentially removing it from the atmosphere). Once in the ground-water system, The Villages of Sandy Creek and Lacona obtain their tritium will continue to decay with a half-life of 12.4 water supply from well fields in the northeastern part of the years, so the decay must be taken into consideration when Village of Lacona (southern well field) and from a relatively determining the approximate age of the ground-water new well (northern well field) 1.25 mi north of the villages samples. The decay curves drawn from values for tritium in and near the northern border of Oswego County (figs. 1 precipitation for Ottawa, Canada (fig. 6, Ian Clark, University and 7), all of which are completed in the Tug Hill glacial-drift of Ottawa, Canada, written commun., 2006), show the aquifer. The wells are installed in nearshore glaciolacustrine paths along which tritium values decay over time within the beach, glaciofluvial outwash, and alluvial inwash deposits ground-water system. consisting mostly of silty sand and gravel; however, the
10,000 Tritium concentration in precipitation Tritium decay curves Tritium decay curve for 21 Tritium Units (note the many concentration peaks that are intercepted and many possible dates) 1,000 Values of about 21 TU for wells OW 497, OW 499, and OW 500 near Pulaski, N.Y.
100 IN TRITIUM UNITS (TU) CONCENTRATION OF TRITIUM, CONCENTRATION
10 1952 1962 1972 1982 1992 2002 YEAR
Figure 6. Tritium decay curves for Ottawa, Canada (modified from Clark, 2006) and value of tritium in wells OW 497, OW 499, and OW 500 sampled on May 7, 2004 near Pulaski, N.Y. 16 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
76°05'45" 76°03'15"
43°41'11"
Lindsey Creek
Lacustrine sediments grade from medium sand in the east to silty sand in the west
Northern Lindsey Creek Well Field OW436
OW490
OW502
B’
Little Sandy Creek OW348 Southern Well Field OW427 OW484 OW424 OW425 OW347 OW487 OW489 B OW485 OW486 { OW488
Sandy Creek 0 0.250.5 MILES 43°38'41" 0 0.50.25KILOMETERS Base from U.S. Geological Survey 1:24,000 Digital Line Graph, Geology from Miller, 1980a Universal Transverse Mercator projection, Zone 18N, NAD 1927
EXPLANATION GEOHYDROLOGIC UNITS IN THE TUG HILL NON AQUIFER GEOHYDROLOGIC UNITS GLACIAL-DRIFT AQUIFER LODGMENT TILL–Gravel and cobbles embedded in a matrix LACUSTRINE FINE TO MEDIUM SAND–Sand that was of clay, silt, and fine sand that were deposited at the base of deposited just off shore from proglacial Lake Iroquois the glacier; poorly sorted; compact and relatively impermeable beach ABLATION TILL–Gravel and cobbles embedded in a matrix BEACH SAND AND GRAVEL–Ridge of well-sorted coarse of clay, silt, and fine sand that were on top of and within the sand to cobble gravel deposited at the shoreline of glacier and laid down after the ice melted; poorly sorted; proglacial Lake Iroquois non-compact, and relatively impermeable OUTWASH AND INWASH SAND AND GRAVEL–Moderately BEDROCK–Ordovician age siltstone and shale sorted gravelly beds and sandy beds deposited by meltwaters from the glacier and inwash from PRODUCTION WELL FOR VILLAGES OF SAND CREEK AND streams draining the Tug Hill Plateau to the east OW487 LACONA, N.Y.–Well number assigned by U.S. Geological Survey NUMERICAL MODEL BOUNDARY (USGS) B B’ TEST WELL AND WELL NUMBER–Well number assigned by LINE OF GEOLOGIC SECTION–Section shown OW486 USGS in figure 8 DOMESTIC PRIVATE WELL AND WELL NUMBER–Well number APPROXIMATE LOCATION OF PROGLACIAL LAKE OW437 assigned by USGS IROQUOIS SHORELINE
Figure 7. Surficial and bedrock geology, model boundary, and production- and test-well locations, and location of line of geologic section in the Sandy Creek/Lacona, N.Y. study area. Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 17 outwash and inwash deposits contain lenses of silty fine-to- deposits are absent, and the outwash and inwash deposits form medium sand (figs. 7 and 8). In the western part of the aquifer, the aquifer (figs. 7 and 8). The aquifer material in the Sandy beach deposits overlie the outwash and inwash deposits, Creek/Lacona area was deposited about 14,000–15,000 years which, in turn, overlie a thin, clayey silt layer of till that is ago (Ridge, 2003) when Wisconsinan glaciers retreated from on top of shale. The till, where present, forms the bottom of the Tug Hill Plateau, producing meltwater and streams that the glacial aquifer; otherwise, the top of the shale forms the carried sediment from the plateau and deposited it as outwash bottom of the aquifer. In the eastern part of the aquifer, beach and inwash in the study area.
B B’ West East Approximate location of sand and gravel mine Tug Hill Plateau begins here and FEET extends eastward Ground water Losing 625 discharges Stream into drainage Ablation ditch OW 484 till 600 Drainage ditch
Test 575
Drainage ditch bor- ing Production well OW 487 OW 489 OW 485 550 Springs
525 Lodgment till
500 VERTICAL EXAGGERATION x20 0 500 1,000 FEET VERTICAL DATUM IS NGVD 1929 0 100 200 METERS EXPLANATION BEACH SAND AND GRAVEL–Ridge of well-sorted, ORIGINAL LAND SURFACE BEFORE SAND thinnly bedded, medium to coarse sand and fine AND GRAVEL MINING to medium gravel deposited at the shoreline of proglacial Lake Iroquois
OUTWASH SAND AND GRAVEL–Moderately sorted gravelly beds and sandy beds deposited by TEST WELL–Well number assigned by meltwaters from the glacier and inwash from U.S. Geological Survey (USGS) streams draining the Tug Hill Plateau to the east OW 489 TILL (LODGMENT OR ABLATION)–Gravel and cobbles embedded in a matrix of clay, silt, and fine sand that were deposited at the base of (lodgment in the west), or on top of (ablation in the east) the glacier; poorly sorted; compact and relatively impermeable SANDY CREEK AND LACONA PRODUCTION WELL–Well number assigned by USGS
BEDROCK–Ordovician age siltstone and shale OW 487
WATER-TABLE SURFACE–Shows altitude of water levels during average annual recharge water level conditions. Arrow shows general direction of ground-water flow screen
Figure 8. Generalized hydrogeologic section B–B’ near Sandy Creek/Lacona, N.Y. (Location of section on figure 7). 18 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
The beach deposits in the glacial aquifer consist of streams and ground-water-flow directions that cross the thinly interbedded, well-sorted, medium-to-coarse sand and narrow axis of the aquifer from east to west. Ground-water fine-to-medium gravel. The well-sorted and coarse-grained discharges to (1) springs, seeps, headwaters of streams, characteristics of the beach deposits result in the largest and wetlands along the western boundary of the aquifer, horizontal hydraulic conductivity of the deposits that form (2) production wells, (3) manmade ditches, and (4) deeply the Tug Hill glacial-drift aquifer. The beach unit thickness incised streams, such as Little Sandy Creek and Lindsey typically ranges from 10 to 20 ft. Creek in the southern and northern parts of the study area, Both outwash and inwash deposits consist predominantly respectively (fig. 7). Because ground water flows across of gravel with some sandy beds interspersed. The gravel is the narrow axis of the aquifer, flow paths and traveltimes typically coarser and less well sorted, and bedding is less of ground water from recharge to discharge areas are developed, than the overlying beach deposits. The poorly relatively short. sorted characteristics of this unit result in a lower horizontal Wells that tap the aquifer in the Sandy Creek/Lacona hydraulic conductivity than that of the beach deposits. The area are susceptible to contamination because the aquifer is outwash and inwash deposits typically range from 15 to 25 ft (1) unconfined (a protective confining layer such as clay is in thickness. absent), (2) highly transmissive, and (3) thin and narrow with ground-water-flow paths from recharge to discharge areas Hydrology (such as wells) that are relatively short because ground water crosses the short axis (width) of the aquifer, and (4) ground- The Tug Hill glacial-drift aquifer in the Sandy Creek/ water traveltimes from recharge areas to production wells Lacona area (fig. 1) is unconfined and thin, ranging in are relatively short and would allow relatively little time for thickness from 10 to 50 ft; however, the saturated thickness natural degradation of possible contaminants. typically ranges only from 10 to 25 ft and is generally thickest Two north-south trending ditches (20–30 ft deep) along in the western part of the aquifer (fig. 8). In the area east of the western edge of County Route 22 and east of the southern the Village of Sandy Creek (southeastern part of the study well field for the villages of Sandy Creek and Lacona (fig. 9) area), the aquifer is narrow (only about 0.5 mi wide) where the affect the hydrology of that area by locally intercepting beach sand and gravel deposits are located between the Tug the westward flow of ground water. Water in these ditches Hill Plateau on the east and till deposits on the west (figs. 1, 7, is routed by channels to the area just upgradient from the and 8). production wells where some water from these ditches seeps In the northern part of the study area, the aquifer is wider back into the aquifer and is captured by the production because lacustrine sand was deposited to the west of the beach wells (fig. 9). These ditches, in effect, act as short-circuits in deposits. The lacustrine sediments grade from medium sand the ground-water-flow system by shortening the route and in the eastern part of the aquifer near the former shoreline of decreasing the time that ground water travels from the eastern Lake Iroquois to silty sand in the western part of the study to western part of the aquifer. area. For modeling purposes, the western boundary of the Seepage measurements were made in the ditch (solid aquifer is considered to be where lacustrine sand grades into red line on fig. 9) on December 20, 2001, during a low-flow silty sand. period. Measurement results are shown in figure 9. The The aquifer is recharged by (1) precipitation that falls northernmost part of the ditch gained 0.46 ft3/s of water in directly over the aquifer, (2) unchannelized runoff from till an 800-ft-long reach beginning at the intersection of County and bedrock hills that border the eastern side of the aquifer, Route 22 and Center Road where flow began to enter the (3) losses from streams that drain the Tug Hill Plateau and discharge water into the aquifer, and (4) ground-water inflow ditch, south to where the ditch turns to the west, and where 3 from the till and bedrock on the adjoining Tug Hill Plateau. flow was measured at 0.46 ft /s (fig. 9). From County Route 3 Average annual recharge in the area from direct precipitation 22 to just north of the well field, flow increased from 0.46 ft /s 3 over the aquifer is estimated to be 28 in/yr on the basis of an to 0.81 ft /s. As water in the ditch flowed to the south, there 3 3 average annual precipitation of 47 in/yr minus the estimated was a measured loss of 0.36 ft /s (0.81 to 0.45 ft /s), indicating evapotranspiration losses of 19 in/yr (Kontis and others, 2004, that water lost from the ditch was recharging the aquifer pl. 1). Surface runoff in areas underlain by sand and gravel is directly east of the southern well field. This loss of water assumed to be minor; however, more runoff (and less recharge from the ditch indicates that some, or all, of this seepage to the aquifer) probably occurs where the lacustrine deposits loss is available to be captured by the production wells in in the northern part of the area grade westward from medium that area. During this same period, only a small amount of sand to silty fine sand. water was flowing in the southern ditch along County Route The aquifer geometry of the Sandy Creek/Lacona area 22 (dashed red line on fig. 9), but when ground-water levels is a lenticular deposit draped on a gently westward-sloping were higher, substantially more ground water was observed to bedrock surface that is the west flank of the Tug Hill Plateau be discharging into this ditch. During high-recharge periods, (fig. 8) and is relatively narrow (typically 0.5 mi wide) with ground water that is intercepted by the southern ditch flows Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 19
76°04' 76°03'30"
Sand and gravel Center Rd mine 43°39'15" 0.03 0.25 Railroad tracks 0.46 Flow Production- well field 0.81 (This reach of the natural channel (in blue),east of the sand and gravel mine, rarely receives ground-water inflow because the drainage ditches in the gravel pits are deeper than the stream channel.)
Flow 0.45
Sandy Village of Lacona County Route 22 Little Creek 0 0.25 0.5 MILE EXPLANATION 43°38'45" 0 0.25 0.5 KILOMETER Base from Oswego County 2-foot resolution color infrared Orthoimagery, NYS Office of Cyber Security & Critical Infrastructure Coordination, April 2003 EXPLANATION SEEPAGE AREA–Water from drainage ditch enters pond and then seeps into ground GROUND-WATER-FLOW PATH THAT ENDS IN PERENNIALLY FLOWING DRAINAGE DITCH IN GRAVEL PIT–Ditch receives ground-water inflow during recharge conditions. Arrow shows direction of ground-water flow GROUND-WATER-FLOW PATH THAT ENDS WITH GROUND WATER CAPTURED BY A MUNICIPAL PRODUCTION WELL DURING LOW AND AVERAGE RECHARGE CONDITIONS–Ditch receives ground-water inflow during average-to-high recharge conditions. Arrow shows direction of ground-water flow GROUND-WATER-FLOW PATH FROM DITCH AND FROM SURFACE-WATER SEEPAGE AREA–Water from losing reach of drainage ditch and seepage area seeps back into aquifer and flows toward Village of Lacona production wells DRAINAGE DITCH Water seeps into ditch from aquifer to the east Losing reach (water discharges from the ditch into the sand and gravel aquifer) Gaining reach (water discharges from the aquifer into the ditch) Water in channel exits past well field to the west UNNAMED EPHEMERAL STREAM–Tributary to Little Sandy Creek VILLAGE BOUNDARY SANDY CREEK/LACONA PRODUCTION WELLS U.S. GEOLOGICAL SURVEY NESTED PIEZOMETERS 0.45 SEEPAGE MEASUREMENT SITE–Discharge measurements made on December 20, 2001. Number is discharge, in cubic feet per second.
Figure 9. Streamflow measurements, ground-water-flow directions, surface-water seepage area, and production wells in the southern part of the Sandy Creek/Lacona aquifer, N.Y. 20 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
450 ft to the west where it empties into a ponded area, seeping ranged from 5.2 to 5.6 mg/L. Magnesium and potassium into the ground and reentering the ground-water system. were detected at relatively small concentrations, from Results of a 24-hour aquifer test conducted by USGS 4.0 to 5.1 mg/L and from 0.8 to 1.3 mg/L, respectively. in the Sandy Creek/Lacona southern well field in November Concentrations of the anions chloride and sulfate were small, 1988 indicated that the estimated mean horizontal hydraulic ranging from 3 to 4 mg/L and from 7 to 9 mg/L, respectively. conductivity of the beach deposits was 1,600 ft/d (Zarriello, Bicarbonate (estimated from alkalinity) was the dominant 1993). Horizontal hydraulic conductivities were calculated from drawdown data from four wells in the beach deposits. anion in all samples from the sand and gravel aquifer and These calculated horizontal hydraulic conductivity values bicarbonate and sulfate were the dominant anions in water ranged from 480 to 4,200 ft/d (Zarriello, 1993). The horizontal from the bedrock. Unlike ground water from the Pulaski well hydraulic conductivity of the inwash and outwash deposits field where fluoride was not detected, fluoride was detected east of the beach deposits was estimated to be 120 ft/d, which in all samples from the Sandy Creek/Lacona well field at is less than that of the beach deposits because of the finer- concentrations ranging from 0.2 to 1.0 mg/L, indicating that grained nature of the inwash and outwash deposit sediments. the mineral fluoride may be present locally in the sediments The lacustrine deposits in proximity to the ancient comprising the aquifer; however, the concentrations did not shoreline (beach) deposits consist of medium sand that grade exceed the SMCL (2 mg/L). westward to mostly silty fine sand. The horizontal hydraulic Results of analyses for inorganic constituents (table 3) conductivities of the lacustrine sediments were estimated to grade from 80 ft/d where sediments consist of coarse sand in indicate that there are differences in quality between water the off-shore area of the beach to 8 ft/d where sediments grade from the bedrock and water from the sand and gravel in the to fine sand in the western part of the area. Tug Hill glacial-drift aquifer. For water samples collected from the bedrock well (OW 485) at the Sandy Creek/Lacona well field, the dominant cation was sodium. For water samples Water Quality collected from the sand and gravel aquifer, the dominant Four test wells, OW 485, OW 486, OW 488, and OW anion was calcium. The sodium concentration of 63 mg/L 489 (fig. 7), were installed in the Sandy Creek/Lacona study represented in samples from the bedrock aquifer exceeded the area to aid in construction of a numerical ground-water-flow USEPA Drinking Water Advisory for sodium of 60 mg/L. model (see section “Model Design and Model Grid”) and to In water from the bedrock, the sulfate concentration was provide access for the collection of water-quality samples. relatively large (35 mg/L) compared with that in the sand and Results from physical and inorganic analyses of water gravel aquifer (7 to 9 mg/L); the SMCL for sulfate (250 mg/L) samples collected during November 28 and 29, 2001, from was not exceeded. The sodium-sulfate water at OW 485 likely wells OW 486, OW 487, and OW 488 that tap the sand and reflects the chemistry of the shale bedrock in which the well gravel of the Tug Hill glacial-drift aquifer (fig. 10) show that is finished. Iron was detected in all four samples from the the ground water in the unconsolidated deposits is potable (table 3). Relatively low concentrations of some major ions Sandy Creek/Lacona wells, and concentrations ranged from and nutrients in all samples indicate that agricultural fertilizer 0.03 to 0.17 mg/L. The SMCL for iron (0.30 mg/L) was not applications, septic-tank effluent, and the dissolution of road exceeded in any sample. The iron concentration in water from salt minimally affect the ground-water quality in wells at the the bedrock aquifer was slightly greater than that from the Sandy Creek/Lacona southern well field. sand and gravel aquifer. Manganese was detected in two of the four samples; the concentration in the sample from well Physical Properties OW 486 (sand and gravel) was 0.11 mg/L, which exceeded the SMCL for manganese (0.05 mg/L). The concentration in Field pH of the samples from the Tug Hill glacial-drift the bedrock well (OW 485) was 1.3 mg/L. The relatively low aquifer ranged from 6.9 to 7.7, and specific conductance of the concentrations of iron and manganese are typical for ground samples ranged from 205 to 345 µS/cm at 25°C (table 3). The water and are derived from the dissolution of these trace sample from the bedrock well (OW 485) had a pH of 8.2 and a elements from minerals in the sand and gravel deposits. specific conductance of 330 µS/cm (table 3). In the sand and gravel in the Tug Hill glacial-drift aquifer, the predominant nutrient was nitrate, which ranged Inorganic Ions and Nutrients in concentration from less than 0.5 to 0.6 mg/L as N; the nitrate MCL of 10 mg/L as N was not exceeded in any sample. Results from the analyses for inorganic ions in water samples collected from three wells (OW 486, OW 487, and Nitrate was not detected in water from the bedrock well. OW 488) that tap the sand and gravel aquifer (table 3) show Nitrite, ammonia, total organic nitrogen, and orthophosphate that the dominant cation was calcium; calcium concentrations were not detected in any sample. Total phosphorus was ranged from 36 to 55 mg/L, and sodium concentrations detected in one sample at 0.11 mg/L (OW 486). Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 21
SANDY CREEK/LACONA WELL FIELD, TUG HILL GLACIAL-DRIFT AQUIFER
U.S. Geological Survey test wells
Municipal Shallow Elevation Deep well above well #4 well Intermediate NVGD 1929 OW 487 OW 488 well OW 486 OW 485 (in feet) (S&G) (S&G) (S&G) (bedrock) Land surface 553 0
Sand and gravel (S&G) Water Level = Water Level = 543.07 543.76 Water Level = 11/20/01 11/20/01 542.53 11/20/01 15 538
Coarse sand and aquifer pebbly sand Unconfined Depth = 20 ft 529 11.7 yr 24
Cobble to bouldery silty gravel Depth = 29 ft 520 33 15.0 yr Till, gray, silty clay Depth = 34 ft matrix with embedded DEPTH BELOW LAND SURFACE, IN FEET DEPTH BELOW LAND SURFACE,
unit 13.6 yr Con- stones 515 fining 38
Bedrock (siltstone and shale of Ordovician age)
47 496 Depth = 45 ft 32.0 yr EXPLANATION 13.6 yr Apparent age of ground water determined from concentrations of chlorofluorocarbons in water
Screened interval
Figure 10. Wells and apparent ages of ground water determined by concentrations of chlorofluorocarbons in the villages of Sandy Creek and Lacona’s well field near Lacona, N.Y. 22 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
-- -- <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 Total Total Ortho- kjeldahl nitrogen phosphate ) -1 3 -- -- as 183 159 183 244 <0.5 <0.5 <0.5 <0.5 (mg/L (as N) HCO Ammonia Bicarbonate
) 1 3 -- -- 0.6 0.5 0.6 150 130 150 200 Total (as N) <0.2 CaCO Nitrate (mg/L as alkalinity 1 -- -- 210 170 230 240 TDS <0.5 <0.5 <0.5 <0.5 Nitrite Constituents Constituents -- C) -- o 8.3 9.0 1.0 0.2 0.2 0.3 ( 10.1 11.6 Fluoride Temperature Temperature C) o 6 3 4 3 -- 330 290 345 340 205 (µS/cm at 25 Specific Chloride conductance conductance pH -- Total Total 8.2 0.11 7.7 7.4 7.4 6.9 (units) E. Coli, Escherichia coli . Units Information System (NWIS) parameter code; E. Coli, Escherichia , bicarbonate; (00945), USGS National Water <0.50 <0.50 <0.50 -1 phosphorus 3 Date Date 08/16/02 11/28/01 11/28/01 11/28/01 sampled 11/29/01 sampled 08/16/02 11/28/01 11/28/01 11/28/01 11/29/01
, calcium carbonate; HCO 3 23 23 45 45 29 29 19 19 34 34 Well depth Well Well depth Well (feet below (feet below land suface) land suface) type type S&G S&G S&G S&G S&G S&G S&G S&G Shale Shale Aquifer Aquifer Inorganic and physical constituents of ground water from wells in the Villages of Sandy Creek and Lacona, N.Y. study area. of Sandy Creek and Lacona, N.Y. Inorganic and physical constituents of ground water from wells in the Villages
well well USGS USGS county county number number OW 484 OW OW 484 OW OW 485 OW OW 485 OW OW 486 OW OW 486 OW OW 488 OW OW 488 OW OW 487 OW OW 487 OW Water samples were field-collected by USGS personnel; pH, specific conductance, and water temperatures were measured in field by USGS personnel; all constituent concentrations measured by Oswego by USGS personnel; all constituent concentrations measured Oswego temperatures were measured in field conductance, and water by USGS personnel; pH, specific samples were field-collected Water 1 C, degrees Celsius; C, degrees centigrade; TDS, total dissolved solids; mg/L, milligrams per liter; mL, milliliters; --, not measured; µS/cm, microsiemens per centimeter at 25 degrees Celsius; S&G, sand solids; mg/L, milligrams per liter; mL, milliliters; --, not measured; µS/cm, microsiemens centimeter at 25 degrees centigrade; TDS, total dissolved Celsius; C, degrees C, degrees o are milligrams per liter unless otherwise noted. Locations are shown in figs. 2 and 7.] in figs. are milligrams per liter unless otherwise noted. Locations shown Table 3. Table [ and gravel; <, less than; Temp, temperature ; CaCO <, less than; Temp, and gravel; County Department of Health. Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 23
4.4 0.9 0.8 1.3 -- -- <0.5 <0.5 <0.5 <0.5 Caffeine Potassium ------1.3 0.11 <0.02 <0.02 E. Coli (colonies per 100 mL) Manganese 1.3 4.7 4.0 5.1 ------100 mL) Magnesium (colonies per Total coliform coliform Total 1 -- -- 0.17 0.06 0.05 0.03 <.2 <.2 <.2 <.2 Iron Bromide Constituents Constituents ) 3 ------36 42 55 8.3 CaCO filtered Calcium Hardness (mg/L, as -- 9 5.6 7 5.4 8 5.2 -- 35 63 (00945) Sulfate Sodium 2.9 2.2 2.0 2.4 -- -- <0.5 <0.5 <0.5 <0.5 Total Total Silica organic nitrogen E. Coli, Escherichia coli . Units Information System (NWIS) parameter code; E. Coli, Escherichia , bicarbonate; (00945), USGS National Water -1 3 Date Date sampled 08/16/02 08/16/02 11/28/01 11/28/01 11/28/01 11/28/01 11/28/01 11/28/01 sampled 11/29/01 11/29/01 , calcium carbonate; HCO 3 23 23 45 45 29 29 19 19 34 34 Well Well Well Well depth depth (feet below (feet below land surface) land surface) type type S&G S&G S&G S&G S&G S&G S&G S&G Shale Shale Aquifer Aquifer Inorganic and physical constituents of ground water from wells in the Villages of Sandy Creek and Lacona, N.Y. study area.—Continued of Sandy Creek and Lacona, N.Y. Inorganic and physical constituents of ground water from wells in the Villages
USGS USGS county county OW 484 OW OW 484 OW OW 485 OW OW 485 OW OW 486 OW OW 486 OW OW 488 OW OW 488 OW OW 487 OW OW 487 OW Water samples were field-collected by USGS personnel; pH, specific conductance, and water temperatures were measured in field by USGS personnel; all constituent concentrations measured by Oswego by USGS personnel; all constituent concentrations measured Oswego temperatures were measured in field conductance, and water by USGS personnel; pH, specific samples were field-collected Water well number well number 1 C, degrees Celsius; C, degrees centigrade; TDS, total dissolved solids; mg/L, milligrams per liter; mL, milliliters; --, not measured; µS/cm, microsiemens per centimeter at 25 degrees Celsius; S&G, sand solids; mg/L, milligrams per liter; mL, milliliters; --, not measured; µS/cm, microsiemens centimeter at 25 degrees centigrade; TDS, total dissolved Celsius; C, degrees C, degrees o are milligrams per liter unless otherwise noted. Locations are shown in figs. 2 and 7.] in figs. are milligrams per liter unless otherwise noted. Locations shown Table 3. Table [ and gravel; <, less than; Temp, temperature ; CaCO <, less than; Temp, and gravel; County Department of Health. 24 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Geochemical Age Dating collected from well OW 488 (water depth of 20 ft and finished in the upper part of the sand and gravel aquifer) to In the investigation by Komor (2002), CFCs and tritium 15.0 years for water collected from well OW 486 (well depth were used to estimate the apparent recharge age of ground of 29 ft and finished in the lower part of the sand and gravel water in 28 community wells in Oswego County. A numerical aquifer) (fig. 10). Production well OW 487 (depth 34 ft) had ground-water-flow model was constructed previously by the an apparent age of 13.6 years, which is roughly the average USGS to simulate ground-water conditions in part of the of the ages of water from two piezometers (OW 486 and OW Tug Hill glacial-drift aquifer near the Sandy Creek/Lacona 488) that were analyzed. The 20-ft-long screen in production southern well field as part of a wellhead-protection study well OW 487 is open to nearly the entire saturated thickness (Zarriello, 1993). Model results indicated that the time for of the aquifer, and therefore, induces water throughout this ground water to cross the 0.5-mi-wide aquifer and enter the length to flow into the well. This results in a mixed water well field generally would take less than 500 days. From a sample consisting of water from all parts of the aquifer along water sample collected in August 1999 from Sandy Creek/ the screened length. This is probably why the apparent age Lacona production well OW 487, the estimated apparent age of 13.6 years is roughly the average of the samples from the of ground water was 17 years, but the USGS CFC laboratory two piezometers. noted in the analysis results that the water was likely a mixture The sample collected from test well OW 485 finished of ground water from sources having different ages. To resolve in bedrock (depth 45 ft) had an apparent age of 32 years, or the discrepancy between the CFC date of 17 years and model nearly twice as old as water in the sand and gravel aquifer. dates of less than 500 days, additional samples for analysis Substantial differences in apparent ages and concentrations for CFCs and dissolved gases were collected from the aquifer, of several common ions and nutrients (tables 3 and 4) and a more detailed numerical ground-water-flow model was constructed. between water in the sand and gravel aquifer and water in the bedrock unit indicate that little or no water pumped from the production well (OW 485) is derived from bedrock. Dissolved Gases Apparently, the relatively impermeable till layer on top of the bedrock confines water in the rock and restricts flow of For this investigation, five additional samples for ground water between the sand and gravel and bedrock. This is dissolved gases were collected from the Sandy Creek/Lacona supported by the lower hydraulic head (542.53 ft in OW 485) well field. The recharge temperatures calculated from the in the bedrock aquifer compared with the hydraulic head in dissolved gases for samples from the sand and gravel aquifer the overlying sand and gravel aquifer (543.07 and 543.76 ft ranged from 7.8 to 9.2°C. The recharge temperature for the in wells OW 488 and OW 486, respectively) indicating a water sample from the bedrock aquifer was 10.3° C (table 4). The recharge temperatures in table 4 are corrected for excess downward vertical gradient between the sand and gravel air (air-temperature correction data) and are considered aquifer and bedrock. The results of the selected chemical reliable for the calculation of the CFC age. analyses and hydraulic head relations negate the hypothesis that older water from underlying bedrock mixes with young water in the sand and gravel aquifer resulting in mixed age Chlorofluorocarbons water in the production well that is older than that predicted by time-of-travel computations of a numerical ground-water-flow Three CFC isotopes were measured to determine the model (Zarriello, 1993). apparent ages of four ground-water samples collected during Typically, one would expect that ground water would fall 2001 (OW 486, OW 487, and OW 488) and one ground- water sample collected during summer 2002 (OW 485) become progressively younger eastward (upgradient) from (figs. 7 and 10, and table 4). For the samples collected at the the well field near Lacona, N.Y., because it has been in Sandy creek/Lacona well field, results of the analyses for the system for a shorter time and has recharged the aquifer CFC-12 and CFC-113 were in close agreement and were used more recently from precipitation falling on the land surface; to determine the apparent age of the water. Values of CFC-11 however, the CFC results from test well OW 484 (finished typically resulted in ages older than that indicated from the at the bottom of the sand and gravel unit, depth 23 ft) in the CFC-12 and CFC-113 analyses, indicating that CFC-11 may central part of the aquifer (figs. 7 and 8) indicate that the water have been degraded; therefore, results of analyses for CFC-11 from well OW 484 is older than water from the well field. were ignored. The sample collected from well OW 484 had an apparent age At the Sandy Creek/Lacona southern well field, the of 17.4 years (table 4). The reasons for this discrepancy are apparent ages of ground water (at the time of sampling in discussed later in the section “Comparison of Geochemical 2001) determined from averaged CFC-12 and CFC-113 Age Dating and Ground-Water Traveltimes Computed from concentrations (table 4) ranged from 11.7 years for water Numerical Modeling”. Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 25 13.6 11.7 15 32 17.4 (in years) from averaging concentrations CFC-11 and CFC-113 (since sampled date) Apparent age of water 32.1 56.9 56.9 82.8 296.4 CFC-113 atmospheric 1 463.1 502.3 434.2 152.5 252.3 CFC-12 mixing ratio (in pptv) 48.6 141.3 184.5 131.9 479.3 Average calculated Average CFC-11 (2) 7.8 8.9 9.2 NP NP NS (°C) 1997 10.3 CFC-113 Average Average Contam. Contam. recharge temperature temperature
4 .00 .00 (1) NP NS CH 0.00 (mg/L) 12.39
1987.7 1983.0 1992.0 1 CFC-113 CFC-113 Contam.
2 ------(2) 4.61 0.77 NS CO 1 15.23 10.33 (mg/L) CFC-12
2 .00 O (1) NS 2.76 2.11 0.77 1988 1989 1987 1971 1983 (mg/L) CFC-12 Average model piston-flow dates Average Average concentration Average Ar .73 .73 .72 (2) NP NP NP NP NS 1994 0.73 (mg/L) CFC-11 (Dual dates are possible for CFC-11, CFC-12, and CFC-113) (1) N2 NS 1977 1982 1976 1969 1980 20.12 20.50 20.60 20.91 (mg/L) CFC-11 Date Date sampled sampled 12/3/2001 12/3/2001 8/16/2002 C, degrees Celsius; pptv, parts per trillion by volume; CFC, chlorofluorocarbons; NP, not possible; NS, sampled; Contam., contaminated. Units are milligrams per liter unless CFC, chlorofluorocarbons; NP, parts per trillion by volume; Celsius; pptv, C, degrees 11/27/2001 11/30/2001 o 11/27/2001 11/30/2001 12/3/2001 12/3/2001 8/16/2002 type type S&G S&G S&G S&G S&G S&G S&G S&G Shale Shale Aquifer Aquifer Tritium, chlorofluorocarbon (CFC), and air-temperature correction data for wells in the Sandy Creek and Lacona, N.Y. study area. correction data for wells in the Sandy Creek and Lacona, N.Y. chlorofluorocarbon (CFC), and air-temperature Tritium,
well well USGS USGS OW 487 OW 488 OW OW 486 OW OW 485 OW OW 484 OW OW 487 OW OW 488 OW OW 486 OW OW 485 OW OW 484 OW county county Average of values for two samples. for two of values Average number number 1 otherwise noted. Locations are shown in figs. 2 and 7.] in figs. otherwise noted. Locations are shown Table 4. Table [mg/L, milligrams per liter; 26 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Sulfur Hexafluoride Boundary Conditions
Each of the samples analyzed for SF6 during this study Several types of boundary conditions were specified in were contaminated, probably by natural sources of SF6, and the model (figs. 11 and 12). Natural boundaries were used produced unrealistic ages. Therefore, no reliable information where possible, but an arbitrary boundary was used to limit was obtained from this environmental tracer. the northwestern extent of the model because the aquifer sediments progressively decrease in grain size (from coarse sand to fine sand) westward from the beach deposits in Simulation of Ground-Water Flow the east. A three-dimensional, finite-difference, ground-water-flow model (MODFLOW) developed by the USGS (McDonald No-Flow Boundaries and Harbaugh, 1988) was used to simulate ground-water flow No-flow boundaries were assigned to (1) the contact in the Sandy Creek/Lacona study area. Contributing areas to between the sand and gravel aquifer and the upland till on the the production wells then were calculated using MODPATH Tug Hill Plateau, and (2) the till drumlins to the west of the (Pollock, 1989), a semi-analytical, particle-tracking program aquifer (figs. 7, 11, and 12) because flow between these units that uses heads and fluxes predicted through steady-state was assumed to be negligible compared to the flow within the simulation by the ground-water model to trace the flow sand and gravel aquifer. Also, a no-flow boundary was used lines in the aquifer. A pre- and post-processing program, between the aquifer and underlying till and bedrock (fig. 12). Groundwater Modeling System (GMS) version 6.0 (build date Bedrock was not included as a geohydrologic unit in this 2006), was used to input data for MODFLOW and MODPATH model because of the accepted assumption that shale transmits and generate output graphics and statistical analyses. The relatively little ground water in comparison to the overlying calibrated model then was used to determine contributing sand and gravel units in this area. areas under various pumping scenarios and to facilitate the interpretation of the geochemical age-dating data collected for Specified-Flux Boundaries this study. A specified-flux boundary was used to represent areal recharge from precipitation that falls directly on the aquifer Model Design and Model Grid (fig. 12). Where the aquifer consists of very permeable sand and gravel and medium-to-coarse grained sand, surface runoff Previously, Zarriello (1993) constructed a two- was assumed to be negligible. Average annual recharge from dimensional flow model (modeled as one layer with a grid of direct precipitation over the aquifer is estimated to be 27 in/yr 34 rows and 25 columns) using MODFLOW and MODPATH on the basis of an average annual precipitation of 46 in/yr to represent 1.19 mi2 of the sand and gravel aquifer in the minus the estimated evapotranspiration losses of 19 in/yr Sandy Creek/Lacona area. For this investigation, a three- (Kontis and others, 2004, pl. 1). However, in the northwestern dimensional, five-layered model was used to represent the part of the study area, the lacustrine sediments grade from aquifer and was constructed also using MODFLOW and coarse sand near the former shoreline of proglacial Lake MODPATH; however, the modeled area was enlarged to Iroquois to progressively finer and less permeable sand (fine 4.8 mi2 to include a new production well (OW 490) that and very fine sand) to the west (fig. 7). In the area underlain was constructed for the villages about a mile north of the by fine and very fine sand, not all precipitation infiltrates into existing well field (figs. 7 and 11). The model grid included the soil because of the less permeable nature of the fine- 99 rows and 91 columns in each of the five layers (each layer grained sediments. In these areas underlain by fine sand, some contained 4,642 active cells) for a total of 23,210 active precipitation runs off the land surface and flows into streams cells with a uniform cell size of 164 ft (50 m) by 164 ft. resulting in less recharge to the aquifer in these areas than in Five layers were used to increase the vertical resolution of those underlain by coarse sand, and sand and gravel. It was the aquifer and to permit ground-water-flow paths to have estimated by best-fit procedure in the calibration process vertical-flow components. The grid was oriented such that that the areas underlain by fine sand receive approximately the columns were aligned with the north-south axis of the 11 in/yr of recharge from precipitation that falls directly over aquifer and the rows were aligned roughly with the east-to- the aquifer. west direction of ground-water flow (fig. 11). The conceptual A specified-flux boundary was used to simulate recharge model representation is shown in figure 12. The horizontal and from surface runoff and ground-water seepage from adjacent, vertical hydraulic conductivities of the sand and gravel aquifer unchanneled uplands into the aquifer in layer 1 of the model were assumed to be isotropic (equal physical properties along (figs. 11 and 12). The drainage area of the unchanneled rows and columns and between layers). hillsides was delineated from USGS 7.5-minute digital Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 27
76°05'45" 76°03'15"
43°41'11"
Lindsey Creek
OW 436
OW 502 OW 490
C’ OW 348
OW 484 OW 427 OW 424 OW 347 OW 487 C OW 489
OW 486 485 549
0 0.25 0.5 1 MILE Little Sandy Creek
43°38'41" 00.25 0.5 1 KILOMETER
Base from U.S. Geological Survey 1:24,000 Digital Line Graph, Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION SPECIFIED-FLUX CELL–Simulates recharge to the aquifer along the DRAIN CELL–Simulates discharge from the aquifer by small eastern boundary from unchannelized surface flow on till-covered streams and springs, mainly on the western aquifer boundary bedrock on the Tug Hill Plateau and along the face of aquifer where gravel mining has removed PRODUCTION WELL CELL–Simulates discharge from production material well screened for the entire thickness of aquifer C C’ STREAM CELL–Simulates stream losing or gaining water to the LINE OF CONCEPTUAL SECTION SHOWN IN FIGURE 12 aquifer as a head-dependent flux boundary WELL USED FOR CALIBRATION–Water-level elevation used to calibrate ACTIVE CELLS–Thicker line is model boundary flow model. Well data shown in table 5a
Figure 11. Three-dimensional-flow model cells, simulated cell types, and location of conceptual section C–C’ for the Sandy Creek/Lacona, N.Y. study area. (Conceptual section in figure 12). 28 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York No-flow boundary Hill (between Tug upland till and sand and gravel aquifer) East
C’ Recharge from unchanneled runoff from Hill the Tug Plateau Losing stream Losing ACTIVE CELLS–Darker line represents model boundary RECHARGE TO THE AQUIFER DISCHARGE FROM THE AQUIFER INACTIVE CELL–aquifer absent or impermeable till present from surface to bedrock stream Losing C–C’ shown in figure 11). Ground water drainage ditch the aquifer into discharges from fully penetrating along aquifer face mining Aquifer removed
by gravel
EXPLANATION
Gaining stream Gaining Bottom of unconsolidated aquifer (till or bedrock), no-flow model boundary model no-flow bedrock), or (till aquifer unconsolidated of Bottom
Recharge from direct precipitation and snowmelt on land surface
Losing stream Losing
Fully penetrating production well production penetrating Fully Springs NOT TO SCALE SPECIFIED-FLUX CELL–Simulates recharge to the aquifer along eastern boundary from unchannelized surface flow on till-covered Hill Plateau bedrock on the Tug PRODUCTION WELL CELL–Simulates discharge from municipal production well screened for the entire thickness of aquifer STREAM CELL–Simulates stream losing water to the aquifer as a head-dependent flux boundary DRAIN CELL–Simulates discharge from the aquifer by small streams and springs, mainly on the western aquifer boundary and along face of aquifer where gravel mining has removed material 1 2 3 4 5 C Conceptual generalized, cross-sectional representation of ground-water-flow model showing cell types and directions of recharge Conceptual generalized, cross-sectional representation of ground-water-flow West
Model Layers Figure 12. area. (Line of section discharge to the aquifer in Sandy Creek and Lacona, N.Y. No-flow boundary (between till drumlins and sand gravel aquifer) Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 29
line graphs, and the recharge from these upland areas was Hs = head in the stream (ft); calculated from equation 1. The calculated recharge from the Ha = head in the aquifer (ft); unchanneled upland area was then divided by the number of and bordering active model cells to obtain the recharge rate for C = streambed conductance (ft2/d), defined as each cell. The daily amount of recharge from runoff from str the vertical hydraulic conductivity of the adjacent unchanneled hillsides can be calculated by the streambed, times the width of the stream following equation: reach, times its length, divided by the PETDA thickness of the streambed. R ( - ) ´ , (1) = t Streams are represented only in layer 1 of the model. where Each stream cell was assigned values for the following: R = average annual rate of recharge from runoff (1) layer, row, and column number; (2) average water-surface from unchanneled hillsides, in ft3/d; elevation (estimated from USGS digital elevation models, P = average annual precipitation of 46 in. or DEMs); (3) streambed conductance; and (4) elevation (3.8 ft); of bottom and top of the streambed (bottom estimated by ET = average annual evapotranspiration (1.5 ft); subtracting an average depth from top elevation from DEM). DA = drainage area of hillside; An initial value of 5 ft/d was estimated for the vertical hydraulic conductivity of the streambed, and an assumed and streambed thickness of 1.5 ft was used to compute streambed t = 365 days. conductances. Streambed conductance values ranged from A specified-flux boundary, represented by discharging 10 ft2/d for the small streams to as much as 25 ft2/d for the wells (pumped wells) was used to represent the municipal largest stream in the study area (Little Sandy Creek). production wells (fig. 12). During 2005, pumping was Little Sandy Creek forms the southern boundary of the alternated between the two well fields on a daily basis, modeled area where the stream flows on till in the eastern part therefore, total ground-water withdrawal was equally divided of the study area and then flows on bedrock in the western part between the two well fields (Jill Mattison, Villages of Sandy of the study area (fig. 7). Lindsey Creek forms the northern Creek/Lacona Town Clerk, oral commun., 2005). Total boundary of the model, where in the eastern part of the aquifer daily average ground-water use for the villages is about it loses water to the aquifer. The creek then flows on bedrock 270,000 gal/d. in the central reach before it flows on lacustrine sediments in the western part of the aquifer. Both creeks were represented Head-Dependent Flux Boundaries in the model by the river package (fig. 11). Typically, streams that originate in the Tug Hill Plateau The Drain Package in MODFLOW was used to simulate are major sources of recharge to the Tug Hill glacial-drift ground-water seepage into a manmade ditch and to the many aquifer because they lose water as they transverse the eastern springs and headwaters that originate along the western margin part of the permeable sand and gravel aquifer (Miller and of the aquifer (figs. 7 and 11). Drains are head-dependent others, 1989). The River Package in MODFLOW was used flow boundaries similar to stream boundaries except they only to simulate flow of water between the aquifer and streams, permit water to discharge from the aquifer. Drains are only including the recharge from upland tributaries that lose water active when the water table in the aquifer rises up to the level where they flow over the aquifer. The river package also was of the drain. The heads in drain cells that represent headwater used to represent the reach of manmade ditches where it was streams and springs were determined from land-surface uncertain whether the ditches gained or lost water. The reach elevations from USGS 1:24,000-scale topographic maps. of the manmade ditch that fully penetrates the aquifer was The head in drain cells that represent the manmade ditch was represented by a drain (fig. 12). determined through elevation levels run for this investigation. Leakage between streams and the aquifer depends on the head difference between the stage (water surface) of Model Calibration the stream and the water table of the aquifer adjacent to the stream and is adjusted according to a streambed-conductance The model was calibrated to average annual recharge term. The amount of leakage is computed using Darcy’s Law, conditions. Simulated heads were compared to measured as follows: heads in 11 wells (table 5a) and simulated streamflow was compared to measured streamflow in Lindsey Creek (table 5b) Q = C (H – H ) , (2) str s a in the northern part of the study area and in the sand and where gravel quarry ditches in the southern part of the modeled Q = leakage to or from the aquifer through the area (fig. 9). Most of the available data used to calibrate the streambed (ft3/d); model were for the southern part of the study area where this 30 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Table 5a. Difference between measured and simulated heads (3) channel elevations of streams and springs that represent at 11 selected non-pumped wells in the Sandy Creek/Lacona, the intersection of the water table and land surface. Model N.Y. model study area. calibration was considered satisfactory where (1) simulated heads were within 3 ft of the measured heads, (2) the square [ft3/s, cubic feet per second; RMS, root mean square; well head difference root of the arithmetic mean of the squares (root mean values are observed head minus simulated head, in feet. Well locations square—a statistical measure of the magnitude of a varying shown in figs. 7 and 11] quantity that is especially useful when variates are positive and negative) of differences between measured and simulated Model cell location Head heads were less than 1.5 ft, and (3) the model-computed Model Model Well number difference in non-pumped well discharge and the discharge measured in streams and drains row column were within 10 percent of each other. 38 67 OW 436 0.05 As the previous ground-water model of this area 45 62 OW 490 1.45 (Zarriello, 1993) had demonstrated, the relatively thin 45 58 OW 502 1.11 saturated thickness of the aquifer (less than 5 ft to 35 ft) limited the range of horizontal hydraulic conductivities and 66 68 OW 348 -1.39 rate of recharge values that could be used before modeled 71 67 OW 484 2.34 hydraulic-head elevations in cells would decrease below 72 57 OW 427 1.64 model layers. This modeling limitation results in dry model cells and model instability. The distributions of horizontal 75 67 OW 347 2.98 hydraulic conductivities for each different type of glacial 78 55 OW 425 2.10 deposit (fig. 7) are shown in figure 13 and were estimated 78 56 OW 424 1.87 from surficial geologic interpretations by Miller and others 79 56 OW 486 2.50 (1989) and estimates produced by Zarriello’s (1993) model of the aquifer system. 79 56 OW 485 1.03 The simulated water budget for the aquifer is summarized RMS 1.84 in table 6. Most recharge to the aquifer (61 percent) occurs through precipitation that falls directly over the aquifer. Losing Table 5b. Difference between measured and simulated streams also are a major source of recharge that contribute streamflow in the Sandy Creek/Lacona, N.Y. model study area. 38 percent of the total recharge to the aquifer. The remaining 1 percent of water entering the aquifer is from unchanneled [ft3/s, cubic feet per second] runoff and ground-water inflow from till uplands to the east. Most discharge (86 percent) from the aquifer occurs as Streamflow measurement Measured flow Simulated flow seepage to the numerous headwater streams in the western 3 3 location (ft /s) (ft /s) part of the study area. The production wells and ground water Lindsey Creek—northern -0.331 -0.341 that discharges into the manmade ditches along County Route part of the study area 22 each account for another 5 percent of the discharge. The Sand and gravel quarry 0.46 0.41 least amount of discharge (4 percent) occurs via discharge into ditch along County Little Sandy Creek. Route 22 The configuration of the contours representing the water 1Negative value indicates loss of streamflow and a positive value indicates a table is shown in figure 14. In general, the water table slopes gain in streamflow. from east to west. The head drop across the aquifer is 250 ft with the highest heads (a little above 610 ft) being in the eastern part of the aquifer and the lowest heads (slightly below investigation concentrated on the hydrology of the southern 360 ft) in the northwestern part. production wells, and where there was a discrepancy between ground-water traveltimes determined by chemical data (Komor, 2002) and those determined by previous numerical Model Sensitivity simulation of ground-water flow (Zarriello, 1993). The Horizontal and vertical hydraulic conductivity values northern area was included in this model because during the and recharge rates in the model were varied within reasonable period of this investigation, the Villages of Sandy Creek and ranges to test the sensitivity of these model parameters. Lacona had installed another production well in that area, and Whereas recharge rates and hydraulic conductivity values were this well is used as an alternate supply for the villages. varied, a constant pumping rate of 98 gal/min was maintained Starting values for heads used for the numerical at production wells OW 487 and OW 490 in the south and ground-water model were derived from (1) heads computed north well fields, respectively, for a total pumping rate for from the model by Zarriello (1993), (2) heads from several both wells of 196 gal/min, or 0.44 ft3/s (0.29 Mgal/d) (table 6). observation and production wells in the study area, and Results of the sensitivity analyses are depicted in figure Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 31
76°05'45" 76°03'15"
43°41'11"
Lindsey Creek
1 2 3 4 5 6 7 8 9 10 D D’
Inactive model area
Little Sandy Creek 0 0.25 0.5 1 MILE 43°38'41" 00.25 0.5 1 KILOMETER
Base from U.S. Geological Survey1:24,000 Digital Line Graph, Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION 10 HYDRAULIC CONDUCTIVITY ZONE–Shows area of similar hydraulic conductivity used in calibration of ground-water-flow model. (see table below for values) D D' West East Hydraulic Conductivity Zone 1 2 3 4 5 6 7 8 9 10 (Values are in feet per day) LAYER 1 9 12 25 50 60 80 1,000 120 120 110 LAYER 2 9 12 25 50 60 80 1,000 120 120 110 LAYER 3 9 12 25 50 60 70 120 120 120 110 LAYER 4 8 10 20 40 50 70 110 120 120 110 LAYER 5 8 10 20 40 50 70 110 120 120 110
Figure 13. Hydraulic conductivities for different aquifer deposits and model layers used in ground- water-flow model calibration. 32 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Table 6. Steady-state ground-water budget for the ground-water-flow model of the Tug Hill glacial-drift for average annual recharge conditions.
[ft3/s, cubic feet per second; Mgal/d, millions of gallons per day]
Amount Percent of Ground-water budget component ft3/s Mgal/d total budget Recharge to the aquifer Precipitation on the aquifer 5.91 3.83 61 Upland sources Seepage losses from streams 3.72 2.41 38 Unchanneled runoff and ground-water inflow from uplands 0.05 0.03 1 Total 9.68 6.27 100
Discharge from aquifer Discharge from aquifer to drains (headwater streams) 8.37 5.42 86 Pumping wells 0.44 0.29 5 Discharge from aquifer to manmade ditches 0.52 0.33 5 Discharge from aquifer to streams 0.35 0.23 4 Total 9.68 6.27 100
15. The model was very sensitive to changes in horizontal 500 dry cells. Horizontal hydraulic conductivity could be hydraulic conductivity and recharge rate values, especially in increased by no more than 25 percent of the calibrated values the parts of the aquifer that consist of lacustrine and outwash/ before substantial amounts of dry cells occurred, and the inwash deposits, and was less sensitive in the permeable model became unstable. The model was relatively insensitive beach deposits in the central part of the aquifer. Hydraulic- to changes in horizontal hydraulic conductivity of the beach head changes were least sensitive to variations in the vertical deposits. Hydraulic conductivity was calculated for the beach hydraulic conductivity. deposits during an aquifer test at the southern well field (Zarriello, 1993) and is, therefore, fairly well defined (see Horizontal Hydraulic Conductivity “Hydrology” section, earlier in this report).
Decreasing hydraulic conductivity values by 50 percent Vertical Hydraulic Conductivity caused simulated heads to rise several feet above the final calibrated heads in areas of the aquifer underlain by sand Increasing the vertical hydraulic conductivity from and gravel that are distant from the production wells and in the calibrated value of 0.1 ft/d to 0.2 ft/d, then to 0.5 ft/d, the northwestern part of the study area that is composed of resulted in little change to the simulated heads in the model. lacustrine sand deposits. However, the simulated heads near Decreasing the vertical hydraulic conductivity from the production wells were lower than measured heads, indicating calibrated value of 0.1 ft/d to 0.08 ft/d, then to 0.06 ft/d, also that horizontal hydraulic conductivity values of 50 percent less resulted in little change to the simulated hydraulic heads in the than those used in the final calibrated model were probably model. The scant change in the simulated hydraulic heads with unrealistically low. Decreasing values of horizontal hydraulic positive or negative changes in vertical hydraulic conductivity conductivity reduces the rate at which water is transmitted indicates that this parameter is the least sensitive of the parameters tested for the model. through the aquifer; therefore, the effect on heads under steady pumping rates is to increase drawdown and thus, increase the size of the area affected by the production wells. Recharge Increasing the horizontal hydraulic conductivity by Increasing the final calibrated model recharge rate 50 percent caused many heads in the outwash/inwash area in (28 in/yr) by 1.25 times (35 in/yr) caused hydraulic heads in the eastern part of the study area to drop below the bottom the lower areas of the aquifer in the northwestern part of the of the cells in all layers of the model resulting in more than study area to rise above land surface. Hydraulic heads in the Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 33
76°05'45" 76°03'15"
43°41'11" 360 Lindsey Creek
400
610 500 OW 436
(560) 450
OW 490 OW 502 (550)
(548) 400
550
OW 348 (581)
600
OW 427 OW 484 (547) (583) OW 425 (544) OW 424 (544) OW 347 OW 485 (578) (543)
OW 486 600 OW 487 (544)
Little Sandy Creek 0 0.25 0.5 1 MILE 43°38'41" 00.25 0.5 1 KILOMETER
Base from U.S. Geological Survey 1:24,000 Digital Line Graph, Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION 550 WATER-TABLE CONTOUR–Shows altitude of water table derived from calibrated three-dimensional ground-water-flow model. Contour interval 10 feet. Datum is NGVD 1929 PRODUCTION WELL OW 436 WELL USED FOR MODEL CALIBRATION–Upper number is assigned (560) U.S. Geological Survey well number, lower number in parentheses is measured hydraulic head. Datum is NGVD 1929
Figure 14. Water-table contours based on the calibrated, three-dimensional-flow-model output in the Sandy Creek/ Lacona, N.Y. study area. 34 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
4.5
Horizontal hydraulic conductivity 4.0 Recharge Vertical hydraulic conductivity
3.5
3.0
2.5
2.0 ROOT-MEAN SQUARE OF DIFFERENCE BETWEEN ROOT-MEAN CALCULATED AND MEASURED HYDRAULIC HEAD CALCULATED
1.5 0.4 0.5 0.6 0.7 0.8 0.9 1 1.5 2 MULTIPLICATION FACTOR
Figure 15. Results of sensitivity analyses of hydraulic head in the Tug Hill glacial-drift aquifer in the Sandy Creek/ Lacona, N.Y. study area.
model areas with larger horizontal hydraulic conductivity (135,000 gal/d) equal to a pumping rate of about 100 gal/min (beach and outwash/inwash areas) were relatively insensitive for 24 hours of pumping. to increases in recharge rate. Heads in the beach and outwash/ The contributing areas to the main production well inwash areas did not start to rise above the modeled land (OW 487) in the southern well field and to the new production surface until the initial recharge rate (28 in/yr) was doubled well (OW 490) in the northern part of the study area were estimated using calibrated model output from MODFLOW (56 in/yr). Decreasing the calibrated recharge rate (28 in/yr) by and post-processing with the particle-tracking routine 10 percent was the lowest rate that could be sustained before MODPATH (Pollock, 1989). The traveltimes within these the model cells began to go dry. Further decreases in recharge contributing areas are based on an assumed porosity of 0.3, resulted in dewatering of significant areas in the eastern part of which is reasonable for these types of deposits (Freeze and the modeled area. Cherry, 1979). The areas that provide recharge to the two production wells were determined by particle tracking in Simulation of Ground-Water Withdrawals and the forward direction toward the pumped wells (areas of discharge). Particles were placed on the top face of layer Determination of Contributing Areas 1 cells representing the water table, and the path lines that terminated at the pumped wells were used to identify the area In 2005, withdrawal from the Villages of Sandy contributing water to the pumped wells. For time-of-travel Creek/Lacona production wells was about 270,000 gal/d analyses, 20 particles were started in the center of the cell (0.27 Mgal/d). Withdrawals were alternated from the north containing the pumped well and run in the reverse direction and south well fields on a daily basis to reduce the stress at toward the origin of recharge. Particles tracked in the reverse each area and allow ground-water heads to recover during the direction were used to simulate time-of-travel zones of 1 to non-pumping days. For steady-state model simulation of the 60 days (about 3 months) or less, 61 to 180 days (0.5 years) pumping, half of the total pumpage was applied to each well or less, 181 to 365 days (1 year) or less, 366 to 600 days Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 35
76°05'45 76°03'15"
43°41'11"
Lindsey Creek
OW 490
OW 487
Little Sandy Creek 0 0.25 0.5 1 MILE 43°38'41" 00.25 0.5 1 KILOMETER
Base from U.S. Geological Survey 1:24,000 Digital Line Graph, Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION SIMULATED AREA CONTRIBUTING TO PUMPED WELL–Area derived from particle-tracking results based on calibrated three-dimensional-flow-model output for wells pumping at 100 gallons per minute. Number of days represents range of time for water particle to travel to the well from within that colored area. OW 487 Production well 1-60 days 61-180 days 181-365 days 366-600 days 601-730 days 731-910 days (north well only) (north well only)
Figure 16a. Simulated areas contributing to production wells pumping at 100 gallons per minute, derived from particle-tracking results based on calibrated, three-dimensional-flow-model output in the Sandy Creek/Lacona, N.Y. study area. 36 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
76°05'45" 76°03'15"
43°41'11"
Lindsey Creek
OW 487
Little Sandy Creek 0 0.25 0.5 1 MILE
43°38'41" 00.25 0.5 1 KILOMETER
Base from U.S. Geological Survey 1:24,000 Digital Line Graph, Universal Transverse Mercator projection, Zone 18N, NAD 1927 EXPLANATION SIMULATED AREA CONTRIBUTING TO PUMPED WELL–Area derived from particle-tracking results based on calibrated three-dimensional-flow-model output for wells pumping at 200 gallons per minute. Number of days represents range of time for water particle to travel to the well from within that colored area.
OW 487 1-60 days 61-180 days 181-365 days 366-545 days Production well
Figure 16b. Simulated areas contributing to the southernmost production wells pumping at 200 gallons per minute, derived from particle-tracking results based on calibrated, three-dimensional-flow-model output in the Sandy Creek/ Lacona, N.Y. study area. Hydrogeology of the Sandy Creek/Lacona Water-Supply Area 37
(south production well only), 601 to 730 days (2 years) or age (13.6 years) of water from production well OW 487, less, and 731 to 910 days (2.5 years; north production well which has a 20-ft screen and draws water from most of the only) (fig. 16a). In general, simulated traveltimes were saturated thickness of the aquifer (fig. 10). The dates of the relatively short. Simulation indicates it takes approximately ground-water samples from discrete sections of the sand and 500 days for most of the water to flow from the eastern gravel aquifer indicate that there was an insufficient amount boundary of the aquifer to the production wells. In 2006, of young water that could potentially mix with old water in the northern production well was discontinued because of the sand and gravel aquifer to result in the CFC-derived age elevated manganese concentrations, forcing the Villages of of 500 days or less indicated by ground-water traveltimes Sandy Creek and Lacona to increase pumping in the southern derived from numerical model results. Water-quality data from well field to a rate of 200 gal/min. The configuration of the well OW 485 (finished in bedrock) indicate that the water simulated contributing area under the increased pumping had different chemical characteristics than water in the Tug conditions (200 gal/min) at the southernmost well field is Hill glacial-drift aquifer, and had an apparent age (32 years) shown in figure 16b. about twice that of water from the sand and gravel aquifer. Therefore, the different water chemistry, the CFC age, and the Comparison of Geochemical Age Dating and hydraulic-head relations between the bedrock aquifer and the overlying Tug Hill glacial-drift aquifer indicate that little, if Ground-Water Traveltimes Computed from any, older water moved upward from the bedrock unit into the Numerical Modeling younger water in the overlying sand and gravel aquifer. Typically, one would expect to encounter progressively Results from the numerical ground-water-flow model younger water east of the well field, upgradient from the by Zarriello (1993) and from the model constructed for this production well, and closer to the sources of recharge, such study indicate that for the longer flow paths originating at the as losing streams, runoff, and ground-water inflow from eastern margin of the aquifer, it would take approximately the Tug Hill Plateau. However, a sample collected from test 500 days for ground water to reach the southern well field well OW-484 (figs. 7 and 8) in the central part of the aquifer and progressively less time as distance to the well field had a CFC-derived apparent age of 17 years, indicating that decreases. However, results for a water sample collected water in this area to the east of the production well was not from the southern well field by Komor (2002) and analyzed younger than water from the well area as would be expected. for CFCs and tritium indicate that the apparent age of the The 17-year age from CFC measurements indicates that the water was about 17 years. Several methods were used in an apparent age of some of the recharge water to this test well attempt to resolve the discrepancy between ground-water was greater than would be expected. Relatively old (32 years) traveltimes determined by numerical ground-water-flow water from the underlying bedrock was ruled out as a source modeling and those determined by geochemical dating, of old water in the production well because of differences in including (1) analyzing additional water samples from wells chemistry, CFC-derived ages, and hydraulic-head relations, in the southern well field finished at various depths, including therefore, there likely is another source of water that recharged one well finished in bedrock, for common ions, nutrients, the aquifer in this area and increased the CFC-derived age. and CFCs; (2) analyzing for CFCs in samples collected from an additional test well finished upgradient (east) from the Sandy Creek/Lacona well field and approximately halfway Flow-Model Traveltimes and Chlorofluorocarbon between the well field and the eastern boundary of the Age Comparisons aquifer; and (3) constructing a three-dimensional, ground- water-flow model. In order to explain the source of the older water determined by analyses for CFCs in the southern well field, a hypothesis was proposed that states the longer traveltimes may Apparent Ground-Water Ages be the result of (1) slower vertical components of ground- At a production well in the southern well field, results water flow, and (2) slightly longer flow paths than those in of analyses for CFCs in a sample collected August 24, Zarriello’s model (1993). To test this hypothesis, traveltimes 1999, by Komor (2002) indicated that the apparent age of determined from particle-tracking analyses were compared to the water was 17 years, slightly older than, but somewhat ground-water ages determined from CFCs. similar to, the apparent age of water (13.6 years) collected on The time for ground water to travel across the southern November 27, 2001, from production well OW 487 (table 4). part of aquifer from east to west computed from the The apparent ages of ground water obtained from wells with calibrated model results is 500 days. To simulate a scenario short-screened intervals (1.5-ft-long screens) finished in of slower ground-water-flow paths, the following model discrete parts of the aquifer—OW 488 finished in the upper parameters were adjusted: (1) vertical anisotropy was set to part of the Tug Hill glacial-drift aquifer (11.7 years) and 1:10, (2) hydraulic conductivity values were set to one-half OW 486 finished in the lower part of the sand and gravel of the calibrated values, and (3) porosity was increased by aquifer (15.0 years)—when averaged, are close to the apparent 33 percent—raised from 0.3 to 0.4. 38 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Results from these changes indicate that the longest time CFCs are not thoroughly explained by any of the following for ground water to travel across the aquifer from east to west hypotheses: (1) mixing of young and old water within the would be about 4 years, still substantially younger than the Tug Hill glacial-drift aquifer, (2) mixing of water in the Tug CFC-derived apparent age of 11.7 to 15 years. Even when Hill glacial-drift aquifer with water from the underlying model parameters were adjusted to their feasible limits to bedrock aquifer, or (3) adjustment of parameters in the produce the slowest possible flow paths, the numerical model numerical ground-water-flow model within reasonable limits could not simulate times-of-travel of ground water that were close to that of the CFC-derived apparent ages. (to effectively slow ground-water-flow velocities). Therefore, The discrepancy between ground-water ages in the another hypothesis was needed to account for water that Sandy Creek/Lacona study area determined from the ground- is relatively old entering the Tug Hill glacial-drift aquifer water-flow modeling results and those computed from as recharge.
Tug Hill glacial-drift aquifer Tug Hill Plateau (sand and gravel) Upland stream Drumlin Till-mantled (till) bedrock hillside Precipitation that falls directly over the sand and gravel aquifer
Beach ridge
Losing stream Pumped well Springs Water in aquifer is mixture of old and young water
Not to scale EXPLANATION
BEACH SAND AND GRAVEL–Ridge of well-sorted coarse SOURCES OF OLD RECHARGE TO SAND AND GRAVEL AQUIFER–Water is sand to cobble gravel deposited at the shoreline of older than 1 year proglacial Lake Iroquois Ground-water inflow from relatively impermeable till that seeps into OUTWASH AND/OR INWASH–Fans of dirty sand and gravel upland streams and along the margin of the sand and gravel aquifer deposited by meltwater and by recent streams that drained Tug Hill Plateau Ground-water inflow through relatively impermeable fractured shale that discharges into upland streams and along eastern margin of the LODGMENT TILL–Gravel and cobbles embedded in a matrix aquifer of clay, silt, and fine sand that were deposited at the base of the glacier, poorly sorted, compact, and relatively impermeable Seepage from losing upland stream where it flows over the aquifer BEDROCK–Ordovician age siltstone and shale SOURCES OF YOUNG RECHARGE TO SAND AND GRAVEL AQUIFER–Water is younger than 1 year FRACTURES OR BEDDING PLANES IN BEDROCK Precipitation that falls directly over the aquifer WATER TABLE REGIONAL GROUND-WATER FLOW IN BEDROCK Unchanneled runoff from adjacent hillsides
Figure 17. Conceptual diagram showing possible sources of older ground water recharging the sand and gravel aquifer near Lacona, N.Y. Summary 39
Two alternative hypotheses that account for the relatively Also, a resolution of the discrepancies between apparent old water are (1) relatively old ground water that moves ages of water in the Tug Hill glacial-drift aquifer previously slowly through the till and bedrock in the Tug Hill uplands determined using numerical ground-water-flow models, and discharges to upland streams during base-flow conditions and those determined by use of CFCs was needed because that then lose water to the sand and gravel deposits where the aquifer is the source for the Sandy Creek/Lacona and they flow over the Tug Hill glacial-drift aquifer (38 percent of Pulaski production wells. Uncertainty about the ground-water estimated total recharge; fig. 17 and table 6), and, to a lesser traveltimes and contributing areas to these production wells extent, (2) ground-water inflow into the Tug Hill glacial-drift necessitated re-evaluations of previous ground-water-flow aquifer from adjacent till and bedrock uplands (1 percent of modeling results and CFC-derived ages of water. estimated total recharge; fig. 17 and table 6). During base The hydrogeology of two well sites, one for the Village flow, the water that discharges into upland streams is from till of Pulaski and the other for the Villages of Sandy Creek and and bedrock of relatively low horizontal and vertical hydraulic Lacona in Oswego County, N.Y., were investigated separately conductivity. This water moves very slowly through these in two projects during 2003–04 and 2001–03, respectively. poorly permeable units and, hence, is relatively old—possibly The production wells for both sites are completed in the several decades old. In small streams, once the ground water Tug Hill glacial-drift aquifer, which is composed mainly of discharges from the till and bedrock, it likely is in the channel Pleistocene, glaciofluvial, and glaciolacustrine sands and for only several hours to less than a day, and may not be in gravels. The water table in each area regionally slopes from contact with the atmosphere long enough for CFCs to degrade east to west. or for concentrations of CFCs to be reset from exposure to the At the Pulaski area, 11 new shallow test wells were atmosphere (L. Neil Plummer, USGS, oral commun., 2005). installed and monitored for water levels and temperature. A In Cortland County, N.Y., to determine the apparent age nearby stream also was monitored to assess stream-aquifer of water in a small upland stream, a sample was collected interaction. For this investigation, a three-dimensional, where the stream begins to exit a till-mantled bedrock upland five-layered, ground-water-flow model was constructed for the drainage area. The stream then flows out over a sand and Sandy Creek/Lacona area based on new test wells screened gravel valley-fill aquifer, where it begins to recharge the at discrete zones in the aquifer; and CFC-derived ages of aquifer. CFC concentrations indicated that the apparent water from these test wells were determined and compared age of the surface water was 18 years (USGS CFC Lab, to model output. Samples analyzed for inorganic constituents written commun., 2005). This relatively old water seeps and physical properties were collected from test wells and into the aquifer and mixes with younger ground water that production wells in both areas to help characterize the originates from recent precipitation that falls on the aquifer. geochemistry of the aquifer and determine sources of recharge Consequently, in this type of hydrogeologic setting, where to the aquifer and production wells. some recharge is derived from losing streams that drain The Village of Pulaski water supply is drawn from four till-mantled bedrock uplands, the water that is sampled in shallow, concrete-lined production wells that range in depth the aquifer is a mixture of water of different ages. Using from 8 to 12 ft and discharge into a centralized treatment the concentration of CFCs to determine the ages of mixed system before being gravity fed to the village. Four 2-in. test water results in an average apparent age. Additional data are wells were installed in 2003, upgradient from each production needed for the Sandy Creek/Lacona study area to determine well, and heads were measured and samples collected to more definitively whether older water from losing streams determine physical properties and concentrations of inorganic that drain the Tug Hill Plateau till and bedrock uplands constituents, nutrients, bacteria, SF6, dissolved gases, and (and from adjacent till and bedrock that border the aquifer) CFCs during winter 2003 and spring 2004. Concentrations mixes with younger water in the sand and gravel aquifer in of inorganic constituents and values of physical properties Oswego County, which is similar to the one near Cortland, indicated that the water was potable and a calcium bicarbonate N.Y., in order to explain the ground-water-age discrepancy type. In three of the four wells sampled, the only inorganic between ages determined from the modeling results and those constituent that slightly exceeded the USEPA Secondary determined using the CFC data. Maximum Contaminant Level (SMCL) was iron. The presence of relatively small concentrations of the other major ions and nutrients in the samples indicated that contamination Summary from human practices such as septic-tank effluent, road-salt applications, or agricultural applications within the area were Concerns about contamination of aquifers by chemical not affecting the quality of the aquifer water near the well field spills and agricultural practices highlighted the need for an at the time the aquifer was sampled. Additionally, coliform understanding of the ground-water-flow system in Oswego bacteria was detected in three samples at counts ranging from County, N.Y., especially in parts of the county where there is 3 to 24 colonies per 100 mL of water, whereas E. coli were not unconfined sand and gravel in the Tug Hill glacial-drift aquifer detected in any samples and, therefore, did not pose a health that is vulnerable to contamination from surface activities. risk at the time of sampling. 40 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
The hydraulic gradient in areas away from the production flows toward the production wells, thereby short-circuiting the wells is about 0.01 ft/ft, but the gradient likely increases closer ground-water system. to the wells and streams. Temperature probes and water- Four test wells (three wells in the sand and gravel aquifer level-data loggers were installed in (1) one production well, and one in bedrock) were installed in the Sandy Creek/Lacona (2) in a stream 20 ft south of the well, and (3) in a piezometer study area to aid in construction of the numerical ground- between the stream and the well to determine whether there is water-flow model, determine vertical distribution of heads, interaction between the surface water and ground water. The calibrate the model, and provide access to collect water-quality temperature and water-level data indicated that there was no samples. Results of the chemical analyses of ground-water infiltration of water from the stream to the well. samples indicate that at the Sandy Creek/Lacona well field the Results of a numerical ground-water-flow model ground water is potable and a calcium-bicarbonate type that constructed by Malcolm Pirnie, Inc., indicated that the time- meets all State and Federal water-quality guidelines. A sample of-travel for water following the longest ground-water-flow from the test well that is finished in bedrock, however, was a path, roughly 7,000 ft from the west edge of the aquifer sodium-bicarbonate/sulfate type, and the sodium concentration to the Village of Pulaski well field, was less than 1 year to slightly exceeded the SMCL. arrive at the well field. In contrast, geochemical-age dating CFCs were measured to determine the apparent ages using concentrations of CFCs and tritium to estimate the age of four ground-water samples collected during fall 2001 and of ground-water recharge indicated that the apparent age of summer 2002. At the southern well field, the apparent ages of ground water discharging to the Pulaski production wells ground water ranged from 11.7 years for water collected from was 24 years. To resolve the difference in apparent ground- a well 20 ft deep and finished in the upper part of the aquifer water ages derived from the numerical ground-water-flow to 15.0 years from a well 29 ft deep and finished in the lower model and the CFC concentrations, a second set of water part of the sand and gravel aquifer. In production well OW samples was collected from four test wells in November 2003 487 (34 ft deep with a 20-ft screen that almost spans the entire and May 2004 and analyzed for CFCs and tritium. The data saturated part of the aquifer), the apparent age of water was indicted that the apparent age of the water sampled ranged 13.6 years, roughly the average age of the sampled shallow from approximately 25 years to 1 year or less, indicating a and deep piezometers. In a test well (23 ft deep and screened mixing of waters of different ages. Older water (CFC-derived at the bottom of the sand and gravel in the central part of the ages indicate water from the mid-1970s) might indicate the aquifer) the apparent age was 17.4 years. This older age for presence of deep flow paths within the aquifer that supply water from the central part of the aquifer was not expected water to the wells, and younger water (CFC-derived ages because the well is upgradient from the well-field area. estimated to be 1 year or less) might indicate relatively A three-dimensional, finite-difference ground-water-flow shallow flow paths and quick traveltimes within the aquifer. model was used to simulate ground-water flow in the Sandy The Villages of Sandy Creek/Lacona obtain their water Creek/Lacona study area in order to determine contributing from wells in the northeastern part of Village of Lacona and areas under two pumping scenarios and to interpret the from a well 1.25 mi north of the villages near the northern discrepancy between ground-water ages determined by border of Oswego County. Both wells are in Pleistocene previous ground-water-flow models and CFC-derived ages. glaciolacustrine beach, glaciofluvial outwash, and alluvial Contributing areas to the well fields were calculated from the inwash deposits. The beach deposits consist of well-sorted calibrated model output using a particle-tracking program. sand and gravel, whereas the outwash and alluvial inwash Calibrated model output indicated that (1) 61 percent of deposits consist of silty sand and gravel, with lenses of silty recharge to the aquifer occurs through precipitation that falls fine-to-medium sand. The saturated aquifer thickness ranges directly over the aquifer, (2) 38 percent of the simulated from 10 to 25 ft and generally is thickest in the western part recharge is from losing streams flowing over the aquifer, of the aquifer. The aquifer is recharged by precipitation that and (3) the remaining 1 percent of water entering the aquifer falls directly over the aquifer and by upland sources that is from unchanneled runoff and ground-water inflow from include (1) runoff and ground-water inflow from unchanneled till uplands to the east. Most discharge (86 percent) from hillsides of the Tug Hill Plateau to the east of the aquifer, the aquifer occurs as seepage to the numerous headwater and (2) upland tributaries that typically lose water where streams in the western part of the study area. The municipal they flow over the eastern part of the aquifer. Average annual pumping wells and the discharge into the manmade ditches recharge in the area from direct precipitation over the aquifer along County Route 22 each account for 5 percent of the is estimated to be 27 in/yr. Wells in the aquifer are susceptible discharge (a total of 10 percent) from the modeled area. The to contamination because (1) the aquifer is unconfined, (2) it least amount of discharge (4 percent) occurs via discharge into is highly transmissive, and (3) the time-of-travel of ground Little Sandy Creek. water from recharge areas to the production wells is relatively Particle-tracking analyses indicate simulated recharge to short. Also, a drainage ditch that fully penetrates the central the production wells is from parabolic areas extending from part of the aquifer drains the entire thickness of the aquifer in the wells towards the eastern boundary of the aquifer. In the this area and routes this water westward to an area near the southern part of the modeled area, the longest flow paths to the production wells, where some water recharges the aquifer and production well (starting from the eastern aquifer boundary) Selected References 41 range from 3,000 to 4,000 ft and take from about 1.5 to 2 years Busenberg, Eurybiades, and Plummer, L.N., 2000, Dating for the farthest particles to reach the production wells from the young groundwater with sulfur hexafluoride: Natural eastern boundary of the aquifer. In contrast, from a previous and anthropogenic sources of sulfur hexafluoride: Water study, the apparent age of pumped water at the well field Resources Research, v. 36, no. 10, p. 3,011–3,030. determined from CFC concentrations, was about 17 years. As Cadwell, D.H., and Muller, E.H., 1986, Surficial geologic described above, this study yielded apparent ages ranging from map of New York: New York State Museum—Geological 11.7 years to 15 years. Survey, Map and Chart Series 40, Finger Lakes Sheet, New Two hypotheses, which are (1) mixing of water in the York State Geological Survey, 1 sheet, scale 1:250,000, sand and gravel in the Tug Hill glacial-drift aquifer with older digital compilation by Beckie Ugolini, 1998, from New water from the bedrock aquifer, and (2) ground water is older York State Geological Survey 90-meter DEMS. because it moves more slowly through the aquifer (this was simulated by adjusting model parameters within reasonable Clark, I.D., 2006, Course notes for GEO 4342, Groundwater limits to effectively retard ground-water velocities), could Geochemistry, accessed June 18, 2006, at http://www. not explain the discrepancy between the ages simulated by science.uottawa.ca/est/eng/Prof/clark/GEO4342/Ch%20 ground-water-flow modeling and those measured by CFC 8%20Groundwater%20renewability.pdf age-dating methods. An alternative hypothesis is that part of Elkins, J.W., 1999, Chlorofluorocarbons, in Alexander, the water that enters as recharge to the sand and gravel in the D.E. and Fairbridge, R.W., eds., The Chapman & Hall Tug Hill glacial-drift aquifer is initially relatively old water. encyclopedia of environmental science: Boston, Ma., Possible sources of old recharge to the Tug Hill glacial- Kluwer Academic, p. 78–80. drift aquifer might include streams, which during base-flow Freeze, A.R. and Cherry, J.A., 1979, Groundwater: Englewood conditions consist of old ground water discharged from Cliffs, N.J., Prentice Hall, Inc., 604 p. upland till and bedrock units draining the Tug Hill uplands. In the model simulation, about 38 percent of recharge to the Heaton, T.H.E., 1981, Dissolved gases—Some applications Tug Hill glacial-drift aquifer is from these upland streams to groundwater research: Transactions of the Geological flowing across the aquifer. This process of recharge from old Society of South Africa, v. 84, p. 91–97. water derived from losing streams that drain upland areas was Heaton, T.H.E., and Vogel, J.C., 1981, “Excess air” in similarly documented in an aquifer study near Cortland, N.Y., groundwater: Journal of Hydrology, v. 50, p. 201–216. which had a similar hydrologic setting, and where water in the upland streams was dated at 18 years. Additionally, lateral Komor, S.C., 2002, Ground-water age dating in community ground-water inflow from adjacent till and bedrock uplands wells in Oswego County, New York: U.S. Geological may account for some of the older recharge to the sand and Survey Open-File Report 01–232, 16 p. gravel aquifer. Kontis, A.L., Randall, A.D., and Mazzaferro, D.L., 2004, Regional hydrology and simulation of flow of stratified-drift aquifers in the glaciated northeastern United States: U.S. Acknowledgments Geological Survey Professional Paper 1415–c, plate 1. Malcolm Pirnie, Inc., 2000, Wellhead protection study— The authors thank Auralie-Ashley Marx, the Oswego Village of Pulaski, New York: Latham, N.Y., Malcolm County Environmental Program Director, and Chris Williams Pirnie, Inc., 24 p. of the Oswego County Health Department for their efforts in assisting with collection and analysis of water samples. McDonald, M.G., and Harbaugh, A.L., 1988, A modular Thanks also are given to Gary Stevens, Water Manager of the three-dimensional finite-difference ground-water flow Village of Pulaski Water Department, for providing access model: U.S. Geological Survey Techniques of Water- to the wells in the village’s well field and for providing Resources Investigations, book 6, chap. A1, 586 p. valuable information. Miller, T.S., 1980a, Surficial geology of part of Sandy Creek quadrangle, Oswego County, New York: U.S. Geological Selected References Survey Open-File Report 80–1114, 1 sheet. Miller, T.S., 1980b, Surficial geology of the Richland quadrangle, Oswego County, New York: U.S. Geological Busenburg, Eurybiades, and Plummer, L.N., 1992, Use Survey Open-File Report 80–763, 1 sheet. of chlorofluorocarbons as hydrologic tracers and age- dating tools—the alluvium and terrace system of Central Miller, T.S., 1982, Geology and ground-water resources of Oklahoma: Water Resources Research, v. 28, no. 9, Oswego County, New York: U.S. Geological Survey Water- p. 2,257–2,283. Resources Investigations Report 81–60, 37 p. 42 Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York
Miller, T.S., Sherwood, D.A., and Krebs, M.M., 1989, U.S. Environmental Protection Agency, 1997, Guidelines Hydrogeology and water quality of the Tug Hill glacial for preparation of the comprehensive state water quality aquifer in northern New York: U.S. Geological Survey assessments (305(b) reports) and electronic updates: Water-Resources Investigations Report 88–4014, 60 p., Washington D.C., U.S. Environmental Protection 24 pls., scale 1:24,000. Agency, Office of Water, EPA 841-B-97-002A and EPA 841-B-97-002B, PL 95–217. Office of the Federal Register, National Archives and Records Administration, 1990, 40 CFR 52. U.S. Environmental Protection Agency, 2000, Water quality conditions in the United States—Profile From the 1998 Plummer, L.N., and Busenberg, Eurybiades, 1999, National Water Quality Inventory Report to Congress: Chlorofluorocarbons: Tools for dating and tracing Washington D.C., U.S. Environmental Protection Agency young groundwater, in Cook, P., and Herczeg, A., eds., Report EPA–841–F–00–006. Environmental tracers in subsurface hydrology: Boston, Ma., Kluwer Academic Publishers, chap. 15, p. 441–478 . U.S. Environmental Protection Agency, 2002, Drinking water advisory—Consumer acceptability advice and health effects Plummer, L.N., Michel, R.L., Thurman, E.M., and Glynn, analysis on sodium: Washington D.C., U.S. Environmental P.D., 1993, Environmental tracers for age dating young Protection Agency, Office of Water, EPA 822–R–02–032, ground water, in Alley, W.M., ed., Regional ground-water 34 p. quality: New York, N.Y.,Van Nostrand Reinhold, chap. 11, p. 255–294. U.S. Environmental Protection Agency, 2004, Proposed radon in drinking water rule: Washington D.C., U.S. Pollock, D.W., 1989, Documentation of computer programs Environmental Protection Agency, Office of Water, accessed to compute and display pathlines using results from the May 14, 2004, at http://www.epa.gov/safewater/radon/ U.S. Geological Survey modular three-dimensional finite- proposal.html. difference ground-water flow model: U.S. Geological Survey Open-File Report 89-622, 49 p. U.S. Environmental Protection Agency, 2004, 2004 Edition of drinking water standards and health advisories: Washington Reilly, T.E., Plummer, L.N., Phillips, P.J., and Busenberg, D.C., U.S. Environmental Protection Agency, Office of Eurybiades, 1994, The use of simulation and multiple Water, EPA822-R–04–005, Winter 2004. environmental tracers to quantify groundwater flow in a shallow aquifer: Water Resources Research, v. 30, no. 2, U.S. Geological Survey, 2003, CFC webpage, accessed p. 421–433. October 29, 2003, at http://water.usgs.gov/lab/cfc Ridge, J.C., 2003, The last deglaciation of the northeastern Zarriello, P.J., 1993, Determination of the contributing area United States: a combined varve, paleomagnetic and to six municipal ground-water supplies in the Tug Hill calibrated 14C chronology, in Cremeens, D.L. and Hart, glacial aquifer of northern New York, with emphasis on the J.P., eds., Geoarcheology of landscapes in the glacial Lacona-Sandy Creek well field: U.S. Geological Survey northeast: New York State Museum Bulletin, v. 497, Water-Resources Investigations Report 90–4145, 51 p. p. 15–45. Zaugg, S.D., Sandstrom, M.W., Smith, S.G., and Fehlberg, Rozanski, K., Gonfiantini, R., and Araguas-Araguas, L., 1991, K.M., 1995, Methods of analysis by the U.S. Geological Tritium in the global atmosphere—Distribution patterns Survey National Water Quality Laboratory—Determination and recent trends: Journal of Physics, G: Nuclear Particle of pesticides in water by C-18 solid-phase extraction and Physics, v. 17, p. S523–S536. capillary-column gas chromatography with selective-ion monitoring: U.S. Geological Survey Open-File Report Stute, M. and Schlosser, P., 1999, Atmospheric noble gases, 95–181, 49 p. in Environmental tracers in subsurface hydrology, Cook, P., and Herczeg, A., eds.,: Boston, Ma., Kluwer Academic Publishers, Chapter 11, p. 349–377. U.S. Environmental Protection Agency, 1996, Drinking water regulations and health advisories: Washington D.C., U.S. Environmental Protection Agency, Office of Water, EPA 822-B-96–002, Oct. 1996, 11 p. For additional information write to: New York Water Science Center U.S. Geological Survey 30 Brown Rd. Ithaca, N.Y. 14850
Information requests: (518) 285-5602 or visit our Web site at; http://ny.water.usgs.gov Miller and others—Hydrogeology of Two Areas of the Tug Hill Glacial-Drift Aquifer, Oswego County, New York—Scientific Investigations Report 2007-5169