Niagara Power Project FERC No. 2216





HTML Format.  Text only


Prepared for: New York Power Authority 

Prepared by: URS Corporation; Gomez and Sullivan Enigeers, P.C.; and E/PRO Environmental Consulting, LLC


August 2005




Copyright © 2005 New York Power Authority



Aquifer                               Saturated rock or sediment sufficiently permeable to supply economic quantities of water to wells or springs

Aquitard                             A low-permeability unit that can store groundwater, but that transmits groundwater slowly

Artesian water                    Groundwater that is under pressure sufficient to raise it above the level at which it is encountered in a borehole or well

BTEX                                   Abbreviation for benzene, toluene, ethylbenzene, and xylenes, all classified as volatile organic compounds, constituents of motor fuels, and commonly used as indicators of soil and water contamination by such fuels.

Barometric pressure            Atmospheric pressure measured by a barometer.  Changes in atmospheric pressure are capable of inducing changes in water elevation

Bedding plane                     In sedimentary rocks, the planes or surfaces that separate individual layers, beds, or strata that tend to split more or less horizontally or parallel to ground surface

Bedrock                              Solid rock either exposed at the surface of the earth or overlain by unconsolidated material

Bioherm                             A carbonate rock formation in the form of an ancient reef or hummock, consisting of the fossilized remains of corals, algae, mollusks, and other sedentary marine life, and commonly surrounded by rock of a different lithology

Boundary Effect                 The potentiometric head in the aquifer is affected by fluctuating water levels in a surface water body. This condition forms a boundary of the aquifer. Energy in the form of pressure waves is transmitted from the surface water body into the aquifer with the energy gradually diminishing with distance from the fluctuating source.

Clast                                  an individual grain or constituent particle of a rock

Confining unit                    A low-permeability material that lies adjacent to an aquifer and confines groundwater within the aquifer.  It may lie above or below the aquifer.

Crinoid                               any of a large class (Crinoidea) of echinoderms, such as the sea lily, usually having a somewhat cup-shaped body with five or more feathery arms atop a stalk.  Found from Ordovician to the present.

Discharge                           The process by which water is removed from a groundwater system along a discharge area, which may include a spring, seepage from an excavation face, or inflow to a stream

Dolomite                            A limestone rock that contains magnesium carbonate, or the mineral dolomite that is not easily weathered or dissolved

Drainage system                 As applied to groundwater, a mechanical system that locally increases flow and facilitates the area drainage of groundwater

Escarpment                        A steep-faced linear ridge frequently presented by the abrupt termination of sedimentary rock layers

Flow pattern                       The direction of movement of groundwater both horizontally and vertically.  Flow patterns may change with depth and geologic unit

Fractures                            Breaks in rock occurring at a variety of possible angles due to intense folding or faulting, or in response to glacial unloading or stress release

Forebay Efficiency              The relationship between the cyclic water level fluctuations in the forebay and corresponding groundwater fluctuations observed in each well.  The efficiency of the well is a mathematical representation of the “connectedness” of a well and is calculated by dividing the net change in groundwater elevation (for half of one complete cycle) by the corresponding net change in forebay level.  Efficiency values range between 0.0 and 1.0 (i.e. 0 to 100%) with 0 indicating no connection and 1.0 indicating a complete connection.

Geologic Dip                      The angle at which a stratum or bedding plane of a sedimentary rock is inclined from the horizontal

Gradient                             In an aquifer, the rate of change of total head per unit of distance of flow at a given point and in a given direction (i.e., upgradient or downgradient)

Groundwater                      Underground water occupying openings, cavities, and spaces in rock and sediments

Groundwater divide            The line of separation between groundwater flow systems.  It marks the high point of groundwater elevations, with lower groundwater elevations and flow moving away from this divide

Groundwater Modeling       A three dimensional numerical model was used to simulate groundwater conditions in an aquifer. The model utilized boundaries, multiple layers, rows and columns that formed individual cells. The modeling tool was used to estimate flow rates and flow direction within the aquifer.

Groundwater Sink               The conduit drainage system acts as a groundwater sink that is, it has a drain effect on groundwater in the vicinity of the drainage system.

Grout curtain                      Grout-filled segment approximating an impermeable wall in bedrock, formed by the pressure-injection of cement grout into a linearly spaced sequence of boreholes.

Headwater                          The upstream end or upper tributaries of a stream or river

Hydraulic boundary            A boundary to the flow of water, such as a groundwater divide or low-permeability rock unit

Hydraulic conductivity        Rate at which a fluid moves through a given permeable material under a hydraulic gradient (driving force) equal to 1.0 (i.e., rise equals run).  Ranges of hydraulic conductivity have been determined for various geological materials

Hydraulic gradient              The slope of an underground water surface expressed as the change in total head (i.e., groundwater surface elevation) with change in distance in a given direction

Hydraulic head                   The pressure exerted by a fluid upon a unit area (surface) due to the height at which the fluid level stands above the surface.  Usually expressed as pounds per square inch, sometimes as actual feet of head or fluid column. total head or head

Hydrogeology                     The study of geological factors relating to the occurrence and movement of underground water and its relationship to surface water and rainfall

Infiltration                          The flow of water downward from the land surface into and through the underlying soil or rock

Joints                                 Fractures in rock that occur more or less vertical to bedding, along which no appreciable movement has occurred.

Lacustrine                          Sediments deposited in a lake, consisting of layers of clay, silt, and fine sand

Leakance                            The ability of an aquitard to transmit vertical flow between two horizontal aquifers. Leakance is defined as the ratio of its vertical hydraulic conductivity to its thickness. It is synonymous with coefficient of leakage.

Limestone                          A bedded, fine-textured sedimentary rock consisting chiefly of calcium carbonate.

Linear drain                        A linear feature towards which groundwater converges and is discharged, as at the face of a linear excavation or escarpment face

Non point - source              A diffuse, indefinable area over which discharge of a fluid or other substance occurs.

Operational Cycle               Daily fluctuation in electrical power demands are reflected in water level changes in the Lewiston Reservoir. During low demand at night water is added to the reservoir. During high demand hours during the day water is drained from the reservoir for power generation.

Overburden                        Loose, unconsolidated material (soil) that rests upon solid rock

Packer test                          A test of the transmissivity of rock in proximity to a borehole.  A pair of connected, expandable/retractable rubber glands (called a packer assembly) is set within a borehole to isolate a single test interval. The inflated packer assembly prevents water from moving above or below the test interval in the borehole. Water is injected under pressure through the packer assembly and into the rock formation.

Permeability                       Capacity of a soil or rock to transmit a fluid.  Depends upon the size and shape of the pores and their interconnection.  It is measured by the rate of fluid movement in the porous medium.

Piezometer                         A tube set into a borehole, through which tube the distance of the piezometric surface below ground surface may be measured.

Piezometric surface             A surface that represents the level to which water will rise in a tightly sealed piezometer or well.  The water table is a piezometric surface for an unconfined aquifer.  Also called a potentiometric surface

Point drain                         A point to which groundwater converges and is discharged, as to a supply well or quarry excavation

Point source                       A discrete, identifiable point or area from which a discharge of a fluid or other substance occurs, commonly into air or a water body.

Recharge                            The process by which water is added to groundwater, which may include the downward infiltration of precipitation or inflow from streams or other surface water bodies

Reservoir Efficiency            The relationship between the cyclic fluctuations in the reservoir and corresponding groundwater fluctuations observed in each well.  The efficiency of the well is a mathematical representation of the “connectedness” of a well and is calculated by dividing the net change in groundwater elevation (for half of one complete cycle) by the corresponding net change in reservoir level.  Efficiency values range between 0.0 and 1.0 (i.e. 0 to 100%) with 0 indicating no connection and 1.0 indicating a complete connection.

SVOCs                               Abbreviation for semivolatile organic compounds.  SVOCs tend to evaporate or volatilize relatively slowly at standard conditions of temperature and pressure.

Sedimentary rock                Rock formed by the accumulation of sediments or chemical precipitates (e.g., gypsum) that forms bedding layers

Shale                                  A sedimentary rock made up of clay- and silt-sized particles, hardened into rock

Sinusoidal fluctuation         Changes in a characteristic (such as water level) that, when plotted, appear as a regularly undulating, smooth, up-and-down curve about a central horizontal axis (sine curve)

Spatially and temporally      Water withdrawn from wells located some distance from each other and withdrawn at different times.

Stratigraphic Unit               Recognizable unit consisting of stratified, mainly sedimentary, rocks grouped for description and mapping over an area

Stylolite                             A surface or contact, usually between two layers of carbonate rock, that is marked by an irregular, interlocking penetration, in cross-section resembling a row of interlocking columns or teeth on one side fitting into their counterparts on the other side.

Subcrop                              A “subsurface outcrop” that describes the areal limits of a truncated rock unit at a buried unconformity surface.

Till                                     Sediments deposited by the glacial ice sheet, consisting of a mixture of clays, silt, and sand, with cobbles and boulders

Topographic boundary        A physical feature of the land surface that forms a boundary, such as a ridge or stream

Transmissivity                    The rate at which water is transmitted through a unit width of the aquifer under a unit hydraulic gradient.

Unconfined Aquifer            An aquifer not confined from above by low-permeability material, having a water table surface between unsaturated material above and saturated material below

VOCs                                 Abbreviation for volatile organic compounds.  VOCs tend to evaporate or volatilize readily at standard conditions of temperature and pressure.

Water table                         The upper surface of the zone of groundwater saturation.  Above the water table, the pores in soil or rock are unsaturated, i.e., not completely filled with water.

Weir                                   A device, usually a low dam, placed across a stream or flow section to control and measure flow volume


The New York Power Authority (NYPA) is in the process of relicensing the Niagara Power Project (NPP), in Lewiston, New York.  As part of the relicensing process, NYPA is developing information related to various aspects of the NPP, including assessment of the effect of Project operations on groundwater flow patterns and groundwater quality.  This report describes the activities and presents the findings of the groundwater investigation conducted in the Project vicinity from spring 2003 through spring 2004.

The investigation area is bounded to the north by the Niagara escarpment, to the east by the Tuscarora Nation eastern boundary/Cayuga Creek, to the south by the upper Niagara River, and to the west by the lower Niagara River.

Investigation Activities

The scope of work developed to perform this evaluation included the following tasks:

1.      Groundwater flow modeling – this task involved using an existing regional groundwater flow model prepared by the United States Geological Survey (USGS), and preparing a focused model of groundwater flow in the vicinity of the Lewiston Reservoir.  The model was used to evaluate NPP effects on groundwater flow patterns.

2.      Groundwater monitoring well installation – this task involved drilling and installing nested groundwater monitoring wells throughout the study area in order to evaluate NPP effects on groundwater levels and groundwater quality.  A total of 91 nested piezometers were installed at 17 groundwater sampling locations.

3.      Groundwater level monitoring – this task involved completion of a groundwater level monitoring program to assess groundwater level fluctuations and groundwater flow in the vicinity of the conduits and reservoir.  Monitoring was performed using water level probes and pressure transducers equipped with dataloggers. A detailed study of water level fluctuation effects on groundwater quality in the vicinity of the conduits was also conducted using electronic water quality probes/dataloggers.

4.      Falls Street Tunnel flow monitoring – this task involved placement of flow metering equipment at three locations within the Falls Street Tunnel (FST)/South Side Interceptor (SSI) system in order to assess groundwater infiltration to the FST near the conduits as well as evaluation of factors contributing to groundwater infiltration.

5.      Surface water and groundwater quality sampling program – this task involved completion of three sampling events, with each event including collection of groundwater samples from the 91 piezometers installed as part of this investigation, and collection of surface water samples from 11 locations.

6.      Preparation of this report presenting the investigation activities and findings.

Groundwater Modeling

Groundwater modeling activities included: (1) conversion of the existing USGS model from MODFLOW computer code to Groundwater Modeling System (GMS) environment, (2) verification of accurate model conversion through comparison of flow budgets and predicted hydraulic heads for the two model formats, (3) preliminary modeling, and (4) focused modeling of the Lewiston Reservoir area.

Results of the focused model of the Lewiston Reservoir area indicate that the reservoir area of influence on groundwater flow is limited north and east of the reservoir, due to the presence of groundwater flow divides in these areas.  The northern groundwater flow divide (based on the results of numerical modeling) is located from 1,000 to 4,000 feet south of the Niagara escarpment, throughout the length of the escarpment, within the study area.  The presence of the northern groundwater divide prevents reservoir-influenced groundwater flow from migrating to and discharging from the Niagara escarpment.  A groundwater flow divide located approximately 1,500 feet east of the reservoir, determines the eastern extent of reservoir-influenced groundwater flow.  Vertically, the maximum extent of reservoir-influenced groundwater flow is seen in the upper weathered bedrock water-bearing zone, and decreases for each successively deeper water-bearing zone.  The two deepest identified water-bearing zones, the Gasport water-bearing zone and the Gasport/DeCew water-bearing zone, are not influenced by the reservoir, and therefore, there is no groundwater flow divide.  East of the reservoir, groundwater flow for these zones is inferred to be from the northeast, toward the reservoir.  East of the reservoir, the vertical extent of reservoir influenced groundwater flow extends to a depth of approximately 30 to 60 feet below ground surface.

Drilling and Well Installation

In order to evaluate Project effects on groundwater levels and groundwater quality, a total of 17 nested groundwater monitoring wells was installed in the vicinity of the NPP.  Eleven nested wells (GW03-001 through GW03-011) were installed in the vicinity of the Lewiston Reservoir, and six (GW03-012 through GW03-017) along and adjacent to the NYPA conduits.

Continuous rock coring was completed at each of the 17 well locations in order to log (observe and describe) the lithology within the borehole and identify water-bearing zones to target for piezometer placement.  The results of geologic descriptions are presented in geologic boring logs, and lithology was correlated between well locations as shown in geologic cross-sections presented in this report.  Throughout the area, bedrock is overlain by overburden materials consisting of glacial sediments (till and lacustrine silt and clay).  Thickness of overburden observed during this investigation ranged from approximately 3 feet to 40 feet.  The relatively simple bedrock stratigraphy identified during the drilling program exhibited a slight southward dip of approximately 30 feet per mile, with beds subcropping progressively northward at the overburden/bedrock contact.  At each nested well location, borings were completed through the entire thickness of the Lockport Group dolomite, which consists of the following dolomite formations (in order from oldest to youngest): Gasport, Goat Island, Eramosa, and Guelph.  At all locations, the borings were terminated in either the DeCew dolomite or the Rochester shale, both of the Clinton Group, which underlies the Lockport Group.

Water-bearing zones were identified by visual observation of fracturing and/or by packer testing of selected boreholes.  Discrete piezometers were placed to intercept nine targeted regional water-bearing zones, mainly associated with bedding planes at stratigraphic contacts.  A total of 91 discrete piezometers were placed at 17 nested well locations installed during this investigation.  The selected regional water-bearing zones were chosen to be representative of groundwater flow model layers presented in the USGS groundwater flow model.  One additional water-bearing zone, the Gasport/DeCew, situated beneath the lowest modeled water-bearing zones, was also targeted.  Due to the gentle southward dip of regional stratigraphy, the number of water-bearing zones decreases northward.  Near the Niagara escarpment, only two water-bearing zones within the Lockport, the weathered bedrock and Gasport/DeCew water-bearing zones, were present.  The uppermost water-bearing zone, (the weathered bedrock) is independent of stratigraphy and occurs throughout the investigation area.

Groundwater Level Monitoring Program

To evaluate NPP effects on area groundwater levels, a groundwater-level monitoring program was implemented.  The program consisted of water level measurement using a combination of electronic pressure transducers with dataloggers, manual water level measurements, and permanent surface water level gauges.  Water level data collection began in July 2003 and continued through May 2004.  Locations where water levels were measured included piezometers at nested groundwater monitoring wells, existing NYPA observation wells, the two NYPA conduit pump stations, and permanent gauges at surface water bodies (Lewiston Reservoir, forebay, and upper Niagara River).

Groundwater level data collected from monitoring wells throughout the study area were analyzed and compared to water level data for surface water bodies directly influenced by Project operations (e.g., forebay, reservoir, and conduits).  Responses to Project-induced water level fluctuations were evaluated in terms of reservoir and forebay efficiencies, presented as a percentage of the surface water level change, with 0% being no response, and 100% representing a groundwater level change magnitude equal to the surface water level fluctuation.

Efficiencies calculated for wells located near the reservoir ranged from 0.03% to 13.42%.  Generally, the wells showing the highest reservoir efficiency were NYPA observation wells located immediately adjacent to the reservoir.  Groundwater level fluctuations resulting from daily reservoir fluctuations observed in these wells were on the order of 1 to 2.5 feet.  Of the piezometers installed in bedrock as part of this study, the highest efficiencies were exhibited in the upper weathered bedrock zone, which has the most direct contact with reservoir water.  Groundwater level fluctuations observed in these piezometers (installed at distances greater than 600 feet from the reservoir) were on the order of one foot or less.  In general, fluctuations in groundwater levels around the reservoir showed limited reservoir influence relative to natural seasonal effects observed to be on the order of approximately 7 feet.

Groundwater levels north of the reservoir along a north-south line defined by wells GW03-001 to GW03-004 confirm the presence of the northern groundwater flow divide in the upper weathered bedrock water-bearing zone as well as in the underlying Gasport/DeCew water-bearing zone.  Based on water level data from these wells, the groundwater divide at this location is between 1,000 feet and 3,500 feet north of the reservoir.

Forebay water level fluctuations directly impact groundwater levels along the conduits.  Water levels measured in piezometers installed along the conduits were seen to fluctuate as much as approximately 7 feet due to forebay fluctuations.  This effect, transmitted via the conduit drainage system (CDS), affects hydraulic gradients within the CDS as well as groundwater flow to and from the CDS.  When forebay water levels are low, hydraulic gradients within the conduits are stronger from the south to the north (i.e., towards the forebay), and groundwater flow gradients from the surrounding groundwater toward the CDS are greater.  Conversely, higher forebay water levels create a decreased northward hydraulic gradient, and even reverse to the south at times.  Under these conditions groundwater flow gradients toward the conduits from surrounding bedrock are reduced , and even turn  away from the conduits at times.  The extent of the effect of the conduits on groundwater has been estimated by others to be approximately ½ mile on either side of the conduits.

Average forebay water levels are 5 feet lower during non-tourist season than during tourist season.  As a result, CDS hydraulic heads are lower during non-tourist season, which may increase the amount of groundwater flow to the conduits.

Falls Street Tunnel

As part of the evaluation of NPP effects on groundwater infiltration into the FST, a tunnel reconnaissance was performed to assess existing tunnel conditions and evaluate possible flow measuring opportunities.  During tunnel reconnaissance, more significant groundwater infiltration was observed upstream (east) of the conduit crossing than was observed in the downstream (west) portion.  Based on these observations, a flow monitoring program was developed that incorporated measurement of flows at three locations within the FST/SSI system.  The sum of these measurements were taken as representative of all FST groundwater infiltration at the conduit crossing.

The flow monitoring program consisted of installing and maintaining electronic flow meters at three locations: (1) a depth/velocity probe placed at the flow measuring weir at Drop Shaft 12, (2) an area/velocity probe placed in the SSI west bypass tunnel, and (3) an area/velocity probe placed in the SSI east bypass tunnel.  Maintenance activities during the flow monitoring program included checking the datalogger status and downloading data for each of the three meters every three days.  A rain gauge was also installed at Pump Station A in order to track precipitation events.  The FST flow monitoring program was performed from October 20, 2003, through November 25, 2003.  This flow monitoring period was designed to coincide with the change from tourist season to non-tourist season on November 1, 2003.

Results of flow monitoring indicated that during tourist season, total dry-weather groundwater infiltration into the FST is approximately 5,000 gallons per minute (gpm) (7.2 million gallons per day [mgd]), while during non-tourist season, total dry-weather groundwater infiltration is approximately 4,100 gpm (5.9 mgd).  Groundwater infiltration east of the FST/conduit crossing is approximately twice that west of the crossing.  Evaluation of infiltration rates and local groundwater levels indicate a direct correlation of FST infiltration with fluctuations in forebay levels.

Results of an evaluation of possible groundwater flow pathways indicate that 75 to 85% of infiltration into the FST likely comes from the external drainage system associated with the NYPA conduits (Conduit Drainage System, or CDS).  Approximately 15 to 25% of the flow is estimated to originate as flow from the Niagara River through the high-transmissivity zone in the Lockport aquifer. 

Surface Water and Groundwater Quality

To evaluate groundwater quality in the vicinity of the Project, a surface water and groundwater sampling program was implemented.  This sampling program involved collection of groundwater samples from the 91 piezometers installed in nested groundwater monitoring wells and collection of surface water samples from 11 locations.  Sampling events were conducted at approximately 3-month intervals starting in September 2003.  Sampling events conducted during this study were completed in September/October 2003, November/December 2003, and February/March 2004.

Surface Water Quality

Surface water samples were collected from six river-sourced locations (i.e., from locations whose primary source is the upper Niagara River) in the vicinity of the NPP.  These were: (1) the upper Niagara River itself, at the conduit intakes, (2) one each from the two conduits at the air intake vents located immediately north of Royal Avenue in the City of Niagara Falls, (3) one from the forebay at the Robert Moses Power Plant, and (4) two from the Lewiston Reservoir—one from the west side near the Lewiston Pump Generating Plant, and one from the east side at the eastern reservoir dike.  Five local-sourced (i.e., area creeks and swamps/ponds near Lewiston Reservoir whose source is not the upper Niagara River) surface water samples were collected from Fish Creek (one upstream and one downstream of the reservoir), Gill Creek (one upstream and one downstream of the reservoir), and the wetland at the head of Gill Creek located on the Tuscarora Nation.

Surface water sampling analytical results indicate that river-sourced water is generally free of significant contamination.  Monomethyl mercury was detected in at least one sampling event at very low concentrations in all surface water sample locations except the forebay and the western reservoir sample location (SW03-006).  Detection of monomethyl mercury at these very low concentrations in surface waters throughout the study area is likely indicative of the regional presence of inorganic mercury due to naturally occurring sources (atmospheric deposition) or exogenous sources (industrial sources).  Other than monomethyl mercury, no chemical contaminants were detected in surface water samples collected from the upper Niagara River at the intakes or the forebay during any of the three sampling events.  Two pesticide compounds, delta-BHC and 4,4-DDT, were detected in very low concentrations in a conduit sample (4,4-DDT), and in both reservoir sample locations (delta-BHC).

In the water samples collected from Fish and Gill Creeks, relatively low concentrations of acetone, and carbon disulfide were detected. Lead was detected in the surface water samples collected from the wetland at the head of Gill Creekduring the October and November 2003 sampling events.

Groundwater Quality

Groundwater in the study area may be conceived of as two flow systems: a freshwater system, and a (deeper) saline flow system.  Generally, the groundwater in the freshwater flow system is moderately to highly mineralized, containing (1) sulfates dissolved from soluble gypsum within the dolomite, and (2) calcium and magnesium bicarbonate, also dissolved from dolomite. 

Groundwater Quality in the Conduit Vicinity

Groundwater in the vicinity of the conduits is affected by contaminant plumes originating from various active and inactive hazardous waste sites in the region.  The worst contamination adjacent to the conduits was identified in the once heavily industrialized southern portion of the study area near the upper Niagara River.  In this area, the conduits pass beneath chemical storage and manufacturing facilities, landfills, and other sites of present or past industrial operations.  The predominant compounds identified in groundwater in the vicinity of the conduits were chlorinated volatile organic compounds (VOCs) and non-chlorinated VOCs.  Compounds detected in this area with the greatest frequency and at the highest concentrations were 1,2-dichloroethene (cis), trichloroethene (TCE), tetrachloroethene (PCE), vinyl chloride, and benzene.

The worst contamination, consisting mainly of chlorinated and non-chlorinated VOCs, was identified in the southern portion of the conduit right-of-way (well locations GW03-014 through GW03-017).  Contaminants detected in this area with the greatest frequency and at the highest concentrations were detected in the Gasport/DeCew water-bearing zone.  With a few exceptions, the water-bearing zones intersecting the conduits and CDS exhibited lower overall contaminant concentrations.  This was possibly due to dilution of contaminated groundwater by the influence of Niagara River water within the CDS.

Contaminants detected in the northern portion of the conduit right-of-way (piezometers GW03-012 and GW03-013) were predominantly gasoline-related compounds (including benzene, toluene, ethylbenzene, and xylenes, or BTEX), all detected at relatively low concentrations.  Other analytes detected included carbon disulfide, caprolactum, cadmium, lead, phthalates, phenols, MEK, PCE, and TCE.  Piezometer GW03-012A-P3, screened in the Gasport/DeCew water-bearing zone in the vicinity of the forebay, exhibited the highest contaminant concentrations in the northern portion of the conduit right-of-way.

Observed variations in contaminant concentrations for some of the water-bearing zones intersecting the conduits seem to indicate an effect related to Treaty flow changes between tourist and non-tourist season.  For instance, contaminant concentrations detected at piezometer GW03-015B-P5 increased from non-detect in October (tourist season) to 720 ug/L in December (non-tourist season).  The cause of this change is likely to be lower hydraulic heads within the CDS (caused by lower forebay levels during non-tourist season) creating a steeper groundwater flow gradient from the surrounding bedrock toward the conduits, and thus transporting greater amounts of contaminants from surrounding plumes. . Results of a detailed study of water level fluctuation effects on groundwater quality in the vicinity of the conduits confirm that short-term forebay induced CDS water level fluctuations influence water quality immediately adjacent to the conduits. Groundwater quality was shown to fluctuate with water levels indicating a greater river water influence during high CDS water levels, and a greater groundwater influence during lower CDS water levels. During tourist season, when average forebay levels are higher, the gradient toward the conduits is reduced and groundwater quality is less influenced by surrounding contaminant plumes.

Groundwater Quality in the Reservoir Vicinity

The most predominant chemical contaminants detected in groundwater in the vicinity of the Lewiston Reservoir were gasoline-related VOCs, such as BTEX and methyl tert-butyl ether (MTBE).  BTEX compounds and MTBE were detected in all water-bearing zones except 8A (Eramosa Unit A), with the highest concentrations east of the reservoir.  As these compounds were not detected in water samples collected from the Lewiston Reservoir itself, their detection in piezometers installed near the reservoir is likely related to the presence of fuel service stations in the area, with possibly leaking underground storage tanks.

The most significant detections of gasoline related compounds were in the deeper water-bearing zones, particularly in piezometers screened in the Gasport/DeCew water-bearing zone (e.g., GW03-005A-P2, GW03-007A-P3, and GW03-009A-P3).  Significant concentrations of MTBE were also identified in overlying water-bearing zones at well GW03-009, located southeast of the reservoir.  The predominant contaminants detected in samples collected from the Gasport/DeCew water-bearing zone were the BTEX compounds, whereas MTBE was the predominant compound detected in the overlying water-bearing zones.  Benzene concentrations detected in the Gasport/DeCew water-bearing zone ranged from 20 ug/L in piezometer GW03-007A-P3 to 280 ug/L in GW03-005A-P2.  Relatively low concentrations of MTBE (less than 5 ug/L) were also detected in water-bearing zones, downgradient of GW03-009 at location GW03-010.  The two deepest zones, the Gasport/DeCew water-bearing zone and the zone immediately overlying it, which exhibit the highest gasoline contaminant concentrations, are not subject to reservoir-induced groundwater flow.

Significant BTEX concentrations at reservoir area piezometers were generally detected in deeper saline water-bearing zones that exhibit high chloride levels (greater than 500 mg/L).  The upper weathered bedrock water-bearing zone generally exhibited low contaminant concentrations.  Gasoline related contaminants likely migrate downward through vertical fracturing to the deeper zones.  Higher contaminant concentrations in the deeper zones may be caused by relative lack of groundwater flow in these zones, which would allow contaminants to accumulate, whereas the relatively higher groundwater flow in the upper water-bearing zones may act to dilute or disperse contaminants in these zones.

Non-gasoline-related analytes detected in samples collected from reservoir area wells (GW03-001 through GW03-011) included acetone, methyl ethyl ketone, 1,1,1-trichloroethane, chloroform, carbon disulfide, arsenic, cadmium, lead, monomethyl mercury, bis (2-ethylhexyl)phthalate, phenol, and caprolactum.  The source or sources of these contaminants is unknown.  With the exception of monomethyl mercury, none of these contaminants was detected in water samples collected from the Lewiston Reservoir itself.

Contaminants detected in Lewiston Reservoir sediments, with the exception of lead and arsenic, were not detected in groundwater samples collected from nearby piezometers.  Lead and arsenic were detected in several reservoir area piezometers associated with different water-bearing zones within the groundwater flow influence of the Lewiston Reservoir.  The presence of monomethyl mercury in several reservoir vicinity piezometers may also be indicative of the presence of elemental mercury, possibly from naturally occurring sources or from reservoir sediments, where mercuric compounds were also detected.  Monomethyl mercury was detected in reservoir water samples as well.  It should be noted that these three analytes (arsenic, lead, and monomethyl mercury) were also detected in piezometers that are not within the apparent direct influence of reservoir-induced groundwater flow (namely, the Gasport/DeCew water-bearing zone), a fact that suggests the possibility of another source.



The New York Power Authority (NYPA) is engaged in the relicensing of the Niagara Power Project in Lewiston, Niagara County, New York.  The present operating license of the plant expires in August 2007.  As part of its preparation for the relicensing of the Niagara Project, NYPA is developing information related to the ecological, engineering, recreational, cultural, and socioeconomic aspects of the Project.  This report presents the results of detailed investigations of effects of Project operations on groundwater flow and quality in the vicinity of the Project.

1.1         Project Background

The 1,880-MW (firm capacity) Niagara Power Project (NPP) is one of the largest non-federal hydroelectric facilities in North America. The Project was licensed to the Power Authority of the State of New York (now the New York Power Authority) in 1957.  Construction of the Project began in 1958 and the electricity was first produced in 1961.

The Project has several components (Figure 1.1-1).  Twin intakes are located approximately 2.6 miles above Niagara Falls.  Water entering these intakes is routed around the Falls via two large low-head conduits to a 1.8-billion-gallon forebay, lying on an east-west axis about 4 miles downstream of the Falls.  The forebay is located on the east bank of the Niagara River.  At the west end of the forebay, between the forebay itself and the river, is the Robert Moses Niagara Power Plant (RMNPP), NYPA’s main generating plant at Niagara.  This plant has 13 turbines that generate electricity from water stored in the forebay.  Head is approximately 300 feet.  At the eastern end of the forebay is the Lewiston Pump Generating Plant (LPGP).  Under non-peak usage conditions (i.e., at night and on weekends), water is pumped from the forebay via the plant’s 12 pumps into the 22-billion-gallon Lewiston Reservoir, which lies east of the plant.  During peak usage conditions (i.e., daytime Monday through Friday), the pumps are reversed for use as generators, and water is allowed to flow back through the plant, producing electricity.  The forebay, therefore, serves as headwater for the Robert Moses plant and tailwater for the LPGP.  South of the forebay is a switchyard, which serves as the electrical interface between the Project and its service area.

For purposes of generating electricity using the Niagara River, two seasons are recognized:  tourist season and non-tourist season.  By the provisions of the 1950 Niagara River Water Diversion Treaty, at least 100,000 cfs must be allowed to flow over Niagara Falls during daytime and evening hours in the tourist season (April 1 – October 31), and at least 50,000 cfs at all other times.  Canada and the United States are entitled by international treaty to produce hydroelectric power using the remaining flows.

Water level fluctuations in the Chippawa-Grass Island Pool (in the upper Niagara River) are limited by an International Joint Commission directive to 1.5 feet per day unless conditions triggering special provisions occur.  It is important to note that water level fluctuations in both the upper and lower Niagara River occur due to a number of factors other than operation of the NPP.  These may include wind, natural flow and ice conditions, as well as operation of power plants on the Canadian side of the river.   

Water-level fluctuations in the lower Niagara River (upstream of the RMNPP tailrace) from all causes can be as great as 12 feet per day.  Most of this daily fluctuation is due to the change in the treaty-mandated control of flow over Niagara Falls.  Water level fluctuations downstream of the RMNPP tailrace are much smaller.  The average daily water level fluctuation 1.4 miles downstream of the RMNPP tailrace, during the 2002 tourist season, was approximately 1.5 feet.

Operation of the NPP can result in water level fluctuations in the Lewiston Reservoir of 3-18 feet per day, and as much as 38 feet per week.

1.2         Study Objectives

The objectives of this investigation were developed in consultation with the stakeholders as part of the scoping of this study. The objectives of this investigation are to:

·        Document existing information on groundwater hydraulic influence;

·        Use existing information to determine the effect of river water fluctuation on groundwater flow patterns along the Niagara River;

·        Determine interaction between conduit-transported river water, flow at conduit weir(s), flow within the conduit external drainage system, infiltration of groundwater and/or surface water into the Falls Street Tunnel (FST), and infiltration of groundwater into the forebay;

·        Determine effects of Project features and operations on Tuscarora Nation, the Town of Lewiston and other surrounding communities groundwater flow and water quality (chemical and biological);

·        Assess the impact on water quality and flow of surface waters receiving groundwater due to Project operations.

1.3         Investigation Area

The investigation area is bounded to the north by the Niagara escarpment, to the east by the Tuscarora Nation eastern boundary/Cayuga Creek, to the south by the upper Niagara River, and to the west by the lower Niagara River (Figure 1.1-1).

1.4         Report Organization

This report is organized into seven sections:

1.0       Introduction

2.0       Environmental Setting

3.0       NYPA 2002 Groundwater Study

4.0       Investigation Activities

5.0       Results

6.0       Discussion

7.0       Conclusions

Section 2.0 presents a summary of the environmental setting within which this study is being conducted.  This section includes discussion of natural features such as topography, surface water, and hydrogeology.  Also discussed are manmade features such as the NPP itself (conduits, forebay, and Lewiston Reservoir), other features exhibiting hydraulic influence such as sewers, tunnels, and the LaFarge Redland Quarry, and existing groundwater contaminant plumes within the study area.

Section 3.0 presents a summary of conclusions for the report, Groundwater Flow Investigations in the Vicinity of the Niagara Power Project (URS et al. 2003).  This report describes the study, conducted in 2002, of NPP effects on water levels in the vicinity of the Project.  It was prepared as a preliminary look at NPP effects in order to establish context for the more detailed study discussed in this report.

Section 4.0 presents the scope of work conducted for the various data collection efforts included in this overall study of NPP effects on groundwater.  This section discusses investigation activities performed for (1) groundwater modeling, (2) geologic investigation and groundwater monitoring well installation, (3) a groundwater level monitoring program, (4) investigation of groundwater infiltration into the Falls Street Tunnel (FST), and (5) a water quality sampling program including collection of both surface water and groundwater samples.

Section 5.0 presents the results of the data collection efforts described immediately above.  The results are presented in the same order as the tasks presented in Section 4.0.

Section 6.0 presents a detailed discussion of the results of all of the data collection efforts in the context of the relationship between the NPP and groundwater levels, groundwater flow, groundwater infiltration into the FST, and surface water and groundwater quality in the vicinity of the NPP.

Section 7.0 presents conclusions developed as a result of this detailed evaluation of the relationship between the NPP and groundwater flow and groundwater quality.

Figure 1.1-1

Investigation Area

[NIP – General Location Maps]


2.1         Natural Elements

To properly assess the effects of the operation and features of the Niagara Power Project (NPP) on local groundwater flow it is necessary to have an understanding of the natural elements that affect flow, such as topography, hydrology, geology, and hydrogeology within the investigation area.

2.1.1        Topography

Except for the Niagara escarpment, the Falls, and the Niagara River gorge, the area around the Project is of relatively low relief (Figure 2.1.1-1).       Niagara Escarpment

The Niagara escarpment intersects the river on a generally east-west axis downstream of Niagara Falls, with its greatest relief (200 to 250 feet) at the river.  South of the escarpment, surface elevations range from El. 575 feet (United States Lake Survey Datum [USLSD] 1935, the datum used for construction of the Project) along the upper Niagara River to El. 650 feet on Tuscarora Nation lands to the north (Figure 2.1.1-1).  Immediately north of the escarpment, surface elevations are approximately El. 350 feet. 

The carbonate bedrock units (dolomites and limestones) in the investigation area are more resistant to weathering than the shale units above and below and consequently form the most predominant part of the Niagara escarpment.       Niagara Falls and Niagara Gorge

The Niagara gorge runs north-south, with its southern end at the Falls and its northern end at the Niagara escarpment.  The depth of the gorge is approximately 250 feet below the surrounding land surface.  Approximately halfway along the gorge, downstream of the Falls, is the Whirlpool, a unique structure formed following the most recent glaciation when the river intersected a pre-glacial riverbed.

2.1.2        Surface Water

All surface water bodies in the Project vicinity drain northward to Lake Ontario either directly or indirectly via the Niagara River.  Streams in the investigation area drain either south or west to the Niagara River       Niagara River

For the purposes of discussion, the Niagara River may be subdivided into the upper Niagara River (i.e., the river upstream of the Falls) and the lower Niagara River (i.e., the river below the Falls) (Figure 2.1.1-1).  Flow at the river’s head at Buffalo, averages 212,300 cubic feet per second (cfs). 

The upper Niagara River, including the Chippawa-Grass Island Pool, is located in the southern portion of the investigation area.  The Chippawa-Grass Island Pool is an approximately 3-mile long reach extending from the northern tip of Grand Island to the International Niagara Control Structure (a linear array of 18 sluice gates extending about halfway across the river from the Canadian side, 4,500 feet upstream of the Falls).  The International Control Structure, whose primary purpose is to regulate flow over Niagara Falls, also regulates surface elevation of the Pool.  From the Chippawa-Grass Island Pool, Niagara River water enters the conduits and flows to the forebay under the force of gravity alone.  By international agreement, flow over Niagara Falls must be at least 100,000 cfs during tourist season daylight hours, and at least 50,000 cfs at all other times, with remaining flow to be divided equally for power production purposes between the Niagara Power Project on the U.S. side of the river, and the two Sir Adam Beck plants on the Canadian side.

As mentioned above, flow over Niagara Falls is controlled primarily by the International Niagara Control Structure.  This structure also helps regulate water surface elevation in the Chippawa-Grass Island Pool.  According to international agreement, the change in surface elevation of the Chippawa-Grass Island Pool may not exceed 1.5 feet over a 24-hour period.

Below the Falls, namely in the lower Niagara River, flow continues northward through the gorge and past the escarpment, with discharge to Lake Ontario.  River water diverted for power generation above the Falls by both U.S. and Canadian hydro plants is returned to the river below the Falls (near the escarpment) via plant tailraces.       Gill Creek

The approximately 7-mile long Gill Creek originates in a wetland on the Tuscarora Nation.  It flows along the south dike of the Lewiston Reservoir and then south, eventually discharging to the upper Niagara River downstream of the intakes (Figure 2.1.1-1).

A small amount of Project water is discharged from the Lewiston Reservoir to Gill Creek in order to maintain a continuous flow in this small stream.  This augmentation flow ranges from a high of approximately 3 cfs in the summer  to a low of zero in winter and spring.  NYPA provides this augmentation flow from approximately June through September as part of an agreement with the City of Niagara Falls.  As noted in ENCRPB 1974, the purpose of this augmentation is to enhance the recreational use of Gill Creek as it flows through Hyde Park (in the City of Niagara Falls) by reducing stagnation and improving the appearance of the creek.       Fish Creek

Fish Creek, which is approximately 5 miles long, flows westward from headwaters in the center of the Tuscarora Nation along the east and north dikes of the Lewiston Reservoir, and eventually discharges to the lower Niagara River downstream of the NPP tailraces (Figure 2.1.1-1)       Cayuga Creek

South of the Tuscarora Nation, Cayuga Creek demarcates the eastern boundary of the investigation area.  Based on its classification as a groundwater drain in the model developed by Yager (1996), the creek is assumed to be a hydraulic boundary for the upper 45 feet of the Lockport Group, and therefore a logical eastern limit of the hydrogeologic investigation area.  Cayuga Creek flows to the south (roughly parallel to Gill Creek, and about 15,000 feet to the east).  It discharges into the upper Niagara (via the Little River) at Cayuga Island (Figure 2.1.1-1).

2.1.3        Hydrogeology       Stratigraphy

Due to extensive rock exposures in the Niagara River gorge and the many rock borings that have been completed for local environmental investigations, the regional geology of the investigation area is well understood (Figure 2.1.3-1).


Throughout the region, a relatively thin layer of unconsolidated overburden overlies generally horizontal layers of sedimentary bedrock.  The overburden layer in the Project vicinity is 5 to 15 feet thick, although in some places (southeast of the Project, along Tonawanda Creek) it reaches 80 feet.  Three types of unconsolidated deposits comprise the overburden: (1) glacial till; (2) layers of lacustrine, or lake-deposited, clays, silts, and fine sands; and (3) lenses of sand and gravel (Johnston 1964).  The till, which typically overlies bedrock directly, consists of an unsorted mix of boulders, clay, silt and sand deposited by the glacial ice sheet.  The lacustrine deposits, which typically overlie the till, were deposited as bottom sediments in temporary lakes formed at the terminus of the glacier as it melted.  Sand and gravel lenses are found wherever the lacustrine and till deposits were reworked at the margins of temporary lakes.  They are uncommon in the Project vicinity.


Bedrock found within the investigation area is Silurian in age (deposited between 400 and 440 million years ago) and is classified into two groups: The Clinton Group and the Lockport Group (Brett et al. 1995)  (See Table 2.1.3-1 and Figures 2.1.3-2).  The predominant group in the investigation area is the Lockport Group, which consists of nearly horizontal massive- to medium-bedded, argillaceous (i.e., containing significant amounts of clay) dolomite (Figure 2.1.3-2).  The four formations compromising the Lockport Group are from bottom to top (i.e., from oldest to youngest):  Gasport Formation (dolomite), Goat Island Formation (dolomite), Eramosa Formation (dolomite), and Guelph Formation (dolomite).  Within the investigation area the thickness of the Lockport Group is approximately 150 feet.  The upper layers decrease in thickness and pinch out in the northern half of the investigation area.  Along the Niagara escarpment the thickness of the Lockport Group decreases to about 20 feet, and only the Goat Island and Gasport formations are observed.  Below the Lockport Group lie the DeCew dolomite and Rochester shale, the upper two of the five formations of the Clinton Group.  The DeCew Formation is typically 10-feet thick within the investigation area; the thickness of the Rochester Formation was not determined during this investigation but is reported to be 62-feet (Brett et al. 1995).       Groundwater


The presence of significant volumes of clay and silt in overburden soils in the Project vicinity, with inherent low hydraulic conductivities, prevents the overburden from being an economically important source of groundwater.  Given this slow rate of water movement, overburden deposits in the area are considered a confining unit, limiting the recharge that can occur into underlying bedrock.  Overburden groundwater flow is locally controlled, with minor horizontal flow along the more permeable, albeit infrequently occurring seams.  In some instances, bedding for underground utility lines, or former stream channels that cut into the surficial soil, may act as preferential pathways for horizontal flow.  Despite this, direction of groundwater flow in the overburden, although limited in volume, is predominantly downward, recharging the underlying bedrock aquifer through infiltration of rainfall and snowmelt by the force of gravity.  Preferential localized recharge of the bedrock aquifer occurs where natural soils have been disturbed, as by land filling or by excavation for structural foundations.  Where such activities have disturbed or removed the confining overburden layer, direct vertical groundwater migration pathways have been created between the surface and the underlying bedrock.

The nature of water-bearing openings occurring in the Lockport Group results in the following bulk hydrogeologic character: (1) a permeable zone in the top 15 feet, consisting of gypsum cavities and both vertical and horizontal bedding joints that have been widened by solution processes, and (2) several deeper, essentially horizontal, permeable zones characterized by bedding joints and surrounded by very low-permeability rock.  The bedding planes are approximately parallel to ground surface, with a slope or geologic “dip” to the south of about 30 feet per mile.

In the upper, more highly fractured zone, groundwater occurs under artesian, semiartesian, or unconfined conditions (artesian conditions predominating).  In the lower water-bearing zones, conditions are exclusively artesian.  The lower horizontal zones generally act as separate artesian aquifers (Johnston 1964) (Figure 2.1.3-4).

Johnston (1964) identified an area of high groundwater yield in the Lockport Group bedrock near the north shore of the upper Niagara River approximately two miles upstream of the Falls.  The area is approximately 2,000 feet wide and trends to the northeast from the shoreline for approximately 2 miles.  The point where the FST crosses over the twin conduits lies within this more highly fractured zone (Figure 2.1.3-5).

Bedrock Stratigraphy, Water-Bearing Zones, and Groundwater Model Layers

Bedrock of the Lockport Group comprises the only regionally extensive aquifer system in the Niagara Falls area.  The water-bearing zones of this aquifer consist of vertical and horizontal fractures and joints, dissolution zones along bedding planes, and small cavities from which gypsum has been dissolved.  Nine regionally extensive horizontal water-bearing zones have been defined within the Lockport Group; they are represented by model layers in a three-dimensional groundwater flow model presented in USGS Water-Supply Paper 2487  (Yager 1996) (Figure 2.1.3-6).  The model defines 10 layers with unique characteristics: the top most layer, Layer 1, represents unconsolidated glacial sediments and Layer 2 represents a stratigraphically independent layer of weathered bedrock.  Layers 3, 4 and 5 are bedrock formations that are not present within the investigation area.  Layers 6, 7, 8, 9, and 10 represent bedrock formations of the Lockport Group and in some cases are further subdivided into Regional Water-Bearing Zones (RWBZs [8A, 8B, 9A and 9B]).  The underlying Clinton Group, which has sufficiently low permeability to act as an aquitard, is not included or defined in the model.

2.2         Human Elements

2.2.1        Niagara Power Project       Conduits

NYPA’s two parallel low-head conduits, which carry water below ground for approximately four miles from the upper river to the forebay (Section, convey an average of 70,000 cfs .  The conduit intakes are located on the river approximately 2,000 feet upstream of the mouth of Gill Creek (Figure 1.1-1).  The conduits pass beneath the FST about 3,000 feet north of the intakes.  Constructed of reinforced concrete, the conduits are installed in 100- to 160-foot-deep trenches excavated in the Lockport Dolomite.  They are each horseshoe-shaped in cross-section, (flat on the bottom, arched on top), 46 feet wide at the base by 66.5 feet high in the center (Figures 2.2.1-1 and 2.2.1-2).

Conduit inspections were performed in 1989 and 1994 using a manned submersible vehicle.  Inspections were performed while the conduit intake gates were closed.  Because of equipment limitations, only approximately 34% of the conduits were inspected in 1989.  For the 1994 inspection, a custom umbilical cable was manufactured to allow inspection of the full length of both conduits.  Conduit 2, the western conduit, was inspected on September 17 and 18, 1994, and Conduit 1 was inspected on October 15 and 16, 1994 (BMD-Candive, 1994).  During the inspections, three areas of water inflow were observed at cracks in the conduit floor slabs of Conduit 1.  The approximate locations of these areas where groundwater inflow was observed within Conduit 1 are shown on Figure 2.2.1-3.  During inspections of Conduit 2, no areas of groundwater inflow were observed.       Conduit Drainage System

A network of gravity drains outside the walls and beneath the floor of the conduits serves to relieve hydrostatic pressure from groundwater (Figure 2.2.1-4).  The conduit drainage system (CDS) consists of four pipes paralleling the conduits; two 1- by 2-feet rectangular corner drains and two 24-inch half-round pipes beneath the floors.  The system also contains lateral pipes connecting all four horizontal drains at 40-foot intervals. Vertical drains located every 10 feet along the conduit walls connect the corner drains to the conduit backfill.  The CDS is hydraulically connected to two pump stations (Section, one near each end of the conduit system.       Pump Stations

Pump stations lie at two locations along the conduits.  The pump stations are designed to relieve excess hydrostatic pressure from around the conduits.  Under typical operating conditions, a system of weirs in each pump station acts to passively relieve excess hydrostatic pressure by allowing groundwater from the CDS to flow into the conduits when the CDS water level exceeds the weir elevation. In the event the conduits ever needed to be dewatered, the pump stations also accommodate pumping equipment to actively pump water from the CDS and relieve excess hydrostatic pressure from around the conduit walls. Pump Station A is located north of the conduit intakes, and Pump Station B is located just upstream of the forebay.  The pump stations consist of a system of concrete sumps extending to just below the full depth of the conduits (Figure 2.2.1-3).  The main sump, measuring approximately 26- by 40 feet, is connected to the CDS via a pair of 24-inch concrete pipes.  The two outer sumps, each measuring 4.5 by 8.33 feet, are connected to their corresponding east and west conduits via a pair of 5- by 7-foot connecting pipes.  The main and outer sumps are connected to each other via weirs equipped with flap gates.  The weir elevations for Pump Stations A and B are at El. 560.00 and 550.00 feet, respectively.  The main sump formerly housed the pumping equipment used for dewatering.  In 1974, however, to protect pumps and electrical equipment from corrosion damage, all such items were removed from both stations.  The gravity drainage system remains hydraulically connected to both conduits (via pump station sumps and weirs) and to the surrounding bedrock (Figure 2.2.1-3), which has been shown to be hydraulically connected to both the Niagara River and the forebay (Miller and Kappel 1987; and Yager and Kappel 1998).       Forebay

The forebay is located at the northern, or downstream, terminus of the conduits, between the Robert Moses Niagara Power Plant and the Lewiston Pump Generating Plant (Figure 1.1-1).  It lies on an east-west axis.  Niagara River water flows from the conduits into the forebay, where it becomes available for passage through the Robert Moses Plant or for pumping into the Lewiston Reservoir.  The forebay is approximately 4,200 feet long, 500 feet wide, and 110 feet deep.  The walls and base of the forebay consist of bedrock primarily of the Lockport Group, although a portion of its base, near the conduit outlet, is in the Rochester Formation (Figure 2.2.1-2).  Daily water level fluctuations in the forebay are dependent on the seasonal diversion schedule, the demand for power generation, and the flow of the Niagara River.  During 2003, water levels in the forebay ranged from elevations of approximately El. 537 feet to 564 feet for non-tourist season, and from approximately El. 540 feet to 568 feet for tourist season. (Figure 2.2.1-5).

To reduce permeability of fractured rock in the vicinity of the RMNPP, a grout curtain was installed along the walls of the lower Niagara River gorge and the forebay (Figure 1.1-1).  The gorge grout curtain extends across the Robert Moses Power Plant intake structure and extends approximately 2,000 feet on the north and south sides of the Forebay.  The forebay grout curtain extends for approximately 1,500 feet eastward from the gorge along the north wall of the forebay.  The gorge grout curtain extends approximately 275 feet into the Queenston Formation while the forebay grout curtain extends approximately 100 feet into the Rochester Formation.       Lewiston Reservoir

The Lewiston Reservoir occupies approximately 3 square miles of land east of the forebay and the Lewiston Pump Generating Plant (Figure 1.1-1).  The reservoir, constructed entirely above the ground surface, is surrounded by a 6.5-mile-long, 55-foot-high earth- and rock-filled dike containing a clay core.  Releases from the 1,900-acre reservoir are used to supply extra power during peak usage periods (daytime weekdays), with reservoir water being passed through the Lewiston Pump Generating Plant while water from the forebay is being passed through the Robert Moses Plant.  Reservoir water is replenished by pumping from the forebay during non-peak-usage periods (nighttime weekdays, and throughout the weekend).  Since water withdrawn for power generation during the week cannot be replenished as quickly as it is used, a net water loss occurs from Monday through Friday.  The reservoir exhibited a maximum net weekly drawdown of approximately 36 feet in 2002 (URS et al. 2005).  Normal Monday morning levels are restored by pumping from the forebay over the weekend.  In addition to a weekly cycle, reservoir surface water elevations exhibit a daily cycle over a 24-hour period Monday through Friday (URS et al. 2005) (Figure 2.2.1-5).  The average reservoir water level for 2003 was El. 649 feet for non-tourist season and approximately El. 643 feet for tourist season.

To reduce permeability of fractured rock beneath Lewiston Reservoir, a grout curtain was installed beneath the reservoir dike core (Figure 1.1-1).  The grout curtain was installed by injecting grout under pressure into boreholes drilled into bedrock spaced every 15 feet around the perimeter of the reservoir. The typical drill pattern consisted of grout injection boreholes drilled to depths of between 25 feet to 50 feet below the top of rock approximately every 15 feet, and boreholes drilled into the Rochester Shale approximately every 120 feet.

2.2.2        Falls Street Tunnel

The east-west trending FST, an approximately 3.2 mile-long unlined rock tunnel, was hand-excavated through the Lockport Dolomite in the early 1900s (Figure 1.1-1).  On its upstream (eastern) end it measures approximately 6 feet by 7 feet, and on its downstream (western) end, approximately 8 feet by 8 feet.  The purpose of this and other tunnels in the City of Niagara Falls was the transport of wastewater discharging from combined sewers.  In wet weather, the excess FST flow discharges directly to the Niagara River below the Falls.  In 1938, a wastewater treatment plant (primary treatment only) was constructed at the foot of Ashland Avenue, about one mile downstream of the Falls (Figure 1.1-1).  Wastewater was routed to this plant via the 3.3-mile-long Gorge Interceptor (GI) tunnel, which extended from the area of Devil’s Hole southward along the river (Figure 1.1-1).  In the late 1960s, a new 48-mgd treatment plant (on line by 1977) was constructed on Buffalo Avenue (above the Falls), and the Ashland Avenue plant was converted to the 20-mgd Gorge Pumping Station (GPS), designed to pump wastewater from the GI to the new plant via a new Gorge Forcemain.  As part of this overall effort, the Southside Interceptor (SSI) tunnel was completed, paralleling the FST (Figure 1.1-1).  Depending on wastewater volume, flow may be diverted to either the SSI or the FST.  During wet weather, a portion of FST flow is conveyed to the plant and the remainder is discharged to the river.  During normal (i.e., dry weather) conditions, 100% of FST flow is conveyed to the plant (Figure 1.1-1). 

The NYPA conduits pass beneath the FST at Royal Avenue (Figures 1.1-1 and 2.2.1-2).  Since the conduits pass below the FST, the FST had to be cut through at that point to construct the conduits.  The 300-foot section of tunnel opened for conduit construction was afterwards reconstructed as a 7-foot diameter concrete pipe contained within a concrete vault.  Two 100-foot sections of pipe were direct-buried immediately east and west of the vaulted section, as a transition from concrete pipe to the open-rock tunnel 

Groundwater infiltration into the FST has historically been a problem for the City of Niagara Falls.  As described below in Section 4.4, the greatest volume of groundwater inflow has historically been noted at the FST’s crossover of the NYPA conduits (CDM 1982).  In 1989, in an effort to decrease dry-weather flow and increase wet-weather capacity at the plant, the City of Niagara Falls initiated a number of measures to reduce groundwater flow into the FST or to divert it.  These efforts have included:

·        dewatering the vault and injecting grout sealant into vault cracks,

·        installing external seals around pipe joints within the vault,

·        installing new internal end seals where the pipe exits the vault,

·        injecting grout through the pipe in the direct-burial sections,

·        adding a diversion dam.

These efforts reduced inflow within the piped section to zero but diverted groundwater to the upstream and downstream rock sections, ultimately reducing inflow by only about one-half.  In 1994, it was reported (Roll and Lannon, 2001) that 6-7 mgd (about 20 percent of average treatment plant flow) was infiltrating a 200-foot section of the FST immediately downstream (west) of the NYPA conduits, while less than half that amount (approximately 3.5 mgd) was entering the upstream section.  In 2000, a grouting project was completed along the downstream (west) section of tunnel, reportedly reducing the inflow in that stretch to about 1.6 mgd (about 5 percent of average treatment plant flow). Therefore, not accounting for possible increased infiltration in the eastern stretch due to shifting from the grouted western section, the total infiltration near the conduits after the 2000 grouting effort would have been approximately 5 mgd.

2.2.3        LaFarge Redland Quarry

The LaFarge Redland Quarry is an operating limestone mine with a reported maximum depth of 140 feet below ground surface (approximately El. 484 feet).  It is located approximately 7,500 feet southeast of the Lewiston Reservoir (Figure 1.1-1).  In order to maintain a dewatered (i.e., operational) state within the mine, groundwater is extracted from sumps in the mine and discharged to a tributary of Cayuga Creek.  The extraction and discharge of groundwater at the mine is regulated by State Pollutant Discharge Elimination System (SPDES) permit #NY0025267. As reported by the New York State Public Notice for SPDES Renewal dated September 25, 2002, the mine is permitted to discharge a maximum of 432,000 gallons of water per day (300 gallons per minute) to Cayuga Creek.

2.2.4        Chemical Contamination

Toxic chemicals from various industrial processes in the Niagara Falls area have contaminated the soil and underlying groundwater as a result of either leaks and spills from plant operations, or from the disposal of waste products in lagoons, dump sites, and landfills.  In the area of the Niagara Power Project, 41 sites have been investigated for the presence of hazardous waste (DuPont et al. 1992).  See Table 2.2.4-1, which lists contaminants by group in both soil/waste and groundwater.  In addition to hazardous waste sites, another potential significant source of groundwater contamination is the numerous fuel service stations that are located throughout the study area.       Hazardous Waste Sites

Historically, the most significant hazardous waste sites in terms of their potential to affect groundwater and subsequently the Niagara River include: the Buffalo Avenue Plant (Occidental Chemical), the Necco Park and BFI/CECOS landfills (DuPont Chemical), and the Hyde Park landfill (Occidental Chemical) (USEPA and NYSDEC 2000).  Of these sites, only the Hyde Park landfill is located outside the zone of influence of the underground conduit system as defined by the groundwater flow divide located west of the conduits (see Study Area description in DuPont et al. 1992).  Farther north on the northern side of the forebay, remediation efforts for soil and groundwater contamination continue at the former Stauffer Chemical Plant site.  Figure 2.2.4-1 shows the locations of these hazardous waste sites within the investigation area.  It should be mentioned that, since the 1980s, significant reductions in area wide contamination have been achieved through the implementation of groundwater remediation programs at area hazardous waste sites.       Groundwater Contamination and Migration

Local groundwater studies have shown that, where contaminant migration has occurred, the direction of transport is generally consistent with the known pattern of groundwater flow (Figure 2.2.4-2).  Yager (1996) analyzed the movement of groundwater near hazardous waste sites in the Niagara Falls area as part of a comprehensive groundwater model developed for the Lockport bedrock.  The direction of horizontal flow in the upper bedrock zone at these sites was found to be mainly toward the FST and NYPA conduits, or the Niagara River (Figure 2.2.4-2).  Vertical flow was also found to occur from the upper Lockport to the underlying fracture zones through vertical fractures.

Flow in the underlying fracture zones is generally parallel to the direction of flow in the upper zone, except near the southern end of the underground conduits.  Flow near the southern end of the conduits is captured by production wells at the Olin Plant site.  Toward the north, at the former Stauffer Chemical site, flow is directed toward the forebay canal.  Investigations in the area of the Necco Park and BFI/CECOS landfills have further determined that groundwater in the upper Lockport tends to flow southward toward the FST, whereas groundwater in the underlying fracture zones is directed westward (and possibly northward) toward the conduit drains.  Groundwater that collects in the conduit drain system discharges either toward the FST or to the forebay.

Groundwater contaminant plumes in the area of the Niagara Power Project are most apparent where several sources of contamination exist close to one another.  These include industrialized areas along the river and plumes associated with the Necco Park and BFI/CECOS landfills (Figure 2.2.4-2).  Plumes of many contaminants (chlorinated and non-chlorinated volatiles and semivolatiles, PCBs, pesticides, and metals) were identified in the upper Lockport along the river section west of the underground conduits, as well as in the vicinity of the Necco Park Landfill (DuPont et al. 1992).  The contaminant plumes west of the conduits, in the vicinity of the DuPont and Olin Buffalo Avenue Plant sites, are centered on Gill Creek south of Buffalo Avenue in the vicinity of probable source areas.  Lower concentration trends suggest contaminant migration toward the FST to the north, the conduits to the east, and the Niagara River to the south.  Contaminant plumes centered upon the Necco Park Landfill also suggest contaminant migration toward the John Avenue and Falls Street Tunnels to the south, and to the conduits to the southwest.

East of the conduits, three sites are associated with contaminant plumes (DuPont et al. 1992).  Plumes of volatile organics, including chlorinated solvents, fuel-related benzene, toluene, ethylbenzene, xylene (BTEX) compounds, and chlorinated benzenes and toluenes, were observed to originate at the Buffalo Avenue (Occidental Chemical) and Royal Avenue (Frontier Chemical) sites (Figure 2.2.4-2).  It was observed that the highest levels of these contaminants were confined primarily to the immediate vicinity of these plant sites, with lower levels detected in the vicinity of the FST and the conduits.  Contaminant plumes containing chlorinated semivolatile compounds (e.g., hexachloroethane, hexachlorobutadiene, hexachlorocyclopentadiene, and octachlorocyclopentene) and pesticides/PCBs were also identified in the area of Occidental Chemical’s S-Area landfill (Buffalo Avenue site), with lower concentration trends directed toward the Niagara River and main plant site (Figure 2.2.4-2).  Localized concentrations of BTEX, chlorinated benzene and toluene, and phenols/methylphenols, were also identified in the upper Lockport at Occidental Chemical’s Durez plant, located west of the BFI/CECOS landfill.  As with the other sites described above, the Durez site is located within the dewatering influence of the conduit drains.  It has been determined that groundwater flow at the Durez plant is to the southwest.

Of the approximately 12 Niagara Falls-area sites with elevated levels of contaminants in the bedrock aquifer, two sites, namely, the Hyde Park landfill and the Stauffer Chemical site, are located beyond the influence of the underground conduit drain system (see Study Area description in DuPont et al. 1992).  Migration of contaminants at these sites is primarily toward the Niagara gorge, with some probable migration in the lower Lockport toward the forebay, particularly from the Stauffer Chemical site.  Contaminant plumes at the Hyde Park landfill (i.e., phenols and methylphenols, chlorophenols, chlorinated benzene and toluene, chlorinated semivolatiles, and pesticides/PCBs), and the Stauffer Chemical site (i.e., chlorinated volatiles, chlorobenzene, and BTEX compounds) are shown in Figure 2.2.4-2.  As with the other sites discussed, higher concentrations of chemicals were reported on site near probable source areas, with levels decreasing with distance from the site.  It was found that groundwater contamination at the Stauffer site was confined to the western portion of the former plant site and adjacent Power Authority property, with vertical migration of volatile organics observed farther west, nearer to the Niagara gorge (CRA 1991).

Downward migration of contaminants in groundwater below the Lockport is restricted by the relatively impermeable Rochester Formation.  Contaminants detected in the existing NYPA conduit observation wells (installed at the time of NPP construction), which monitor the entire Lockport Group, have included chlorinated volatile and semivolatile compounds, with maximum concentrations of less than 1.1 mg/L (Dupont et al. 1992).       Remediation Systems and Flow Barriers

Groundwater remediation programs have been initiated within the Lockport Group at all industrial sites where elevated levels of chemicals have been detected.  The locations of area recovery wells for contaminant plume control and remediation are shown in Figure 2.2.4-3.

Most chemicals identified within groundwater plumes mentioned in Section are now largely controlled by the respective ongoing site remediation programs.  These programs will, in effect, control future chemical migration into and through the Lockport Group.  The remaining comparatively small mass of chemicals present in the bedrock aquifer beyond the influence of these remediation systems is expected to become attenuated, or potentially to reach the Niagara River.  The primary migration pathway to the Niagara River is through the underground conduit drains and the FST, which, apart from the river itself, are the main collectors of Lockport groundwater.  All dry-weather flow in the FST is now treated at the Niagara Falls Wastewater Treatment Plant (WWTP) prior to being discharged to the river below the Falls.  As a result of these combined remedial measures, USEPA estimates that discharge of potentially toxic chemicals to the Niagara River has been reduced by about 80%.  The actual reduction is likely greater since this estimate was based on remedial actions occurring only at the priority waste sites and does not take into account the recent completion of remedial systems.  While the FST is operating as an effective collector of upper Lockport groundwater, an unknown amount of flow and associated chemicals is transmitted downward into the Lockport Group (through bedrock fractures) or is discharged to the conduit drainage system and forebay.       Sites of Interest

Stauffer Chemical Plant

      The former Stauffer Chemical Company plant is located immediately north of the Robert Moses Power Plant Forebay (Figure 2.2.4-1).  Between 1900 and 1930, portions of the site were owned by the Titanium Alloy Manufacturing Company, the American Magnesium Corporation and Niagara Smelting Company.  From 1930-1980, the site was owned and operated by Stauffer Chemical Company.  Between 1930 and 1976, the plant produced carbon tetrachloride and various metal chlorides.  The primary chemical feedstock was carbon disulfide, which reacted with chlorine to produce carbon tetrachloride and sulfur chlorides.  Parachlorothiophenol was produced from chlorobenzene and sulfur chlorides.  Bulk methylene chloride and tetrachloroethylene were also brought to the plant and repackaged.  In addition, two areas east of the plant were used as landfills for the disposal of “inert materials”.  The plant ceased operation in 1976 and building structures were demolished in 1980. 

      Several investigations to determine whether previous site operations have impacted soil and groundwater have been conducted. Results of these investigations determined that several VOCs had impacted soil and were migrating to the groundwater table.  A Site Specific Parameter List (SSPL) was developed and includes:

·        Carbon Disulfide

·        Carbon Tetrachloride

·        Chloroform

·        Methylene Chloride

·        Tetrachloroethylene

·        Trichloroethylene

·        Benzene

·        Toluene

According to a Final Site Investigation Report, prepared by Contestoga-Rovers and Associates (CRA) (April 1991), several hydrogeologic units were identified at the site.  Overburden units were divided into fill deposits and glacial/lacustrine clays.  Three bedrock water-bearing units were identified within the Lockport Group.  The Upper Lockport Water-bearing Zone (UWBZ) consists of the overburden and top 25 feet of the Lockport Formation.  The Upper Aquitard separates the UWBZ and lower water-bearing zone (LWBZ).  The Lower Lockport WBZ, which ranges from 50-75 feet below the top of rock.  Beneath the LWBZ and above the Rochester Formation are undifferentiated water-bearing zones and aquitards in the Lockport Formation.  The Lockport/Rochester WBZ located at the base of the Lockport at or near the Rochester Shale contact. 

In 1995, based on the results of site investigations, three soil vapor extraction (SVE) systems were constructed to address soil contamination.  The major “areas” of contamination where identified as Area A (located in the central portion of the site); Area C (the former landfill area), and Area T-4 (located southwest of Area A). 

Area A is the largest, and contains 32 SVE wells, and three dual-phase groundwater/SVE wells.  The area is also covered with a PVC geomembrane liner cover and geotextile cushion.  Significant groundwater present in the Area A SVE system (suspected to be originating from preferential pathways created by underground utilities and structural bedding materials) led to the installation of a clay barrier trench.  The trench (installed in August/September 2000) was excavated to one foot into native soil and ranged in depths from 3.5 – 7.5 feet below grade.  A drainage pipe was installed around the edge of the liner.  No information was noted in reports, as to where this drainage pipe discharges.  A 2000 Annual Operations and Maintenance Report (CRA, 2000) notes that water levels measured on October 24, 2000, indicated 21 out of the 32 SVE wells were dry.  This was the highest number ever observed to date.

Area C is located in the former landfill area that lies within NYPA property immediately east of the Stauffer Plant property.  The original SVE system consisted of three SVE wells, a cover comprised of a PVC geomembrane liner, and geotextile cushion.  In November 2000, confirmatory soil samples were collected from the area.  Analytical results indicated while the majority of the area met soil cleanup objectives of less than 10 ppm total VOCs, an isolated pocket existed within the central portion of the area.  Based upon these results, two additional SVE wells were installed.  According to the 2002 Annual Operations and Maintenance Report (CRA 2003) it appears that only the two newly installed SVE wells now comprise the SVE system.  No additional information is available as to the nature of the old SVE wells.  Additionally, the 2000, 2002, and 2003 Annual Operations and Maintenance Reports indicate an Area B (a second former Stauffer landfill area also located on NYPA property immediately east of the former Stauffer Plant).  However, after review of available materials, no information regarding this area was available. 

The T-4 Area is located southwest of Area A.   The remediation system consists of three SVE wells and one dual-phase groundwater/SVE well.  Following confirmatory soil sampling in November 2000, which indicated soil sample results below the cleanup criteria of 10 ppm, the SVE system was decommissioned in September 2001.  A shallow groundwater extraction well was shut down in mid-2002. 

A DNAPL Recovery System consisting of a wooden equipment shed and a concrete secondary containment pad are around well OW-03.  Prior to the end of 2001, this well was manually pumped and inspected for the presence of product.  At the end of 2001 the well was converted to a permanent extraction well.  No information was available as to the methodology/reasoning for the DNAPL recovery system at this location (i.e., when free-phase product was discovered in this well, how much was recovered).

In order to address identified groundwater contamination, a groundwater extraction well network was implemented.  As of the 2003 Annual Operations and Maintenance Report, there were four deep bedrock groundwater wells, three shallow bedrock extraction wells, plus the shallow well in T-4 area and the three dual-phase wells in Area A.  Groundwater from each well is pumped to an on-site treatment system.  Treated water is discharged through an outfall to NYPA Forebay. 

The 2000 and 2003 Annual Operations and Maintenance Reports (CRA) concentrate on four of the sites WBZs (Upper Lockport WBZ, Lower Lockport WBZ, Lockport/Rochester WBZ and the Rochester WBZ).  For purposes of this investigation, only the first three will be discussed.   Groundwater level potentiometric contours are presented in both reports.  The 2000 report also presents pre-pumping and pumping groundwater contours.  In addition, chemical isocontours were completed for carbon disulfide and carbon tetrachloride and chloroform combined.  Review of the data indicates:

·        Upper Lockport WBZ exhibits little to some local response to pumping.  Pre-pumping conditions indicate groundwater flow from east to west toward the gorge.  Elevated concentrations of all three compounds (carbon disulfide, carbon tetrachloride, and chloroform) within the dual SVE/Extraction wells located within Area A. Chemical isocontours presented in the 2003 Operations and Maintenance Report indicate effective containment of carbon disulfide, and partial containment of chloroform and carbon tetrachloride.

·        Lower Lockport WBZ exhibits high response to pumping.  Pre-pumping groundwater flow is from north to south toward the Forebay.  During pumping, flow is toward the interior of the site.  Elevated levels of carbon disulfide where detected in two wells either within or near the capture zone in Area A.  The 2003 Operations and Maintenance Report indicates that contaminants of the Lower Lockport WBZ are being contained, captured on site, and recovered by the groundwater extraction system.

·        Lockport/Rochester WBZ also exhibits a high response to pumping.  Pre-pumping conditions indicate flow from north to south/southwest toward the Forebay.  Pumping conditions exhibit an inward flow toward extraction wells.  The 2003 Operations and Maintenance Report indicates that contaminants of the Lockport/Rochester WBZ are being contained, captured on site, and recovered by the groundwater extraction system.

Hyde Park Landfill Site

Hyde Park landfill is a 15-acre National Priority List Category 1 site. Approximately 80, 000 tons of waste were disposed there from 1953 until 1975.  The types of waste disposed of at the landfill and chemicals found in the sites soil and groundwater during previous investigations are summarized in Table (Hooker-Hyde Park Landfill).   The clean-up remedies at the site include:

·        A source control extraction system of wells to remove non-aqueous phase liquids (NAPL) from the overburden in the landfill.

·        An overburden drain system surrounding the landfill. It consists of an overburden collection trench that extends around the north, west, and south sides of the Site, and is located within the limits of the overburden aqueous phase liquid (APL) plume.

·        A bedrock remedial system consisting of 17 recovery wells to control a NAPL plume and a 2 well APL recovery well system. This system has been operational since 1994. New recovery wells were added since 1994 to gain complete control of the plume. This is an on-going effort.

·        A landfill cap was installed in 1978 to prevent infiltration. The cap was modernized with a composite clay and plastic liner and an overlying geocomposite drainage layer in 1994.

·        APL (chlorinated acids, hexachloropentadiene, chlorinated toluenes, benzenes, phenolics, and total organic halides) extracted via groundwater extraction pumps and is carbon-treated onsite. About 4,000,000-7,000,000 gallons of groundwater per quarter are treated.

·        Approximately 325,00 gallons of NAPL have been incinerated as of April 2002. NAPL was incinerated at the Occidental Buffalo Avenue plant but is now shipped to Texas for incineration.

·        A network of approximately 170 groundwater monitoring wells is located on and around the site.

·        Restricted access to seeps along the Niagara River gorge face; diversion of seep water and removal of contaminated sediments from the seep face to prevent human exposure.

·        Excavation and removal of 29,200 cubic yards of leachate contaminated sediments from Bloody Run and the collection of leachate at the landfill.

Groundwater recovery wells placed between the landfill and the Niagara River currently prevent groundwater contamination from reaching the Niagara River.  Samples taken from gorge face seeps by the potential responsible party (PRP) have indicated that the seep water currently requires no additional control or remediation, and is similar in composition to groundwater, which is recharged near the gorge.

Groundwater flow modeling (S.S. Papadopulos, 2001) defined 3 bedrock water-bearing zones within the Lockport Group dolomite: upper (Guelph and Eramosa Units A-F, middle (Goat Island), and lower (Gasport and DeCew).  Modeling results indicate that Groundwater flows radially outward from the site:

·        west toward the Niagara Gorge where it discharges in seeps,

·        east toward the NYPA conduits (about 1 mile east)

·        north toward the forebay (about 1 mile north)

·        and toward the south

Modeling results also indicate there is significant downward flow from the upper water-bearing zone to middle and lower water-bearing zones.

Groundwater modeling was used to evaluate the performance of the remedial pumping system for the period between March 1999 and March 2000.  Results of this analysis indicated:

·        in the Upper bedrock zone, only the southern portion of groundwater within the NAPL plume boundary is captured

·        in the Middle bedrock zone, the major portion of groundwater within the NAPL plume boundary is captured

·        in the Lower bedrock zone, the groundwater within the NAPL plume boundary is completely captured

Stauffer Chemical Whittaker Subdivision Site

The Stauffer Chemical Whittaker Subdivision Site was used for disposal of approximately 50,000 to 75,000 cubic yards of waste associated with chemical and industrial plant operations. Disposal activities occurred from approximately 1930 to 1952. The site consisted of an excavated canal, measuring approximately 100 feet wide by 2,000 feet long, located in the Town of Lewiston immediately west of the Tuscarora Nation boundary in the vicinity of Upper Mountain Rd (Figure 2.2.4-1).

According to NYSDEC records, wastes disposed of in the canal included asbestos, concrete cell parts, reactor linings, scrap sulfur, graphite, scrap metal, silicon, zirconium and titanium oxides, flux, cinders, and phenol. Site investigation activities occurred from approximately 1979 to 1993. Contaminants identified in soil and groundwater included PCBs, numerous SVOCs, phenol, sulfur, fluorides, chlorides, mercury, arsenic, trichloroethylene, antimony, chromium, and lead. One area near Upper Mountain Road included elevated levels of PCBs (26 ppm) and mercury (132 ppm) in soil. Groundwater samples in this area contained detectable concentrations of TCE, antimony, chromium, and lead.

The site was delisted from the NYSDEC Inactive Hazardous Waste Sites Registry in 1995.       Fuel Service Stations/Petroleum USTs

Another significant source of groundwater contamination is leaking fuel underground storage tanks (USTs).  Underground releases of gasoline or diesel from leaking fuel USTs can create plumes of non-aqueous phase liquids (namely, gasoline or diesel product accumulation in the subsurface) or plumes of constituent chemical contaminants dissolved in and migrating with groundwater.  Constituent chemicals of gasoline are typically much more soluble and volatile that those found in diesel fuel, and therefore more likely to dissolve into groundwater.  Gasoline-related contaminants typically identified in groundwater include benzene, toluene, ethylbenzene, xylenes, MTBE, cyclohexanes, and numerous other common components.  Fuel service stations are ubiquitous throughout the study area in both the urban and industrialized areas such as City of Niagara Falls, and the more rural areas such as the Tuscarora Nation.       Lewiston Reservoir Sediment Quality

A study was conducted in 2002 to determine the quality of sediment in the Lewiston Reservoir.  The findings were presented in a report entitled Extent of Sedimentation and Quality of Sediment in the Lewiston Reservoir and Forebay (ESI 2005).  Five reservoir sediment samples were collected and analyzed for 18 priority toxic pollutants identified in the Niagara River Toxics Management Plan (NRTMP) and five additional parameters identified by NYSDEC. 

A summary of analytes detected in sediment samples is presented in Table 2.2.4-2.  Results indicated several PAHs and PCBs detected in reservoir sediments.  Total PAH concentrations ranged from 5,500 micrograms per kilogram (ug/kg) to 9,500 ug/kg and total PCB concentrations ranged from 44 ug/kg to 190 ug/kg.  Several metals (arsenic, lead and mercury) were detected in all five samples.  One pesticide compound (mirex) was detected in one sample at a concentration of 53 ug/kg. 

The report concluded based on pH and sediment organic carbon levels that lead and mercury would not be expected to partition into the overlying water column.  Due to the presence of clays in the benthic substrate, arsenic would be unlikely to leach into surface or groundwater.  PAHs and PCBs are expected to remain in sediments due to their low solubility and affinity to sorb to organic carbon in sediments.  Similarly, mirex adsorbs to sediment particles and has a very low solubility (ESI 2005). 

Table 2.1.3-1

Bedrock Stratigraphy of the Niagara Falls Area





Average Thickness (feet)





Vernon Shale

57 (in study area)

Green and red shale



Guelph Dolomite


Brownish-gray to dark gray, fine to medium, thick-bedded dolomite, with some argillaceous dolomicrite, particularly near contact with the Vernon Shale

Eramosa Dolomite


Brownish-gray, biostromal, bituminous, medium- to massive-bedded dolomite, with some argillaceous dolomicrite

Goat Island Dolomite


Light olive-gray to brownish gray, fine to medium crystalline, thick- to massive-bedded saccharoidal, cherty dolomite, with argillaceous dolomicrite near top of formation

Gasport Limestone


Basal unit is dolomitic, crinoidal grainstone, overlain by argillaceous limestone


DeCew Dolomite


Very fine crystalline dolomite, medium to dark gray, thin to medium bedded

Rochester Shale


Dark-gray calcareous shale weathering to light gray to olive

Irondequoit Limestone


Light-gray to pinkish-white coarse-grained limestone

Reynales Limestone


White to yellowish-gray shaly limestone and dolomite

Neahga Shale


Greenish-gray soft fissile shale


Thorold Sandstone


Greenish-gray shaly sandstone

Grimsby Sandstone


Reddish-brown to greenish-gray cross-bedded sandstone interbedded with red to greenish-gray shale

Power Glen Shale


Gray to greenish-gray shale interbedded with light gray sandstone

Whirlpool Sandstone


White, quartzitic sandstone




Queenston Shale


Brick-red sandy to argillaceous shale


¹Designated Albion Group by the U.S. Geological Survey

Source: Yager 1996.

Table 2.2.4-1

Hazardous Waste Sites in the Investigation Area




Waste Description

Contaminants Identified in Soil/Waste 1

Contaminants Identified in Groundwater 1

Basic Carbon


Former disposal or fill area/ Manufacture of carbon product

Coal tar pitch, carbon, graphite

7, 10


Niagara Recycling (BFI/CECOS)



Municipal and industrial waste, chlorinated hydrocarbons



Airco Speer Carbon-Graphite


Landfill/ Operating industrial facility/Manufacture of furnace electrodes

Carbonaceous furnace insulation, asbestos, spent refractory materials



Carborundum Company-Globar


Incineration/ Operating industrial facility/Manufacture of heating elements and electronic components from silicon carbide

Spent halogenated and non-halogenated solvents



City of NF, Buffalo Avenue


Former disposal or filled area/Filled wetland

Incinerator ash, characteristic lead waste


 6,10, 11

TABLE 2.2.4-1 (cont.)

Hazardous Waste Sites in the Investigation Area




Waste Description

Contaminants Identified in Soil/Waste 1

Contaminants Identified in Groundwater 1

Necco Park


Closed Landfill/process waste from Dupont Niagara plant

Brine sludges, barium salts, chlorinated compounds, toluene, methanol and acetone

DNAPL, 1,2,8


DuPont Plant Site


Dump/ Operating industrial facility/Organic chemical manufacturer

Metal cyanide sludges, chlorinated volatiles and semivolatile compounds, pesticides



Forest Glenn Subdivision


Former disposal or filled area/Filled marsh

Phenol, formaldehyde, PVC resins, graphite, carbon



Frontier Chemical- Royal Avenue


Closed Lagoons/ Operating industrial facility/Manufacture of caustic chlorine for mercury cell

Caustic sludges, solvent chemicals (chlorobenzenes, PCE, TCE, benzoyl chloride)


1,2, DNAPL

Hooker Main Plant


Operating industrial facility/Landfill / organic chem. Manufacture

Inorganic and organic sludges (brine, sulfate, phosphate, oxalic acid)


NAPL 1,2,6,8,10

TABLE 2.2.4-1 (cont.)

Hazardous Waste Sites in the Investigation Area




Waste Description

Contaminants Identified in Soil/Waste 1

Contaminants Identified in Groundwater 1

Occidental- Durez Engineered Materials


Operating industrial facility/Manufacture of phenol formaldehyde resins

Phenolic waste



Hooker - Hyde Park Landfill


Closed Landfill

Brine sludges, organic phosphate, metallic and acid chlorides, phenolic tars, chlorination products


 NAPL, 4,5,6,8,10

Hooker - S Area



Organic phosphates, acid chlorides, phenol tars, benzoyl chloride, metal chlorides, chlorinated organics



Olin Corporation- Parking Lot/Plant Site


Operating industrial facility/Dump/Pond

Mercury cell brine sludge, organics (TCP,BHC)



Olin Corporation-Disposal Well


Abandoned water supply well

End liquor (60-65% water, 30% sulfuric acid, 5-10% sodium chlorite)



TABLE 2.2.4-1 (cont.)

Hazardous Waste Sites in the Investigation Area




Waste Description

Contaminants Identified in Soil/Waste 1

Contaminants Identified in Groundwater 1

Olin Corporation -Industrial Welding Site


Former disposal or fill area/ Former pilot research lab and process plant

Brine sludge (w/mercury), waste transformer oil, industrial scrap and fly ash, demolition rubble



Vanadium Corporation of America


Closed Landfill

Ferro chromium alloy dust (K090, D002) and slag, calcium hydroxide, ferro manganese slag



Solvent Chemical Plant


Former industrial facility/Dump/ Manufacture of dichloro, trichlor and tetrachlorobenzene

Lead, zinc, benzene and chlorinated benzenes



Stauffer Chemical


Former disposal or fill area/ Production of CCl4, metal chloride and parachlorothiophenol

Asbestos, scrap metals, sulfur, coke, zirconium and titanium oxides, misc organics (PCE, methylene chloride)



Stauffer Chemical Whittaker Subdivision

Delisted, 1995

Former disposal or fill area

Asbestos, scrap metals, sulfur, coke, zirconium and titanium oxides, phenols




TABLE 2.2.4-1 (cont.)

Hazardous Waste Sites in the Investigation Area




Waste Description

Contaminants Identified in Soil/Waste 1

Contaminants Identified in Groundwater 1

TAM Ceramics, Inc.


Operating industrial facility/Dump /Manufacture of refractory products

Metallic ore residues, barium



Union Carbide Corp. – Carbon Products Division


Operating industrial facility/Dump /Carbon products

General rubble, TCE degreasing sludges, Halowax (D003), chlorinated benzenes



Witmer Road Site


Former disposal or fill area/ open burning by City of NF

Corrosive waste, air pollution control waste




+ - Only iron, magnesium, manganese and sodium exceeded NYSDEC Class GA WQS.

* - Groundwater contaminants associated with adjacent Hyde Park Landfill

** - Groundwater contaminants associated with upgradient DuPont and Occidental Chemical Plant Site

NA - Site information is not available.

1- Chemical designation for contaminants identified

Group 1 - Chlorinated Volatiles                           Group 7 -PAHs

Group 2- BTEX                                                  Group 8 - Chlorinated Semivolatiles

Group 3 - Nonchlorinated Volatiles                     Group 9 - Phthalates

Group 4 - Phenol and Methylphenols                  Group 10 - Pesticides/PCBs/Dioxins

Group 5 - Chlorophenols                                     Group 11 - Heavy Metals

Group 6 - Chlorobenzenes and Chlorotoluenes


DuPont, 1992 and NYSDEC, 2003


Table 2.2.4-2

Analytes Detected in Lewiston Reservoir Sediment




RES-SED05     Reservoir

RES-SED06     Reservoir

RES-SED07     Reservoir

RES-SED08     Reservoir

RES-SED09     Reservoir

RES-SED12     (DUP OF SED-08)  Reservoir

Date Sampled






















360  J

280  J

450  J

260  J

340  J





570  J

490  J

840  J

560  J

640  J





670  J

670  J

1100  J

750  J

840  J





540  J

380  J

690  J

480  J

520  J





640  J

620  J

1100  J

640  J

810  J


Total PAHs


7400  J

6100  J

5500  J

9500  J

6000  J

7200  J













53  J







Aroclor 1242



130  J

79  J


57  J

99  J


Aroclor 1248





44  J




Aroclor 1260


43  J

37  J

28  J





Total PCBs


190  J

170  J

110  J

44  J

57  J

99  J





5.0  J

5.4 J

5.9  J

14.5  J

8.7  J

9.1  J













36.8  J

32.6  J

31.6  J

72.4  J

46.6  J

45.6  J





0.169  J

0.171  J

0.175  J

0.173  J

0.163  J











Total Organic Carbon


15100  J

13400  J

16800 J

15300  J

16600  J

16200  J


Total Volatile Solids








Source: ESI 2005


VOC – Volatile Organic Compound

PAHs – Polynuclear Aromatic Hydrocarbons

PCBs – Polychlorinated Biphenyls

mg/kg – micrograms per kilogram

mg/l – micrograms per liter

mg/kg – milligrams per kilogram

mg/l – milligrams per liter

ng/l – nanograms per liter

pg/g – picograms per gram

%W/W – Percent Weight by Weight

J – Estimated Value

Figure 2.1.1-1

Natural Elements within Project Area

[NIP – General Location Maps]

Figure 2.1.3-1

Bedrock Formations in the Niagara Falls Area as Exposed at the Horseshoe Falls

Drawing modified after Gilbert, 1895 (Figure 5 in Johnston 1964)

Figure 2.1.3-2

Descriptive Stratigraphic Column for the Lockport Group

Source: Brett et al. 1995.  Composite section for Niagara County is shown.

Figure 2.1.3-3

Diagrammatic Stratigraphic Relations in the Lockport Group, with Regional Correlations between Clappison’s Corners, Ontario, and Penfield, New York

Reference Datum is the contact between the Gasport dolomite and the Goat Island dolomite (Brett 1995)

Figure 2.1.3-4

Generalized Stratigraphy and Groundwater Flow in the Vicinity of Niagara Falls

Figure 2.1.3-5

High-Transmissivity Zone and FST/Conduit Crossing

[NIP – General Location Maps]

Figure 2.1.3-6

Relations among Stratigraphy, Regional Water-Bearing Zones, and Model Layers

Source:  Brett et al. 1995, Tepper et al. 1991, Yager 1996.  Use of proposed nomenclature for the Lockport Group does not constitute formal acceptance by the U.S. Geological Survey.

Figure 2.2.1-1

Conduit Cross-Section

[NIP – General Location Maps]

Figure 2.2.1-2

Conduit Longitudinal Section

[NIP – General Location Maps]

Figure 2.2.1-3

Approximate Locations of observed groundwater inflow

[NIP – General Location Maps]

Figure 2.2.1-4

Conduit Drainage System

[NIP – General Location Maps]

Figure 2.2.1-5

Fluctuation in Surface Water Elevation, Lewiston Reservoir and Forebay, Datalogger Data, January - December 2003

Figure 2.2.4-1

NYSDEC-Listed Hazardous Waste Sites

[NIP – General Location Maps]

Figure 2.2.4-2

Groundwater Contaminant Plumes

[NIP – General Location Maps]

Figure 2.2.4-3

Recovery Wells

[NIP – General Location Maps]


A preliminary groundwater study was completed in 2002 as part of the FERC relicensing of the Niagara Power Project (URS et al. 2003).  The study included environmental and engineering investigations to identify the potential effect of Project operations and features on groundwater flow in the Project vicinity.

The 2002 investigation relied on existing wells and gauges for the collection of groundwater level data.  No new wells or piezometers were installed as part of the 2002 investigation.  Data were collected from a variety of sources that included: permanent staff gauges in the Niagara River, abandoned residential wells within Tuscarora Nation Lands, NYPA and government agency owned monitoring wells, and privately owned industrial wells.  An attempt was made to correlate changes in surface water levels attributable to Project operations with changes in groundwater elevation.  An attempt was also made to determine the effect of Project structures such as the forebay, Lewiston Reservoir, and the conduits on the groundwater regime.  As part of this overall effort, observations were made within conduit Pump Station A.

The 2002 investigation concluded that the NPP affects groundwater flow and water levels in the vicinity of the Project. The primary effects identified as part of that study are:

·        Sinusoidal fluctuations in water levels in direct response to forebay and reservoir water level fluctuations.  The observed fluctuations were most pronounced and direct within the CDS and less pronounced in wells in the vicinity of the reservoir.

·        Overall altered potentiometric elevations as a result of project features.  Elevated groundwater levels observed near the reservoir likely resulted from increased recharge to the groundwater system from the Lewiston Reservoir.  Observed levels indicate depressed water levels in the vicinity of the conduits likely as a result of the CDS acting as a regional groundwater flow sink.

Observations made in the 2002 study were compared with observations made in previous studies and were found to be generally consistent with the observations of Johnston 1964, Miller and Kappel 1987, Dupont et al. 1992 and Yager 1996.  One observation made during the 2002 investigation is that Niagara River water appears (at least sometimes) to be exiting the conduits through the CDS in a reverse pattern from system design.

Some other important observations resulting from the 2002 study were:

·        Forebay water level fluctuations have a direct impact on groundwater levels in the vicinity of the conduits for much of the entire length of conduits.  The response time for groundwater level fluctuations in these wells was found to be almost instantaneous with forebay level fluctuations.  One result of this effect on groundwater levels along the conduit is fluctuating hydraulic gradients along the conduit.  When forebay water levels are low, the gradient is typically northward toward the forebay, and when forebay levels are high, the gradient is typically flatter with occasional flow reversals toward the river.

·        Forebay water level fluctuations also directly influence hydraulic heads in water-bearing zones that intersect the CDS.  Forebay-level-induced fluctuations in groundwater levels were observed up to approximately 3,000 feet from the conduits.

·        Groundwater levels along the upper Niagara River exhibit water level fluctuations in response to river water level fluctuations.  For wells located adjacent to the upper Niagara River, but outside of the direct influence of the CDS, groundwater level fluctuations of up to approximately 1.2 feet were observed to correspond with river level fluctuations of approximately 1.5 feet.  Wells located further north (approximately 2,400 feet from the upper Niagara River) show no effect of changes in river water levels.

While it was noted that Project operations do appear to induce sinusoidal fluctuations in groundwater levels, and that Project features do appear to alter the flow path of groundwater near Project features, these effects appear to diminish with distance from the Project, as evidenced by the response of industrial-owned and Tuscarora Nation wells.

·        Due to the limitation in the number of available measuring points and the limited time frame for the 2002 study, further study was recommended to quantify the magnitude and extent of Project operational and feature effects on groundwater flow within the Project area.  The results of the 2002 study helped to form the basis of the current study.



The scope of work developed for this investigation included several tasks designed to evaluate Project effects.  Tasks completed included: (1) conducting groundwater flow modeling to evaluate NPP effects on groundwater flow, (2) installation of nested groundwater monitoring wells, (3) implementation of a groundwater level monitoring program, (4) investigation of groundwater infiltration into the FST at the point where it crosses over the NYPA conduits, and (5) a water quality sampling program, including collection of surface water and groundwater samples.  Each of these tasks is described in detail in the following subsections.

4.1         Groundwater Flow Model

To investigate issues related to the groundwater flow regime in the area of the NPP, numerical modeling of groundwater flow was performed.  URS acquired an existing regional groundwater flow model developed by the United States Geologic Survey (USGS) and encompassing the Niagara Power Project, the City of Niagara Falls, and the surrounding area.  URS converted this model to a Groundwater Modeling System (GMS ) graphical pre- and post-processing environment. The model was then used to evaluate overall groundwater flow patterns in the vicinity of the NPP and to create a focused model of the area surrounding theLewiston Reservoir.

4.1.1        Development and Modification of the USGS Regional Model

In 1996 USGS developed a three-dimensional flow model to estimate rates and directions of the groundwater flow in the Lockport Group, the fractured bedrock aquifer underlying the study area (Yager 1996).  The USGS MODFLOW code was utilized.  The 1996 model has been modified and recalibrated over the years, with the latest changes having been incorporated in 2003 (MODFLOW 2000 version).    The 10-layer model (Figure 2.1.3-6) represents an area of 110 square miles.  The model grid contains 71 rows and 69 columns, with a uniform cell size of 1,000 by 1,000 feet.

Water within the Lockport Group flows through the weathered bedrock surface and several underlying horizontal fracture zones at or near stratigraphic contacts.  The fracture zones are connected by high-angle fractures and by subcrop areas where they intersect the bedrock surface.  At the scale of the model, the fractured bedrock was assumed to act as a porous medium.  The fracture zones were represented by the model layers, and connections between the zones were represented by vertical leakage between the layers.

The Niagara escarpment and the Niagara River gorge form the natural hydrologic boundaries on the northern and western sides of the modeled area, respectively.  No-flow, defined flow, or constant-head boundaries for the southern and eastern sides were selected on the basis of potentiometric surface features indicated on maps of each layer.  The upper boundary was specified as constant-head in areas underlying the Niagara River and its tributaries, and as constant-flow representing the recharge to the weathered bedrock elsewhere.  The bottom boundary was specified as no-flow because an over pressured gas reservoir underlies the Lockport Group and prevents vertical flow from the modeled area.  Several manmade hydraulic structures, including the Lewiston Reservoir, tunnels excavated in bedrock, and an extensive drainage system surrounding the hydropower conduits, were also represented as boundaries.  A grout curtain installed around the Lewiston Reservoir was modeled as a zone of low transmissivity.  The model domain and boundary conditions are shown in Figure 4.1.1-1.  Figure 4.1.1-2 shows a generalized cross-section of the model layers and boundary conditions.

Yager (1996) compared results of the steady-state simulations with (1) the measured potentiometric surface of the weathered bedrock zone, (2) average heads measured by piezometers in fracture zones, (3) low-flow measurements of springs and streams, and (4) measurements of discharge from tunnels and excavations.  Yager used trial-and-error and nonlinear regression to estimate recharge, transmissivity of the weathered bedrock and fracture zones, and vertical hydraulic conductivity of the bedrock.  Nonlinear regression made possible the identification and estimation of values for model parameters to which the measured heads and flows were sensitive.  The model developed as a result of this process is referred to as the “calibrated model” in this report.

Parameters estimated during the calibration process are presented on Table 4.1.1-1.  Modeled distribution of hydraulic heads, as well as the comparison between heads observed in monitoring wells and computed by the calibrated model, are shown in Figures 4.1.1-3 and 4.1.1-4, for the weathered bedrock and Gasport dolomite, respectively.  Table 4.1.1-2 shows the comparison between the modeled and measured flow rates. Yager states that differences between observed and modeled values are likely due to local variations in transmissivity in bedrock that are not represented by the model.

Results of the USGS model indicated that (1) measured flow into the FST exceeds the amount that can be sustained by the aquifer; and that therefore a connection between the tunnel and the Niagara River may be assumed; (2) recharge within the urban parts of the modeled area is greater than in the rural parts, possibly because of losses from the municipal water supply or infiltration from unlined storm sewers that intersect the bedrock; and (3) within the lowlands near the Niagara River, widespread areas may exist in which groundwater flows upward and is discharged through evapotranspiration and surface drainage.  Conclusion (1) is particularly important in the context of this groundwater report.  The inability of the model to account for the large percentage of the flow measured in the Falls Street Tunnel implies that a connection exists between the Niagara River and the Tunnel, and that this connection may not be formed as an effect of the natural hydraulics of the aquifer.  The possible nature of the connection is discussed in Section 6.2.

Another meaningful finding of the model is the distribution of the areas contributing to particular discharge points (Figures 4.1.1-5 and 4.1.1-6).  Table 4.1.1-3 presents the water budget.  Major sources of inflow are recharge, as well as leakage from the Niagara River and the Lewiston Reservoir.  The main sinks are the Niagara River gorge, tributary creeks, and the FST where the two NYPA water conduits pass beneath the FST.

Alternative concepts of the aquifer system that closely reproduced the measured heads and flows used for calibration were identified through nonlinear regression.  The first included additional zones of recharge; the second included different model boundary conditions; and the third included horizontal anisotropy within model layers.  The calibrated model and the two best alternative models were used to estimate groundwater flow rates at 21 waste disposal sites.

URS obtained the latest 2003 model through personal communication with Richard Yager, the model’s author.  The main difference between the 1996 and 2003 models is the increase, by Yager, in the estimated recharge in the rural areas of the model domain of approximately 50%.  This created an increase in the total amount of water entering the system of approximately 6%.  Small changes were also made to some values of transmissivity and vertical leakance.  The resulting values of hydraulic heads are slightly different than those of the original 1996 model, although the flow patterns are preserved. 

4.1.2        Conversion of the USGS Model to the GMS Environment

To make the model useful for the present project, it was decided to convert the USGS model to a Groundwater Modeling System (GMS) environment.

GMS (version 4.0) is a modeling environment developed by the Environmental Modeling Research Laboratory of Brigham Young University, Utah.  GMS provides user interfaces to several groundwater flow and transport models, including MODFLOW.  It facilitates model development and the post-processing of model results.  It has extensive graphics capabilities, as well as interfacing easily with the software packages commonly used to process and present information, such as Geographic Information System (GIS) or ArcInfo.

Conversion of the MODFLOW 2000 model to the GMS modeling environment involved: (1) direct importation of non-calibrated model elements, (2) extraction of calibrated parameter array elements and importation into GMS, and (3) conversion verification by comparing flow budgets and hydraulic heads of observation wells.  Detailed discussion of conversion methods is presented below and in Appendix A.

The converted model served as the basis of the focused model that was later applied to perform a more detailed assessment of groundwater levels and flow near the Lewiston Reservoir.       Conversion Method

URS used GMS, version 4.0, as the platform for the modeling effort.  The file format of the calibrated USGS model, used to create the GMS version, is that of MODFLOW 2000.  Files containing information about model elements that were not the object of calibration are essentially the same as the corresponding files in the GMS format, and can be converted by the straightforward importation through a screen editor.  Files containing results of calibration were converted using the method described in detail in Appendix A.  This section presents a summary of the process.

Model elements subjected to calibration, namely, horizontal hydraulic conductivity (or transmissivity), vertical hydraulic conductivity (or vertical leakance) and recharge, take the form of arrays.  This is the result of the division of the modeled area into a finite number of cells.  For each model cell within a given layer, a single value of each parameter is required by the model, thus creating an array.  The MODFLOW 2000 calibration process creates the required parameter arrays by combining four objects: a group of parameter values, an array of multiplication factors, an array of zone indicators, and a uniform multiplication factor.  The four elements were extracted from the MODFLOW 2000 files and recombined to create arrays of calibrated parameters.  The arrays were then imported into GMS.

The 1996 model was prepared using a MODFLOW option of defining model conductances called “Block Centered Flow” package.  The 2003 model was prepared in terms of another option, namely, the “Layer Flow Package”.  During the conversion process, it was found that the GMS did not support certain elements of the Layer Flow Package.  The conversion was therefore accomplished by using the Block Centered Flow package (the same approach as was utilized in the original 1996 USGS model).

Several methods of solving finite-difference equations are at the basis of the model.  Both the USGS model and the converted GMS model use the method of a “Pre-Conditioned Conjugate Gradient Solver”.  During the simulation using the converted model, it was found that the model conversion could be improved by changing some of the parameters of the solver from the original values used in the USGS model.  The slight change in convergence behavior of the system likely resulted from the use of different flow packages (Layer Flow Package vs. Block Center Flow Package).       Verification of Conversion

To verify that the converted model (GMS model) was equivalent to the 2003 USGS model (MODFLOW 2000 model), both flow budgets and hydraulic heads at locations of observation wells were compared.       Flow Budget

Table 4.1.2-1 shows a comparison of flow budgets between the two models.  The discrepancies in elements of the flow budget are negligible, ranging between zero and 0.06%.  Most likely, they result from the GMS model using slightly different parameters of the Pre-Conditioned Conjugate Gradient Solver package.       Hydraulic Heads

The 2003 USGS model output contains the model-calculated hydraulic heads at 208 monitoring locations.  These can be compared to the heads at the same locations, as calculated by the GMS model.

The comparison is shown on Table 4.1.2-2.  The discrepancies range from zero to 0.7 feet, with most being on the order of 0.01 to 0.1 foot.  The reasons for the discrepancy may be related to the use of slightly different parameters of the solver package.  Moreover, locations of monitoring points were converted from the coordinate system used by the USGS model to the coordinate system used by URS for the NYPA project.  Some degree of error can be expected from the conversion process, resulting in a corresponding error in the hydraulic heads.

Hydraulic heads in the weathered bedrock and Gasport dolomite computed with the converted model are shown on Figures 4.1.2-1 and 4.1.2-2.  These may be compared with the hydraulic heads of the USGS model on Figures 4.1.1-3 and 4.1.1-4.

In addition to weathered bedrock and the Gasport dolomite, two other model layers occur in the area of the Lewiston Reservoir. They are the Eramosa dolomite and the Goat Island dolomite.  Hydraulic heads in these layers are shown in Figures 4.1.2-3 and 4.1.2-4, respectively.

4.1.3        Focused Modeling       Development of the Focused Model

A numerical groundwater flow model relies on the division of the model domain into a finite number of cells.  The size of the model is defined by the total number of model cells, not by the area included in the domain.  To keep the model both manageable and useful, a balance must be achieved between the accuracy (improved by decreasing the cell dimensions) and the model size (decreased by using large cell dimensions).

The regional model encompasses a large area and; therefore, it must rely on a relatively coarse discretization.  This discretization is not sufficient to investigate issues related to the flow patterns in areas whose size is similar to the model cell dimension.  For example, the distance between the groundwater divide that exists in the upper fractured bedrock east of the Lewiston Reservoir and the edge of the reservoir is approximately 1,500 feet.  The model cell dimension is 1,000 feet on a side, meaning that the entire area of the divide is contained within just a few model cells. 

The focused model is effectively a “cut-out” from the regional model.  Its domain is smaller than that of the regional model, allowing for use of a smaller cell dimension without increasing the model size.  The boundaries of the focused model are defined by the regional model.  The advantage of the focused model is that it can “zoom in” on a particular area, allowing for a more detailed analysis of local issues.

URS used this approach to create a local model of the reservoir vicinity, based on boundaries obtained from the regional model.       Model Domain and Grid

The focused model is intended as a tool for analyzing the groundwater flow regime in the vicinity of the reservoir.  The limits of the domain were chosen to coincide either with natural hydrogeologic features or with arbitrary lines where the hydraulic heads and groundwater flow rates were established previously within the framework of the regional model.

The domain of the focused model has been selected based on these criteria.  It is shown in Figure 4.1.3-1.  To the north, the domain is bounded by the Niagara escarpment.  The northwest limit is defined by the Niagara gorge.  To the east, the domain extends to a regional groundwater divide.  The escarpment, the gorge, and the divide are all natural features, also utilized as boundaries of the regional model.  Therefore, the northward, eastward and northwestern extent of the focused model is the same as that of the regional model.

To the south and southwest, the extent of the focused model domain was chosen to include natural features that may potentially affect flow patterns in the study area.  The southernmost such feature is the quarry.  A line immediately south of the quarry defines the southern extent of the focused model.  The location of this line was fine-tuned based on the definition of model layers established within the regional model.  Several of the regional model layers are found only in the southern part.  These layers were eliminated from the focused model by selecting the southern limit of the focused model domain to be north of the area where the layers “pinch out”.  The overall size of the focused model (i.e., the total number of cells) was therefore decreased.  To the southwest, the model extent is terminated immediately east of the NYPA conduits.  The conduits are a major sink in the region, creating an effective hydraulic separation of the Lockport aquifer into the east and west areas.  The domain of the focused model was divided into uniform cells with the dimensions of 250 by 250 feet, arranged into 125 rows and 250 columns.  Vertically, the domain was divided into seven layers, corresponding to Layers 1, 2, 6, 7, 8, 9 and 10 of the regional model [(Layers 3, 4 and 5 of the regional model do not exist within the domain of the focused model) (Figure 4.1.3-2)].  The total number of focused model cells was 218,750.       Boundary Conditions

The northern, northwestern, and eastern boundaries of the focused model domain reflect existing hydrogeologic features.  They are identical with the boundaries of the regional model.  The boundary conditions are therefore also the same (see Figure 4.1.1-1).  The southwestern and southern boundaries were chosen to include specific natural features that may affect flow patterns in the study area.  These boundaries were treated as specified head boundaries, with the values of hydraulic head equal to those determined by the regional model.       Layers

Layer 1 of the regional model represents the sinks and sources located within the overburden.  Flow in the overburden itself is not simulated.  Layer 2 corresponds to the layer of fractured bedrock, extending over the entire model domain.  Subsequent layers 3 through 10 simulate various fractured zones located horizontally within the relatively unfractured rock matrix of the underlying geologic units (Figure 2.1.3-6 and 4.1.3-2).  These geologic units dip to the south at a rate of approximately 30 feet per mile.  As a result, the geologic units, and associated fracture zones subcrop throughout the model domain, with deeper units surfacing progressively to the north.  To account for this, each model layer is terminated north of the outcrop area of the fracture zone simulated by this layer.  Only Layer 2, representing the fractured bedrock, is continuous throughout the model domain.

The focused model retains the same layering system.  Layers 3, 4 and 5 of the regional model, however, subcrop south of the southern boundary of the focused model.  As a result, they do not exist in the area contained within the domain of the focused model.  Therefore, the focused model is constructed of 7 layers.  The correspondence between the layers of the regional and focused models is as follows:

Geologic Formation

Regional Model

Focused Model


Layer 1

Layer 1

Weathered Bedrock

Layer 2

Layer 2

Salina Shale

Layer 3

Not represented

Guelph Dolomite

Layer 4

Not represented

Guelph Dolomite

Layer 5

Not represented

Eramosa Dolomite Unit D

Layer 6

Layer 3

Eramosa Dolomite Unit C

Layer 7

Layer 4

Eramosa Dolomite Units B and A

Layer 8

Layer 5

Goat Island Dolomite

Layer 9

Layer 6

Gasport Dolomite

Layer 10

Layer 7

The horizontal extent of model Layers 2 through 5 of the focused model is the same as the extent of the corresponding layers of the regional model.  The horizontal extent of Layers 9 and 10 was adjusted based on site-specific data obtained from the 2003 investigation.  The regional model relied on the extrapolation of relatively sparse data to define the extent of these layers.  It was found that the actual outcrops of Layers 6 and 7 at the northwestern corner of the model domain (in the vicinity of the Lewiston Reservoir) are approximately 4,000 feet south of the locations used in the regional model.  Subsequent simulations of the focused model indicated that the adjustment had a negligible effect on hydraulic heads produced by the model (on the order of 0.1 to 1.0 feet in the immediate vicinity of the affected region, essentially zero elsewhere in the domain).       Hydrogeologic Parameters

Each model layer is defined by two parameters: transmissivity and vertical leakance.  Transmissivity governs the horizontal flow of groundwater within the layer.  The vertical exchange of water between adjacent layers is determined by the vertical leakance.  Values of these parameters were determined during the calibration of the regional model.  Note that parameter values are not necessarily the same throughout a layer.  In several cases, different model cells within the same layer are assigned different values of transmissivity and leakance.  Values of transmissivity and leakance for each layer are stored as arrays, with each model cell assigned its own value of the given parameter.  

For each layer, transmissivities and vertical leakances of the focused model were assigned by transferring values from the regional model to the focused model.  Because of the different cell sizes of two grids (1,000 by 1,000 feet for regional model, 250 by 250 feet for the focused model) the transfer could not be accomplished directly.  Instead, it was performed by interpolating from the regional model arrays to the focused model arrays. 

Another hydrogeologic parameter used in the model is recharge.  Similarly to the transmissivity and vertical leakance, recharge in the regional model was also obtained during the process of calibration.  Recharge values vary between model cells, and are stored as an array with one value corresponding to each model cell.  The same procedure of interpolation used with transmissivity and leakance was followed in transferring recharge from the regional model to the focused model. 

The interpolation of parameter values from the regional model grid to the focused model grid creates a certain “mathematical” effect, which affects the solution of the focused model.  In areas where the regional model parameters vary from cell to cell, the interpolation into a different (in this case finer) grid of the focused model cannot capture the variation exactly.  Instead, the interpolation produces focused model arrays with the variation somewhat “smoothed” as compared to the regional model.  This effect is typically very minor and does not affect the overall flow pattern.       Initial Hydraulic Heads

In order to initiate the solution, the model requires that a set of initial hydraulic heads be entered.  Inside the model domain, initial heads are adjusted during the simulation until the satisfactory solution is achieved.  The actual values of initial heads inside the model domain are therefore not relevant to the solution.  However, at the specified head boundaries, the initial heads are kept constant, and define the solution.  As a result, hydraulic heads at the specified head boundaries must be known in advance.

The focused model contains specified head boundaries in each of the model layers at the southern and southwestern limits of the model domain.  The hydraulic heads assigned to these boundaries were taken from the regional model.  The hydraulic heads were transferred from the solution of the regional model for each model layer to the corresponding layers of the focused model using the process of interpolation.       Sources/Sinks

The regional model contains numerous sources and sinks, representing both natural and anthropogenic features encountered in the model domain.  Such features include outcrops of water-bearing zones along the Niagara escarpment and Niagara gorge, creeks and swamps, the Lewiston Reservoir, groundwater extraction wells, and the LaFarge Redland Quarry.  These features are shown on Figure 4.1.1-1.

All sinks/sources represented in the regional model are reproduced in the focused model, if they are located within the focused model domain.

Sources/sinks may be represented in either of two ways.  One is to fix the flow of water going into, or out of, a given model cell.  In the regional model, groundwater extraction wells are simulated using this method.  The same method is used and the same flow is assigned to the extraction wells in the focused model.  The second method, used to simulate all other sinks and sources in the regional model, is to specify the flow as proportional to the difference between some given hydraulic head and the hydraulic head within the cell.  This fixes the magnitude of the sink/source as a linear function of the hydraulic head in the model cell containing the feature.

The conversion of the sinks/sources of the second kind from the regional to the local model is not straightforward.  This is because the proportionality constant determining the flow is a function of the number of cells used to represent the feature and/or the cell sizes.  The regional and focused models are based on different cell sizes.  This implies that assigning the same values of the constant of proportionality to represent the same features will not reproduce the effects of the regional model features in the focused model.

Consider a section of a discharging stream modeled as a constant head cell in Layer 1.  The amount of flow that this source cell conveys to Layer 2 is equal to the head differential between two layers multiplied by the proportionality constant (here, vertical leakance) and by the horizontal cell area.  In the regional model, the cells are 1,000 by 1,000 feet.  This is the minimum size of a feature that can be represented.  If the stream section is 100 feet wide and 1,000 feet long (area of 100,000 square feet), it still has to be modeled as a 1,000,000 square foot cell.

In the focused model; however, the cell size is 250 by 250 feet (62,500 square feet).  A 100- by 1,000-foot section of the stream is therefore modeled as 4 cells, total area of 250,000 square feet.  In order to create the same flow as the original 1,000,000 square foot cell of the regional model, with the same head differential between the Layer 1 and Layer 2 cells, the vertical conductance in the focused model must be four times lower than in the regional model.

As a result of this phenomenon, proportionality constants in the focused model had to be adjusted during the construction of the regional model.  This was accomplished by running the model repeatedly and varying the constants until the flow budget of a given sink/source feature in the focused model was within approximately 10% of the flow budget of that feature in the regional model.

Another implication of the difference in cell sizes between the models is the location of each sink/source feature.  The 1,000-foot cell size in the regional model means that features can be located with the accuracy of approximately 1,000 feet of their actual location on the map.  In the focused model, the same features can be located with the accuracy of approximately 250 feet.  Therefore, locations of streams, wetlands, layer outcrops, etc., are slightly different between the two models.       Solution of the Focused Model

The solution of the focused model in the form of hydraulic heads in Layer 2 (fractured bedrock), Layer 5 (Eramosa Dolomite), Layer 6 (Goat Island Dolomite), and Layer 7 (Gasport Dolomite) is shown in Figures 4.1.4-1, 4.1.4-2, 4.1.4-3, and 4.1.4-4, respectively.  Note that these four layers are found within the study area, that is, beneath and immediately east of the Lewiston Reservoir.   The focused model essentially reproduces the solution of the regional model.  See, for example, the hydraulic heads in the fractured bedrock and Gasport dolomite (Figures 4.1.2-1 and 4.1.2-2 of the regional model and Figures 4.1.4-1 and 4.1.4-4 of the focused model).  Small differences exist between the two solutions, although all major features are preserved.  The differences originate from the mathematical artifacts of the conversion between the two grids, as described in the previous sections (smoothing out of parameter variation during interpolation, and adjustments to the location and proportionality constants of sinks/sources).

4.2         Installation of Nested Groundwater Monitoring Wells

The groundwater investigation drilling program featured the installation of up to three piezometers within single boreholes.  This arrangement is referred to as a nested well.  The purpose of installing nested wells was to monitor multiple water-bearing fracture zones at a single location while reducing the number of boreholes that were drilled.  A total of 91 nested bedrock piezometers were installed in 37 boreholes at 17 locations (GW03-001 to GW03-017) across the study area during June through August 2003.  The piezometers were designed to target individual water-bearing horizontal fracture zones identified in the Lockport Group dolomite bedrock aquifer during core analysis or during packer testing of selected boreholes.  At locations where more than three fracture zones were identified, a second or third borehole was drilled.  At locations with multiple boreholes, well Ids were appended with A, B, or C as necessary to distinguish the borehole locations (e.g., GW03-015B). Additional boreholes were installed as necessary (depending on the number of water-bearing zones to be targeted) at the same general locations within approximately 10 to 20 feet of the first borehole location.  A cluster of nine piezometers in three nested boreholes was installed at one particular location (GW03-017).

In addition to the 17 nested bedrock groundwater monitoring well locations, two one-inch piezometers were also installed in shallow overburden soils at location GW03-018 located in a wetland at the head of Gill Creek on Tuscarora lands.

4.2.1        Selection of Nested Well Locations

Nested well locations were selected based on the likelihood of groundwater data from these locations contributing to an evaluation of the effect of the Niagara Power Project operations on the transport of groundwater and contaminants in the investigation area (Figure 1.1-1). 

Eleven nested monitoring wells (GW03-001 through GW03-011) were installed around the Lewiston Reservoir as shown in Figure 4.2.1-1.  These well locations were chosen to evaluate the effect of Lewiston Reservoir operations on groundwater quality and to further evaluate groundwater flow patterns.  Four nested wells (GW03-001 through GW03-004) were installed in a straight line north of the reservoir, with the northernmost nested well located about 600 feet south of the Niagara escarpment.  These wells were situated to more accurately determine the location of a groundwater divide.  This divide marks the location where groundwater moves to the north toward the escarpment or south toward the Lewiston Reservoir.  Five well locations, GW03-005 through GW03-009, were situated east of Lewiston Reservoir to more accurately determine the location of a groundwater flow divide projected to lie in that area.  Five nested wells (GW03-005 through -009) were installed approximately 500 to 2,500 feet east of the reservoir.  These nested wells were situated to examine the extent of influence on groundwater from the operation of the Lewiston Reservoir.  Two nested wells (GW03-010 and -011) were installed approximately 2,000 to 3,000 feet south of the reservoir.  These nested wells were also sited to examine the extent of influence on groundwater from the operation of the Lewiston Reservoir.

Six nested wells (GW03-0012 through -017) were installed along the conduits.  These wells were installed to assess the effect of NPP operations on groundwater levels and flow along the conduits, and to evaluate groundwater quality in this area.

The two nested piezometers set in the wetland at the head of Gill Creek  were installed to assess possible Project and seasonal effects on water levels within the wetland.

4.2.2        Drilling Program

Drilling operations commenced on June 2, 2003.  The drilling rigs consisted of two Versa Drill 2000 Advantage rigs, a Foremost CT 250, and an all-terrain Mobile B-61.  Support trucks carried water tanks used during coring operations.  A pressurized steam cleaner was used to clean and decontaminate the drilling rigs and all equipment used in intrusive activities before drilling commenced and between boreholes. Decontamination activities were conducted using a temporary decontamination station set up at the NYPA storage area.       Surface Casing Installation

Before drilling began, a utilities clearance was performed for each drilling location through the Underground Facilities Protective Organization (UFPO).  In addition, a hand auger was used to excavate a five-foot deep hole at each drilling locations to make certain that no underground utilities were present.  A backhoe was also used at one location instead of a drilling rig to avoid drilling through a water main located near GW03-017.

At the first location, a roller bit was used to drill through overburden glacial deposits to the top of bedrock.  The drillers switched to a 10-inch air hammer on most subsequent holes.  This proved to be a much faster method of drilling through the overburden.  In three boreholes at GW03-015 and one borehole at GW03-017 a 12¼-inch roller bit was used with a soy-based mud additive (Revert) to drill through porous rocky fill.  This drilling method was used to prevent hole collapse and maintain circulation of drilling fluid while drilling through the fill material. 

A temporary 10-inch outside diameter (OD) steel casing was then installed through the overburden borehole to the top of rock.  Soil that accumulated inside the casing was flushed out before coring began. An HQ-size (3.0-inch inside diameter [ID]) core barrel was used to core five feet into rock.  After the core was retrieved, a 10-inch OD air hammer was used to ream the core hole.  A batch of cement-bentonite grout was pumped through a threaded plastic pipe that extended to the bottom of the hole to fill a portion of the borehole.  Following this, an 8-inch OD permanent steel surface casing sealed with a PVC end cap was placed into the 10-inch temporary casing and downward through the grout.  The permanent casing was seated into the bottom of the rock socket.  The grout in the borehole was displaced upward into the annular space between the temporary and permanent casing.  The temporary casing was then withdrawn from the borehole, the level of grout in the borehole dropping as the casing was removed.  Additional grout was added to bring the grout level to the ground surface.  The grout was allowed to cure for a minimum of 48 hours before rock coring commenced.       Rock Coring

A single borehole was cored at each of the 17 locations across the study area. Cores were taken from the Lockport bedrock surface to a depth of five or more feet into the underlying Rochester shale.  At locations where nested wells were installed, no cores were taken from subsequent boreholes that were drilled at the same location.  Instead, these borings were air-hammered from the ground surface to completion depth.

An NQ-size (2.0-inch ID) wireline, eight-foot long, double-tube core barrel assembly was used for coring the bedrock.  The resulting corehole measured approximately 3-inches in diameter.  A wireline core assembly was used for the drilling program because it can be retrieved faster than a conventional core barrel that requires disconnecting sections of drill rods as the core barrel assembly is removed from the hole.  Potable water from a Town of Lewiston hydrant was used as a drilling fluid, or “drill water” (to cool the diamond-impregnated core bit) during the coring operation.  The water also removes finely ground rock particles called rock cuttings or drill cuttings.  The drill water was circulated to the surface where rock cuttings settled into a tub.

A typical individual core section (termed a run) was five-feet long.  (Occasionally, the core barrel would “jam” before completion of the run.)  The driller would remove the core barrel from the hole, carefully remove the core from the inner barrel, and place it in the correct orientation in a wooden core box.  The core was scanned with a photoionizing device to detect volatile organic compounds that may have been trapped in pore spaces and fractures in the rock.  The core box was labeled and photographed.  A URS well site geologist examined the cores and described the rock characteristics and fracturing.  The core boxes were collected at the well site and moved for temporary storage to the above-mentioned NYPA storage yard. 

The total length of rock section cored at individual locations increased to the south across the study area.  The thickening section of rock away from the Niagara escarpment is caused by the dipping of bedrock to the south, and glacial erosion that has leveled the rock surface.  At the northernmost nested well, GW03-004, near the Niagara escarpment, 32 feet of rock was cored.  At the southernmost nested well, GW03-017, near the Niagara River, approximately 148 feet of rock was cored.  Boring logs created from the rock core descriptions are included in Appendix B.       Reaming

The corehole was enlarged (reamed) from a 3-inch diameter opening to an 8-inch diameter opening using an air hammer drilling assembly.  Before reaming, a temporary drilling water containment tub was constructed around the wellhead.  The tub was constructed with a PVC pipe framework and a plastic tarp attached in a watertight arrangement.  The tarp was taped to the wellhead and secured with elastic cords to the plastic frame.  The containment tub, which measured approximately 12 feet long by 8 feet wide and 1.5 feet deep, held 800-1,000 gallons of water and drill cuttings.

The driller used a blast of compressed air to periodically clear cuttings and water from the borehole as the air hammer advanced.  The cuttings and water were captured at the surface and directed into the tub by a rubber and metal shroud that surrounded the wellhead.  The cuttings typically ranged up to ½-inch across.  In fractured zones, chunks of rock as large as 4 inches across were brought to the surface.

Large volumes of water were brought to the surface, particularly at wells installed near the conduits.  Vacuum trucks were used to remove water held in the containment tub while reaming to the well completion depth.  Generally, two 2,000-gallon capacity vacuum trucks were used during borehole reaming.  One truck suctioned water while the second stood by.  The second truck would start working as the first brought its load of water to the six 20,000 gallon tanks stored on Royal Avenue or an additional five tanks stored at the NYPA yard.  At nested well location GW03-016, four vacuum trucks were required to remove water.

The use of an air hammer to ream the boreholes reduces the hydrostatic pressure (pressure exerted by a column of water) on gases like methane and hydrogen sulfide found in some fractures.  In some cases during the drilling program, the pressure reduction allowed these gases to vent to the surface.  While reaming at 80 feet bgs in GW03-011, hydrogen sulfide gas (H2S) was detected.  Rotary drilling with a soy-based additive and water mixture called drilling ‘mud’ was used to ream the hole to its completion depth.  The weight of the drilling mud fluid kept the H2S gas from entering the borehole. Formation gases encountered during the drilling program are discussed further in Section 5.2.6.

4.2.3        Packer Water Injection Testing

Coreholes at ten nested well locations (GW03-001, -003, -005, -007, -010, -011, -013, -015, -016, -017) were tested by injecting water under pressure to sections of the corehole sealed between packers (also referred to as a packer test).  Only the initial corehole was tested at these locations. Subsequent boreholes that were drilled at these locations were not packer-tested.       Purpose

The objective of the packer testing was to aid the identification of transmissive water-bearing fracture zones in the borehole sections.  When the rock is relatively unfractured, as was the case at GW03-003 close to the Niagara escarpment, little water can be forced under pressure into the formation.  When the rock is highly fractured, as at GW03-016, large volumes of water were injected into the formation.

The packer test results, along with information from rock core analysis, were used to correlate identified regional fracture zones with the permeable fracture zones identified in the boreholes.

Packer test results are discussed further in Section 5.2.3.       Methods and Equipment

The coreholes were flushed with water prior to packer-testing to clear cuttings from fractures, but development was not attempted until the hole was reamed to an 8-inch diameter.  The packer-testing was conducted in general accordance with the U.S Bureau of Reclamation procedure USBR 7310-89 (USBR, 1989).  Packer testing was accomplished by lowering a dual packer assembly (called a straddle packer) to the bottom of the corehole at each selected location (see Figure 4.2.3-1).  The straddle packer assembly used was approximately 11.4 feet in length.  It consisted of two inflatable 2-foot long, metal-reinforced, rubber glands (called packers), attached on opposite ends of a 4.4-foot long perforated steel pipe.  Hollow ½-inch ID steel rods were threaded together and connected to the packer assembly as it was lowered into the borehole.

The packers were inflated through plastic tubing connected to a pressurized nitrogen bottle at the wellhead or to an air compressor on the rig.  When the packers were inflated and compressed against the borehole wall, a 5.6-foot test interval was isolated from the rest of the borehole.  Clean potable water was pumped from a holding tank on the rig, pressurized through an FMC beam pump, and injected through the perforated pipe in the packer assembly into the rock formation.  Flow was controlled through a valve on a header assembly at the wellhead. A pressure gauge located near the control valve was used to monitor water pressure.  The driller used the control valve to regulate the flow of water entering the test zone.  The amount of water injected into the formation was measured using an in-line water meter.

Pressure calculations were made before testing began.  The objective was to determine the initial water pressure exerted on the zone being tested and the gauge pressure required at the surface. Care was taken not to use high pressures that could potentially fracture the test interval and cause erroneous readings.  Tests were initially run using three increasing steps of pressure, and pressures were reduced in the same manner.  This method of testing shows whether flow rates are similar at the same pressures on the increasing or decreasing phase of the test.  Localized fractures extending relatively short distances showed a decreasing flow rate over the length of the test.  After a few days of testing, packer test intervals were shortened by using a single pumping pressure (the initially calculated pressure).  Multiple pressure steps were found to yield little useful information beyond that produced during a one-step packer test.

Testing began at the bottom of each borehole.  After the initial test was completed, the packers were deflated and pulled up approximately five feet to test the next zone.  Testing was concluded when the uppermost packer was close to the bottom of the steel casing.  The time taken to test each interval ranged from one minute in zones where no water was injected to sixteen minutes in permeable zones.  Flow rate into the test interval was determined by recording the number of gallons injected on a minute-by-minute basis.  Gauge pressure readings were also recorded.

The packer assembly and rods were decontaminated prior to packer testing using a steam cleaner.  They were steam-cleaned again before testing equipment was moved to the next borehole.

4.2.4        Piezometer Installation

After the core holes were reamed to completion depth, the air hammer was used to remove some of the fine material that had accumulated in fractures in the well bore.  This procedure, called well development, is discussed further in Section

The nested wells were constructed in a nested arrangement with up to three piezometers in a single borehole (see Figure 4.2.4-1).  A total of 91 piezometers were installed during the drilling program.  Five boreholes had a single piezometer, 10 boreholes had two piezometers, and 22 had three piezometers.

The two piezometers installed in shallow overburden soil in the wetland at the head of Gill Creek (GW03-018-P1 and GW03-018-P2) were constructed of  1-inch schedule 40 PVC. Each piezometer consisted of a 1-foot schedule 40 PVC screened interval (0.010-inch slot size) and blank PVC riser.  Temporary 1.5-inch steel casing was driven into the soil using a manual slide hammer.  The piezometers were then placed within the temporary casing and the casing was withdrawn allowing the soils to collapse around the piezometer screens and riser. Piezometers GW03-018-P1 and GW03-018-P2 were installed to depths of 2.14 feet and 4.23 feet, respectively.       Identifying Water-Bearing Zones

Johnston (1964) measured Lockport Group fracture elevations and their lateral extent during the construction of the conduit.  His observations of fracture characteristics and location of seepage points provided a basis for understanding groundwater flow in the Niagara Region that was used by later workers (e.g., Tepper et al. 1990, and Yager 1996). Other bedrock exposures and corehole information also contributed to understanding groundwater flow in the area.  Widespread horizontal bedding plane fractures in the Lockport Group were also observed in outcrops in the Niagara gorge and at the LaFarge Redland dolomite quarry.  Horizontal fractures were identified in numerous rock core samples taken from hazardous waste sites such as DuPont’s Necco Park landfill, located in the southern portion of the study area.  Detailed analysis by geologists of rock cores taken from over 100 wells at the 24-acre Necco Park site showed that some fracture horizons are discontinuous across the site (WCC 1989).  In addition, pumping tests from wells at Necco and surrounding properties show considerable variability in permeability and well yield (DuPont et al. 2002).

This study utilized previously published work, past experience in the area, and recent information gathered from rock core examination, wellhead observations, and packer testing to identify water-bearing fracture zones at the 17 locations in the study area.

Continuous rock core sampling in the Lockport Group was used to correlate lithology and fracture zones across the study area.  Fractures were also noted during air-hammer drilling.  In some instances, drill rods were observed to drop slightly as the air hammer encountered a fracture.  Drilling water circulation loss and increased drilling rates were often noted when the core barrel encountered a significant fracture zone.  Packer testing results were also used to determine fracture zones for piezometer screen placement.

Fracture zones identified in this study were correlated to regional fracture zones identified by Johnston (1964) and Yager (1996).  The screened fracture zones identified in the study appear to be significant and regional in their extent.  Fracture characteristics are discussed in Section       Nested Well Construction

The groundwater monitoring wells were often constructed in a “nested” arrangement of two or three piezometers installed in the same borehole.  A second or third piezometer screen and riser was added if the borehole intersected multiple water-bearing fracture zones.

A typical nested well construction diagram is shown for illustration purposes in Figure 4.2.4-1.  Individual nested well construction diagrams are included in Appendix C.

The piezometers were constructed of 2-inch diameter Schedule 40 PVC continuously wound 0.030-inch slotted screen attached to Schedule 40 solid PVC riser pipe.  The screen sections ranged from 3 to 10-feet depending on the thickness of the fracture zone.  Sand was added to the borehole as filter material around the piezometer screens. Layers of impervious hydrated bentonite clay separated the filter sand from successively higher piezometers if they were installed in the same borehole.

A final layer of bentonite was added to the piezometer above the top piezometer’s filter sand. After the bentonite hydrated, the remainder of the open borehole annular space was completed with cement/bentonite grout to the surface.

Nested wells located in traffic areas were finished with a cylindrical, 12-inch diameter flush-mounted wellhead with a lid secured by bolts.  This type of wellhead is often called a road box.  The road box was placed around the surface casing and surrounded by concrete.  Riser pipes inside the road boxes were fitted with expandable plastic plugs.  The expandable plug provides an airtight seal at the top of the riser pipe.

Nested wells located away from traffic areas were completed with the surface casing extending about 2-feet above the ground.  A wellhead cap was attached to the top of the casing by hex bolts.  The wellhead was secured with a padlock.  Riser pipes inside the steel casing were fitted with expandable plastic plugs.       Nested Well Development

A considerable amount of water and drill cuttings were lifted to the surface during the air-hammer reaming of the core holes.  After the holes were reamed to their completion depths, the air hammer assembly was raised and lowered rapidly in approximately 20-foot long strokes for a few minutes.  This swabbing action agitated water contained in the boring and pulled drill cuttings lodged in borehole fractures into the well water.  The water was then flushed from the borehole with a blast of compressed air through the air hammer. This jetting was done for five to ten minutes, usually until the containment tub filled with water.  In relatively low-yield boreholes, a second round of flushing was carried out.  Water coming from the wellhead was often observed to become less turbid toward the end of flushing.

After the nested piezometers were installed, a second phase of piezometer development was performed.  Submersible pumps were lowered to a few feet from the bottom in the individual piezometers and the groundwater was pumped to the surface, where it was collected in plastic tanks.  The water purged from the borehole and piezometers was later transferred to a larger tank.

Water quality parameters were measured at regular intervals during development of the piezometers.  These parameters included pH, specific conductivity, temperature, turbidity, and oxygen reduction potential.  When the water quality parameters were determined to be stable, piezometer development was completed.  In general, more than three times the amount of water contained within the riser and screen was removed from the piezometers.  In low-yield piezometers, water quality parameters were recorded as the piezometer was pumped to dryness.  Piezometer development logs are included in Appendix D.       Management of Investigation-Derived Waste

Investigation-derived waste (IDW) from the drilling program consisted of trash, drill cuttings, used drilling mud, and water recovered from drilling and well development.       IDW Trash

IDW trash consisted of items such as empty sand and bentonite bags, used plastic tubing, containment tub plastic liners and miscellaneous cardboard and plastic sheeting.  These materials were transported daily to a roll-off staged at the NYPA storage yard.  The trash was disposed of at Modern Landfill in Lewiston, New York.       IDW Drill Cuttings and Drilling Mud

Used drilling mud was stored in 55-gallon drums. The drums were sealed, labeled, and stored near the wells.  Drill cuttings were temporarily staged on the plastic containment tub liners at individual wellheads, covered to protect them from the weather.  Composite samples of drill cuttings from wells drilled at each of the 17 locations were collected in July and August 2003 and submitted to Mitkem Corporation of Warwick, Rhode Island.  The analytical work was in turn subcontracted to Severn Trent Laboratories of Shelton, New York.  The samples underwent a Toxicity Characteristic Leaching Procedure (TCLP) extraction.  The TCLP extract was analyzed for volatile organic compounds (VOCs) by United States Environmental Protection Agency (USEPA) Method 8260B; TCLP semivolatile organic compounds (SVOCs) by USEPA Method 8270C; TCLP pesticides by USEPA Method 8081A; TCLP herbicides by USEPA Method 8151; and TCLP metals by USEPA Methods 60110B/7470A.  The drill cuttings were also analyzed for polychlorinated biphenyls (PCBs) by USEPA Method 8082, and Resource Conservation Recovery Act (RCRA) characteristics including corrosivity by USEPA Method 9045C, reactive cyanide by USEPA Method 9014, reactive sulfide by USEPA Method 9034M, and ignitability by USEPA Method 1030.

A summary table of the analytical results for the drill cuttings is provided in Appendix E.  Complete drill cuttings analytical data tables with validation qualifiers are also provided in Appendix E.

Most of the drill cuttings were removed and transported to Modern Landfill in Lewiston, New York for disposal.  Drill cuttings at remote locations GW03-001 through -008 -and -013 were left at the drill sites.  Plans have been made to remove these cuttings in 2004.  Scale tickets received when the trucks transporting the drill cuttings entered Modern’s facility are included in Appendix F.       IDW Water

As discussed previously, water (a mixture of Town of Lewiston hydrant water and groundwater, with some suspended clay and silt-size rock particles) was removed as it was generated during drilling and well development.  A total of approximately 225,000 gallons of water was recovered. 

Six samples of water and one field duplicate were collected between July 22 and August 8, 2003, and submitted to Ecology and Environment, Inc., of Lancaster, New York.  In order to evaluate the suitability of the water stored for disposal into the city’s sanitary sewer system, the samples were analyzed for an array of parameters requested by the City of Niagara Falls, New York, Department of Waste Water Facilities. The parameters are: VOCs by USEPA Method 8260B; SVOCs by USEPA Method 8270C; pesticides by USEPA Method 8081A; herbicides by USEPA Method 8151; PCBs by USEPA Method 8082; metals by USEPA Methods60110B/7470A; total phosphorus by USEPA Method 365.2; total cyanide by USEPA Method 335.3; total organic carbon by USEPA Method 9060; total recoverable phenolics by USEPA Method 420.2; and total suspended solids (TSS) by USEPA Method 160.2.  The TSS analysis was subcontracted to Waste Stream Technology, Inc., of Buffalo, New York.

A summary table of the detected analytical results for the water samples is provided in Appendix E.  Complete drilling water analytical data tables with validation qualifiers are also provided in Appendix E.

The analytical data were sent to Niagara Falls Wastewater Facility and the Town of Lewiston Water Pollution Control Center for review.  Permits were issued by the City of Niagara Falls and the Town of Lewiston for disposal of the water.  The water disposal permits issued are included in Appendix G.  The approximately 125,000 gallons of water contained in six tanks located on Royal Avenue was pumped into a nearby sanitary sewer manhole (South Side Interceptor) in August 2003.  During the same month, the approximately 100,000 gallons of water stored in five tanks in the NYPA storage yard was transferred by truck to the Town of Lewiston Water Pollution Control Center for treatment and disposal.

4.2.5        Well Location Survey

New York State licensed surveyors from URS used differential level surveying and Global Positioning System (GPS) receivers to survey the locations and elevations of the nested wells.  Surveying fieldwork took place in September and October 2003.  The survey data were checked for accuracy and tabulated on October 2, 2003.  Survey data are provided on Table 4.2.5-1.

The surveyors used a Topcon automatic level instrument to survey ground surface elevations at each nested well location and the elevation of the top of riser pipes at each piezometer installed during the drilling program.  The surveyors started from a record benchmark and referred to a second benchmark. Differences in elevations were balanced (i.e., error-adjusted).  The elevations were referenced to the U.S. Lake Survey (USLSD) 1935 datum.  Elevations are accurate to 0.05’ +/- based upon federally established record area benchmarks.

Horizontal control was established using Trimble 4400 GPS receivers.  The measurements were referenced to North America Datum (NAD) 1927, used when the Project was built.  The measurements were also referenced to NAD 1983, which is currently used.  The GPS receivers use real-time kinematic positioning that is accurate to +/- 1-foot.

4.3         Groundwater Level Monitoring Program

Groundwater elevations were measured in order to evaluate groundwater flow patterns and to assist in the development of a groundwater model for the purpose of further refining the Project area of influence.  A groundwater-level monitoring program was implemented beginning at the end of July 2003 and lasting through May 2004.

Groundwater and surface water elevation data were collected from various locations distributed widely across the investigation area (see Figures 4.3-1 and 4.3-2).  These included a combination of wells and piezometers, either newly installed as part of the Phase I investigation (as discussed in Section 4.2) or existing conduit and reservoir monitoring wells.  Groundwater elevation data were collected using a combination of manual level measurements and electronic data level measurements.  Both methods are discussed in greater detail below. 

4.3.1        Groundwater Level Monitoring Locations

The locations of the newly installed nested wells (GW03-001 through GW03-017) were as follows: locations GW03-001 through GW03-004 were installed at roughly 2,000-foot intervals along a north-south line on the western border of the Tuscarora Nation north of the Lewiston Reservoir; GW03-003 through GW03-008 were installed north and east of the Lewiston Reservoir at approximately 2,000-foot intervals along a north-south trending line within the Tuscarora Nation; GW03-009 through GW03-012 were installed roughly 2,000 feet from the east, west, and south walls of the reservoir; and GW03-012 through GW03-017 were installed from north to south, respectively, at 2,000-foot intervals approximately 50-feet from the east wall of the east conduit (Conduit 1).  Two one-inch piezometers were also installed in shallow overburden soils at location GW03-018 located in the wetland at the head of Gill Creek within the Tuscarora Nation.  Locations are shown in Figure 4.3-1.

Other locations monitored include Pump Stations A and B (including independent monitoring of both the CDS and the conduits); one existing NYPA conduit monitoring well (OW-139); one private well installed on NYPA property (OW-650D), previously monitored in 2002; and permanent surface water level gauges GN_Intakes, GN_Forebay and GN_Reservoir (Figure 4.3-1).  Reservoir dike monitoring wells are shown in Figure 4.3-2.  These locations are discussed in greater detail below.       Nested Groundwater Monitoring Wells

As described in Section 4.2, a total of 17 monitoring locations (GW03-001 through GW03-017) were installed through overburden materials into bedrock (see Figure 4.2.1-1).  Each monitoring location has up to three nests of between 1 and 3 piezometers each, with a maximum of nine piezometers at one location.  As discussed in Section, each piezometer was designed to intersect a specific groundwater model layer as defined by Yager (1996).  A complete discussion of groundwater flow model layers is presented in Section 4.1.

In order to determine groundwater elevations within a wetland located at the head of Gill Creek, two piezometers (GW03-018A-P1 and P2) were installed into the overburden materials within the wetland.  These piezometers were installed by hand (via slide hammer) below ground surface within a mostly filled (via natural processes) former drainage swale located in the western half of the wetland.  Piezometer P1 was advanced to 2.14 feet below the surface and P2 was advanced to 4.23 feet below the surface.  Due to their shallow installation, these piezometers were not used to collect water samples for chemical characterization.  Well construction details are provided in Appendix C.       Conduit Observation Wells and Lewiston Reservoir Wells

Groundwater levels in two previously installed bedrock wells along the conduits (OW-139 [NYPA well], and OW-650D [private well]) were monitored using electronic dataloggers and manual level measurements (Figure 4.3-1).  The NYPA “OW” wells along the conduits were originally installed as well pairs to monitor groundwater near the conduits.  Wells installed near the outer edge of the conduit right-of-way (usually within 50 feet of the conduits) were drilled through rock to a depth below the conduit invert (PASNY circa 1963).  Conduit observation wells were not constructed to monitor discrete hydrogeologic flow zones. 

Electronic data from conduit wells OW-139 and OW-650D were collected continuously at 15-minute intervals.  Electronic data were collected from September 3 through December 8, 2003, at OW-139 and from August 21 through December 8, 2003, for OW-650D.  In addition, dataloggers were calibrated each week with manual level measurements.  After December 8, 2003, the dataloggers were pulled from both locations.   Additionally, no manual measurements were recorded after December 8th.

A subset of Lewiston Reservoir wells with electronic dataloggers was selected to represent the groundwater levels near the reservoir (see Figure 4.3-2).  Datalogger data from the reservoir wells were collected at six-hour intervals.  Reservoir locations were not calibrated or monitored with manual level data.  These are permanent gauges monitored directly by NYPA.  Electronic data for these locations are collected on a continuous basis, although only data from July 2003 through May 2004 were used for this investigation to correspond with the data collection period of this investigation. 

Lewiston Reservoir wells designated “OW” were installed between 1958 and 1960 (during reservoir construction) to provide data for the evaluation of reservoir dike stability.  The OW wells were constructed as 6-inch diameter open hole wells set below the bedrock surface.  Bedrock wells designated as “NW” were installed in 1991 as piezometers using 2-inch ID PVC casings and 20-foot-long 0.01-inch slotted screens.  Differences in well construction may influence comparisons of water levels between wells due to uncertainty in the flow zones intercepted by each well.  Survey data for reservoir and conduit wells is presented on Table 4.3.1-1.       Pump Stations

Water level data were collected from inside both Pump Stations A and B (Figure 4.3-1).  The pump houses were constructed to allow for the release of excess groundwater head in the CDS via a central sump (which is directly connected to the CDS of both conduits) and weirs (Figure 2.2.1-3).  The sumps and weirs permit the flow of groundwater from the CDS into the conduits.  Water level data were collected from both the central sump (GW03-PSA-D and GW03-PSB-D) and the receiving side of Conduit 1 (east conduit) (GW03-PSA-C and GW03-PSB-C) within each pump station.       NYPA Permanent Surface Water Gauges

In order to further understand the interaction between surface water and groundwater, data on surface water levels were collected from existing staff gauges (1) near the NYPA intakes (Gauge ID: GN-RIVER_INT), (2) at the forebay (Gauge ID: GN-FOREBAY), and (3) at the Lewiston Reservoir (Gauge ID: GN-RESERVOIR) (Figure 4.3-1).  During the investigation period (July 2003 through May 2004), gauge data were gathered hourly.  Data from a fourth gauge were not included in the study.  This is the surface water gauge located in the Chippawa-Grass Island Pool (Gauge ID: GN-MATERIAL), which provides the official measurement of water level fluctuations to determine compliance with U.S. and Canadian agreements on the use of the Niagara River for power generation.  It was determined that water levels recorded at the Intakes, Forebay and Reservoir gauges would supply more usable data for the evaluation of Project effects on groundwater within the investigation area. 

Data from the three gauges used for the study were not calibrated with manual measurements.

4.3.2        Manual Water Level Measurements

Manual water level depths were measured from a fixed marked elevation point (from survey data) on each piezometer and/or well.  Depth-to-water measurements were then converted from a fixed measuring point elevation to groundwater elevation data.  Manual water levels were primarily used to maintain the calibration of the vertical location (i.e., elevation) of electronic dataloggers (See Section 4.3.3) used throughout the study and to monitor long-term trends in groundwater elevation within each piezometer.

Initial water level data for Phase I investigation locations (GW03-001 through GW03-017) were collected prior to piezometer development.  Following development, manual water level data were collected with a water level indicator instrument in each piezometer at each location.  Manual levels were also collected from GW03-018A-P1 and P2, OW-139, and OW-650D.  At the pump stations, manual water levels were recorded for the calibration of each of the four dataloggers (GW03-PSA-C and D and GW03-PSB-C and D) only twice, during initial installation of the dataloggers and following replacement of batteries in March 2004 (as discussed below in Section 4.3.3).  As noted earlier, no manual level data were collected from the reservoir wells or permanent surface water level gauges.

Initially, to maintain calibration of datalogger position, manual water level data were collected twice a week through the beginning of November 2003.  Due to the stability exhibited by the datalogger mounts, the schedule was scaled back to once weekly for the remainder of November.  With the exception of GW03-016 and GW03-017 locations, manual water level data collection was completed on November 25, 2003.  Due to datalogger rotation schedule requirements, GW03-016 and GW03-017, OW-139 and OW-650D, were monitored till December 8, 2003.  No manual measurements were collected between December 2003 and March 2004.  With the exception of one round of measurements taken on February 23, 2004 as a precursor to the third groundwater sampling event (see Section 4.5).   Between March and May 2004, manual water levels were collected once a month.

Ideally, between July and December 2003, data on manual levels were to be collected from GW03-001 through GW03-018, OW-139, and OW-650D during each collection round.  Due to logistical difficulties, however, (namely, access to locations, weather, etc) only partial rounds of data were collected during certain weeks.   Tables 4.3.2-1 and 4.3.2-2 show manual groundwater elevation data for each location.

4.3.3        Electronic Water Level Measurements

Continuous groundwater level monitoring was accomplished by means of electronic dataloggers (MiniTroll Pro®, by In-Situ, Inc.®) set to record water pressure and temperature for the purpose of observing both short-term and long-term trends in the fluctuation of water level elevations.  Datalogger pressure readings were converted by the internal datalogger software from pounds per square inch (psi) to feet of water column using standard mathematical conversions (with internal barometric compensation accomplished via an integral tube in the cable).  Water column values were then correlated to water elevations using procedures outlined in Section 4.3.4.

Each datalogger (for locations GW03-001 through GW03-018, OW-139, OW-650D and Pump Stations A and B) was programmed to record feet of water at 15-minute intervals.  Initially, in order to determine that each datalogger was recording correctly and that it remained stationary inside the piezometer or well (i.e., remained calibrated), data were downloaded on a weekly basis.  Downloading was accomplished by means of a pocket iPAQ™ computer, using software and hardware specifically designed to communicate with the dataloggers.  No downloads were attempted between December 2003 and March 2004.  Between March and May 2004, downloads were completed on a monthly basis.

Calibration was accomplished during each download by recording simultaneous manual and datalogger water level measurements.  A schematic of a typical datalogger setup is presented in Figure 4.3.3-1.   Due to access restrictions, dataloggers within Pump Stations A (GW03-PSA-C and D) and B GW03-PSB-C and D) were calibrated only upon installation and during battery replacement. The reservoir dike monitoring wells were not calibrated during the monitoring period.

In the beginning of the monitoring period, dataloggers collected data in a specific piezometer for two weeks.  Following the collection of initial data for each piezometer, and based on the 1996 USGS model (Yager 1996) and corresponding screen intervals, the dataloggers were re-deployed.  Table 4.3.3-1 shows the schedule and rotation of each piezometer and well. 

After a final calibration was performed in 2003, several dataloggers were removed.  However, a total of 18 dataloggers remained in place and continued to record data between December and March 2004.  During this period, no calibrations were conducted.  Electronic data from these locations were downloaded, removed, batteries changed, re-installed, and re-calibrated in March 04.  Table 4.3.3-1 presents which locations remained, in addition to which locations were redeployed between March and May 2004. 

Since the surface water and Lewiston Reservoir monitoring well dataloggers are permanent, they were not rotated as part of this monitoring program.

Between July and December, a few piezometer dataloggers were not deployed.  In some cases (GW03-006A-P3, GW03-014A-P2, GW03-015A-P3 and GW03-017A-P3), this was because the depth to water exceeded the longest available cable length.  In other cases (GW03-012A-P1, GW03-012-P4, GW03-012-P5, and GW03-012-P7), it was because little to no water was observed in the piezometer (i.e., the well was dry). Due to a scheduling error, no datalogger was deployed in piezometer GW03-007-P3.  Dataloggers were not re-deployed in GW03-018-P1 and P2, OW-139, and OW-650D during March through May 2004. 

4.3.4        Data Management

Following each download of electronic data or manual water level collection, all data points were either converted or entered into an Excel spreadsheet.  Manual water levels were entered into a spreadsheet for each piezometer and well.  The measured levels were then subtracted from the fixed measuring point elevation (obtained from survey data) to obtain a value for groundwater elevations.  These data were then transferred into a database and graphed.  The data were cross-checked against the original field notes as a quality assurance measure.

Quality assurance measures for electronic data were more complex.  Electronic data from the dataloggers were converted from a proprietary format to Excel.  These data had, however, been recorded as height of the water column without an elevation reference.  To convert from height of water column (HoWC), to depth to water (DTW), therefore, the depth of the datalogger within each piezometer or well had to be calculated using the simultaneous manual water level DTW and electronic water level measurement of HoWC (Figure 4.3.3-1).  As noted earlier, each time data were extracted from the datalogger, in order to maintain an accurate calibration, these two measurements were made.  The average of the datalogger depth calibration values was then used to determine the depth to water from the top of the piezometer or well by subtracting the HoWC (raw data) from the datalogger’s average depth calibration (sum of simultaneous HoWC and DTW measurements).  This calculation converted data into a depth to water measurement from the top of piezometer or well.  Finally, in order to determine groundwater elevation, the depth to water value was then subtracted from the surveyed measuring point elevation.

After these calculations were completed, the data was imported into a database.  Further quality assurance was completed by graphing each piezometer and well location and comparing to the original spreadsheet and field notes to determine whether reference elevations and depth to water calculations were correct. 

Data from the permanent surface water and reservoir monitoring well locations were provided by the Niagara River Control Center and NYPA, respectively, and were imported into the database with no quality assurance checks.

4.3.5        Field Reconnaissance

Field reconnaissance was performed on Fish and Gill Creek and the LaFarge Redland quarry in order to determine the effects of groundwater flow in these areas.  In addition, a limited seep reconnaissance of the escarpment (limited because of accessibility and safety issues) was also performed.  These are discussed in detail below.       Escarpment Seep Reconnaissance

On June 16, 2003, in an attempt to identify and estimate the frequency and volume of seeps along the escarpment, a limited seep reconnaissance was conducted along the escarpment’s northwest section.  After an initial drive-by reconnaissance proved to be ineffective due to vegetative cover, a walking reconnaissance was conducted.  Two sections were walked: one short section south of Route 104 (underneath the north-south power line near the western border of the Tuscarora Nation), and one long section along the former Hojack Railroad right-of-way (Figure 4.3.5-1).

The short section was walked east to west for 300 to 500 feet approximately 3/4 of the distance up from the bottom of the escarpment (at the interface between the Lockport and Clinton Groups [i.e. the bottom of the Gasport dolomite and top of the DeCew dolomite]).  The long section was walked west to east starting at Route 104 (immediately north of the Niagara Falls Country Club and Mountain View Drive) to the east for approximately 0.75 miles.       Fish and Gill Creek Reconnaissance

The reconnaissance conducted of Fish and Gill Creeks (Figures 4.3.5-2 and 4.3.5-3, respectively) included the completion of a flow profile and volume flow rate calculation for each creek.

The headwaters for both Fish and Gill Creeks are located within Tuscarora Nation Lands east of the Lewiston Reservoir. 

On October 15, 2003, at each creek, a flow profile area was measured along a transect perpendicular to the direction of water flow (Figures 4.3.5-2 and 4.3.5-3).  Along each transect, depth of water measurements (from water surface to creek bed) were made for each horizontal 0.5-foot increment.  Current velocity rate measurements for each 0.5-foot interval were made using a Marsh McBirney Flowmate Model 2000 velocity meter.  Two transects were attempted along each creek but current velocities along the A-A’ transect for Fish Creek were too low for the flow meter to measure and thus measurements at that transect were discontinued.

4.4         Falls Street Tunnel Investigation

A flow study was developed to evaluate the relationship between the NYPA conduits and groundwater infiltration into the FST.  The study involved conducting a tunnel reconnaissance to determine areas of groundwater infiltration, measuring and recording groundwater infiltration flow rates, and evaluating effects of NPP operations on infiltration rates.

4.4.1        FST-Conduit Crossing

Placement of the NYPA conduits beneath the FST necessitated demolition and removal of a section of the FST (Figure 4.4.1-1).  Following placement of the conduits, a section of 84-inch diameter concrete pipe, long enough to span the conduit excavation, was installed where the removed rock section had been.  To carry FST flow during construction, a parallel bypass section was also constructed, also of 84-inch diameter pipe.  In the 1970s, the SSI was constructed by tapping into this bypass and building out both upstream and downstream of it (Figure 4.4.1-1).  The SSI was designed to carry dry-weather combined sanitary/stormwater flows to the wastewater treatment plant. Except for the bypass section, which also consists of 84-inch diameter concrete pipe, the SSI is a rock tunnel lined with 6-foot diameter concrete pipe.  At the crossover, it runs parallel to the FST, approximately 150 feet south of it.  Like the FST, it slopes to the west (i.e., towards the City of Niagara Falls Wastewater Treatment Plant).  It passes over the conduit right-of-way just south of conduit Pump Station A.

In the vicinity of the FST-conduit crossover, both the conduits and the FST are located below the groundwater table.  Even though this makes infiltration likely along the entire length of the FST, the section of the FST in the vicinity of the crossover has historically experienced higher rates of groundwater infiltration than upstream or downstream FST sections.  Since 1989, the City of Niagara Falls has undertaken a series of projects that have resulted in a significant reduction in groundwater infiltration in this section (see Section 2.2.2).  The SSI does not receive significant groundwater infiltration presumably due to the concrete lining method of construction.  The subject of this investigation is the quantification of dry-weather flow within this section of the FST.

The FST crosses over the NYPA conduit right-of-way along the southern edge of Royal Avenue (Figure 4.4.1-1).  To either side of the conduit excavation, the FST is an unlined rock tunnel, approximately 6 by 7 feet in cross-section.  Where a section of tunnel was removed for installation of the conduits, an 84-inch diameter concrete pipe, installed with internal and external seals in a vault, has been placed.  From this replacement section to a point approximately 500 feet west of the conduits, the tunnel has been grouted.  As mentioned above, a similar concrete section was put in place as a bypass when conduit excavation was in progress.

At the FST-conduit crossing, the FST and SSI are joined via two 84-inch diameter pipes (part of original bypass section), located on either side of the NYPA conduits (Figure 4.4.1-1).  The pipes are approximately 160 feet long.  Under dry-weather conditions, all FST and SSI flow is routed to the treatment plant.  Six gates make it possible, however, to divert flow from the treatment plant if necessary.  Two of these gates are built into drop shafts within the FST, two are in the SSI, and one is in each of the 84-inch flow diversion pipes constructed as part of the original bypass (Figure 4.4.1-1).  Under normal conditions, all but one of these gates are fully open, with all flow through the FST and SSI being gravity-controlled only.  (The one gate that is not fully open affects only wet weather flow).  Dry-weather flow in the system is controlled by four weirs or stoplog structures within the FST.  As shown in Figure 4.4.1-1, they are located as follows:

·        in Drop Shaft 12, approximately 300 feet west of the conduits,

·        immediately west of the FST junction with the west flow diversion pipe,

·        in Drop Shaft 13A, immediately west of the FST junction with the east flow diversion pipe, and

·        in Drop Shaft 14B, approximately 1,500 feet east of Drop Shaft 13A.

4.4.2        Dry Weather Flows

Flow direction in tunnel and interceptor is from east to west.  While most sanitary sewer input into the FST east of the site has been rerouted into the SSI, a pipe was nevertheless seen (during the field reconnaissance carried out for this investigation, described in Section 4.4.3) discharging directly into the FST at Drop Shaft 14 (Figure 4.4.1-1).  The pipe, approximately 2 feet in diameter, appears to be coming from the north side of Royal Avenue, the location of a small industrial/commercial facility.  The origin of flow in the pipe is not known.  The flow depth in this pipe during the site visit was on the order of 1 inch.  The entire manhole into which the pipe discharged was filled with hard deposits thought to be calcium carbonate.

The section of the FST east of the conduits, between the 84-inch concrete section and Drop Shaft 14, is known to receive infiltration from groundwater.  This flow is redirected into the 84-inch diversion pipe and the SSI by the 3-foot-high diversion dam located immediately west of the junction of the tunnel and the east flow diversion pipe (Figure 4.4.1-1).  This flow was observed during the site visit.  This flow was reportedly measured by the City of Niagara Falls some years ago (one-time instantaneous sample measurement) at 2.9 mgd (2,000 gpm) (verbal communication with Richard Roll, 9/3/03).

Infiltration into the FST’s concrete section, reportedly measured at one time to be several mgd, has been reduced through sealing and grouting to approximately 100 gpm (Roll and Lannon 2001).  This flow is directed into the SSI by means of the stoplog structure located immediately west of the junction of the tunnel and the west flow diversion pipe (Figure 4.4.1-1).

Infiltration also enters the FST through the section west of the conduits.  In spring 2001, infiltration in this area was estimated at 1.6 mgd (Roll and Lannon 2001).  A weir structure at Drop Shaft 12 creates a backwater, redirecting flow upgradient, towards the stoplog structure near the west flow diversion pipe.  Under design conditions, this flow spills over the stoplog structure and into the SSI via the west flow diversion pipe.  During the site visit, however, water was observed also to be flowing over the weir at Drop Shaft 12.  Flow passing over this weir continues to the Gorge Pumping Station where it is pumped to the treatment plant.

As a result of the use of weirs and stoplog structures to direct the flow from the FST to the SSI, backwater conditions arise at various locations.  These backwater conditions cause significant accumulation of solids, including sediment and debris, in the FST and flow diversion pipes, leading to a reduction in system efficiency.

4.4.3        FST Reconnaissance

On September 22, 2003, a reconnaissance of the FST system in the vicinity of the NYPA conduits was conducted to observe existing conditions and to determine the optimum design of a flow-measuring program.  Approved confined-space entry procedures were used for all tunnel reconnaissance activities.  Confined-space entry procedures employed for this inspection included continual monitoring of tunnel atmospheric conditions; use of standby personnel on the surface, equipped with emergency rescue equipment; and use of constant radio communication between tunnel entrants and surface standby personnel.  Moreover, Wastewater Treatment Plant personnel from the Niagara Falls Water Board (NFWB), which owns and operates the wastewater system in the City of Niagara Falls, were notified of all tunnel inspection activities, and were provided with a cell phone number to call in case of any condition that would necessitate the immediate extraction of tunnel inspection personnel.

The FST system reconnaissance involved tunnel entries at several different locations along the tunnel.  They were, in order of entry: Drop Shaft 14A, Drop Shaft 12, Drop Shaft 13A, the East Bypass Gate Structure, and Drop Shaft 13.

At each location, observations regarding tunnel flow, groundwater infiltration, inflows, and sediment accumulation were recorded.  Observations were made of the FST both up- and downstream of the entrance point as well as within both the east and west SSI bypass tunnels.  The inspections were also documented using a digital camera.  Results of tunnel reconnaissance, including photographs, are presented in Section 5.4.

4.4.4        Flow Measuring Program

Based on observations recorded during tunnel reconnaissance, URS developed a program to monitor flow rates of groundwater infiltration into the FST.  URS determined that, during dry-weather conditions, all groundwater infiltration entering the FST in the vicinity of the NYPA conduits flows to one of three places: (1) over the FST measuring weir located immediately upstream of Drop Shaft 12, (2) through the SSI via the east bypass tunnel, or (3) through the SSI via the west bypass tunnel.  The program developed involved placing flow-monitoring equipment at these three locations in order to monitor infiltration rates.  The program additionally involved making regular visual observations of inflow at the industrial discharge located at Drop Shaft 14, and of any flow entering the FST at Drop Shaft 14A.  Since the above holds true only during dry-weather conditions, data collected during dry-weather conditions were used to estimate rates at which groundwater infiltrates into the FST.  During wet weather, inflow from other sources, such as backup from the SSI into the FST, precludes the accurate measurement of flows.

The flow-measuring program for all three locations was implemented for a period of approximately one month starting in mid-October.  This allowed for measuring groundwater infiltration for approximately 2 weeks before and 2 weeks after November 1.  This was done in order to evaluate the groundwater infiltration rates under both tourist season (April 1-October 31) and non-tourist season (November 1-March 31) operating conditions.       Drop Shaft 12

An existing measuring weir was used to measure flow in the FST at Drop Shaft 12.  The weir at drop Shaft 12 is sharp-crested, with a rectangular opening 60 inches wide and 12 inches high.  A Sigma 920 flow meter equipped with a single depth probe was used to continuously record flows at this location.  The depth sensor was installed upstream of the weir and set to measure the depth of flow above the weir crest at 5-minute intervals.  The flow meter was set up to automatically convert this depth to a flow rate using the equation for a 60-inch rectangular weir with end contractions.

For depths greater than 12 inches above the weir crest, the weir opening is totally submerged and the flow cross section is no longer rectangular. Therefore, calculation of flow for depths greater than 12 inches above the weir crest was not possible. These data; however, correspond to significant wet weather events and are not used in the calculation of groundwater infiltration, which was performed based on dry-weather data.  Detailed discussion of the flow monitoring results is presented in Section 5.4 and Appendix H.       East and West Bypass Tunnels

It was noted during the tunnel reconnaissance that the bottom of each 84-inch SSI bypass pipe was covered by approximately one foot of sediment.  Flow depths over the sediment were observed to be approximately one foot.  The cross-sectional flow created by the flat sediment surface and the 84-inch pipe walls resembles flow in a trapezoidal channel.  In order to obtain flow measurements in both these tunnels, flat, steel-bottom platforms (Figure 4.4.4-1) were designed, fabricated, and installed in both the east and west SSI bypass pipes in order to provide regular flow channels.  The flow channel created by the platforms and the tunnel walls measured approximately 60 inches wide at the base, and varied in width at the top depending on the flow depth.

At each tunnel location, a Sigma 920 flow meter equipped with two area-velocity probes was used to measure flow.  Each sensor measured the depth of flow above the bottom of the flow channel and the depth-averaged velocity. One sensor was placed at the flow channel centerline and the other was placed approximately 1 foot from the eastern tunnel wall (Figure 4.4.4-1).  Data collected over the one-month monitoring period was used to calculate dry-weather flows through the bypass tunnels.  Detailed discussion of the flow monitoring results is presented in Section 5.4 and Appendix H.       Drop Shafts 14 and 14A

As mentioned previously, the flow-monitoring program also included a component for performing visual spot-checks of flow at drop shafts 14 and 14A.  During the tunnel reconnaissance, inflow from an apparent industrial discharge was observed at Drop Shaft 14.  Approximately one inch of flow was observed discharging from the 24-inch pipe.  The flow monitoring program involved using a 5-gallon bucket and stop watch to measure discharge rates into the FST.  The flow-monitoring program also included performing visual spot checks for any possible flow in the FST at Drop Shaft 14A (the most upstream location included in this program).  These spot-checks were performed approximately twice weekly.  Results of these spot-checks are discussed in Section 5.4.       Rain Gauge Installation

The overall purpose of the monitoring program was to estimate groundwater infiltration, which is a long-term phenomenon independent of peak flows that are produced by surface water runoff.  To distinguish between dry-weather and wet-weather flows, a rain gauge was installed at the Project site.  The gauge was mounted on the roof of Pump Station A, located in the area between the NYPA conduits, the FST and the SSI.  Rainfall data gathered by the gauge was used to identify wet-weather flow regimes in the monitored sewer system.  As mentioned above, wet-weather flows were excluded from the analysis of groundwater infiltration.

4.4.5        Field Activities

Field activities performed as part of this flow-monitoring program included tunnel reconnaissance (discussed above), flow meter installation, routine data downloading, flow meter maintenance, and flow meter removal.       Flow Meter Installation

URS installed three flow monitoring meters on October 20, 2003.  As discussed previously, flow metering equipment was installed at three locations; the measuring weir at drop shaft 12, the west SSI bypass tunnel, and the east SSI bypass tunnel.

At the measuring weir, the depth probe was securely attached to the weir plate using a vise clamp with a steel rod extending approximately five feet upstream of the weir (Figure 4.4.5-1).  The depth probe was mounted on a vertical cross-bar attached to the steel rod and extending downward into the water. The depth probe was installed at a depth of approximately 5.5 inches.  This depth was calibrated using the Insight software to read a measurement of 1.25 inches of flow over the weir crest.  The depth probe and flow meter were connected using a 50-foot cable.  To facilitate downloading of data, the flow meter was mounted on a steel hook located just below ground surface on the inside rim of the manhole.

For both the east and west SSI bypass tunnel flow meter installations, the steel plate assemblies were installed first.  To allow for access through the 24-inch manholes, the steel plates were fabricated in four pieces, each measuring approximately 22 inches wide by 60 inches long.  The steel plates were lowered into the tunnels one piece at a time and assembled at the appropriate location in the tunnel.  The plates were designed to fit together smoothly using tongue-and-groove construction.  The pieces were then secured together using 3/8-inch U-shaped rods inserted through holes on either side of the joints.  The flow measuring assemblies were designed so that the dual area-velocity probes were mounted on the downstream edge of the platform formed by the steel plates.  For both the east and west tunnels, the probes were installed with one located at the centerline of the flow area and the other located approximately one foot from the eastern edge of the steel plates.  The probes were installed within the tunnel flow approximately 96 feet downstream (in the west tunnel) and 84 feet downstream (in the east tunnel) of the intersection with the FST (Figure 4.4.1-1).  Once the platforms were in place, they were checked to ensure a level surface and then each plate was secured to the bypass gate structure using a 1/8-inch stainless steel cable.  The area-velocity probe data cables were run from the probes to the flow meters using custom made 175-foot cables.  The flow meters were suspended from the manhole ladder rungs just beneath the ground surface.  At the time of installation, depth readings recorded by the area-velocity probes were calibrated by comparing the readings to depths measured by hand. 

For the east bypass, several sandbags were placed in the tunnel to redistribute flow from the eastern edge toward the center of the channel.       Downloading and Maintenance

During the flow measuring program, data downloads were performed approximately every three days.  Data was downloaded by accessing the flow meters mounted just below each manhole, and connecting a laptop computer using the appropriate cable.  For each downloading event, the following steps were followed:

·        Remove manhole covers and access flow meters,

·        Confirm proper connection and communication with flow meter,

·        Check the current status and record all parameters in the field notebook,

·        Initiate data download,

·        Confirm proper download, and

·        Return flow meter to manhole and replace cover.

A planned tunnel entrance to inspect the condition of the flow meter installations was conducted on October 30, 2003.  This was done to confirm satisfactory condition of the flow meters and to ensure collection of useful data through the transition from tourist to non-tourist season on November 1, 2003.  Entries were made at all three flow monitoring locations.  Activities completed during these inspections included checking the level of the steel plates, checking for and clearing any sediment accumulation on the steel plates and probes, and checking the depth calibration of each probe.

To provide data for future analysis, a velocity profile was also obtained from collected data in each bypass tunnel. The velocity profile was obtained by detaching the center area-velocity probe and recording velocity and depth readings for different locations across the cross-sectional flow channel.  The probe was placed at 6-inch intervals across the back (downstream side) of the steel platform over the entire 60-inch length.

On November 4, 2003, a check of the most recently downloaded data indicated that the flow meters at the east and west bypass tunnels had stopped recording data due to the memory slates being full.  The dataloggers at all three monitoring locations were reset to record data in “wrap” mode rather than “slate” mode.  In slate mode, the datalogger will stop recording data upon reaching a full memory condition. In wrap mode, when the datalogger memory becomes full, the datalogger will begin overwriting previously recorded data.  As a result of the memory filling up and the halt in data recording, there is a data gap of approximately 35 hours for the east and west bypass locations from approximately November 3 to November 4, when the error was discovered.

Once the dataloggers were reset, data for the west bypass tunnel indicated that flow was still not being recorded at this location.  A review of the data revealed that a flow reversal had occurred just prior to the loss of flow signal strength.  This indicated that sediment may have been deposited on the steel platform and the area-velocity probes when the flow reversal had occurred.  To inspect and correct the problem, an unplanned tunnel entrance was completed on November 6.  Upon inspection of the steel platform and probes, it was revealed that significant quantities of sediment had covered the plate and probes.  Once the sediment had been cleared off, signal strength and flow readings returned.       Flow Meter Removal

Satisfactory completion of data collection was determined to have occurred by November 20.  All flow monitoring equipment, including sandbags, steel plates, and related hardware, was therefore removed from the tunnels on November 25, 2003.

4.5         Water Sampling

Three groundwater/surface water sampling events were performed in 2003/2004.  The first sampling event was conducted from September 24 through October 15, 2003 (two surface water sample locations were partially resampled on October 28, 2003), the second from November 24 through December 19, 2003 and the third from February 24 through March 10, 2004.  Sampling was carried out in general accordance with the protocols and requirements of the Work Plan – Groundwater and Surface Water Monitoring Program (URS 2003a) and the Health and Safety Plan, Groundwater Investigation Activities (URS 2003b).  The goal for each sampling event was the collection of groundwater samples from each of the 91 piezometers within 17 nested groundwater monitoring locations (GW03-001 through GW03-017, installed as part of the NYPA relicensing groundwater investigation), and the collection of surface water samples from 11 locations in and around the NPP.  For various reasons, samples could not be collected at all of the locations during either sampling event.  Sample collection information for the September-October, November-December and February-March sampling events is summarized on Tables 4.5-1 ,4.5-2, 4.5-3, respectively.

4.5.1        Surface Water       Locations

The surface water study targeted eleven locations (see Figure 4.2.1-1) for sampling and analysis.  They are:

·        SW03-001: Fish Creek, north and downstream of the Lewiston Reservoir.

·        SW03-002: Fish Creek, east and upstream of the Lewiston Reservoir

·        SW03-003: Gill Creek, east and upstream of the Lewiston Reservoir

·        SW03-004: Gill Creek headwater wetland area, east and upstream of both the Lewiston Reservoir and SW03-003.

·        SW03-005: Gill Creek, south and downstream of the Lewiston Reservoir and upstream of the Gill Creek augmentation outflow from the Reservoir.

·        SW03-006: Lewiston Reservoir, west end near the Lewiston Pump Generating Plant

·        SW03-007: Forebay, west end at the Robert Moses Niagara Power Plant

·        SW03-008: Western Conduit (Conduit 2), collected from the west air intake shaft (located approximately 450 feet north of Royal Avenue)

·        SW03-009: Eastern Conduit (Conduit 1), collected from the east air intake shaft (located approximately 450 feet north of Royal Avenue)

·        SW03-010: Niagara River at Project intakes

·        SW03-011: Eastern end of the Lewiston Reservoir.

These locations are shown on Figure 4.2.1-1.  Since, however, no water was flowing at locations SW03-002 and SW3-003 during the September-October sampling event, or at location SW3-003 during the November-December sampling event (i.e., only standing, stagnant water was present at these locations), no samples were collected at these locations (see Tables 4.5-1 and 4.5-2).

The October 2003 event, which was conducted from October 7 through October 9, 2003, included collection of samples at nine locations.  No samples were collected at SW03-003 (Gill Creek) and SW03-002 (Fish Creek) since these locations were “dry” (no flow with small amounts of stagnant water in shallow puddles).  Due to a sample processing/preparation error at the laboratory, two locations (SW03-001 and SW03-004) were re-sampled for monomethyl mercury on October 28, 2003.

The November 2003 event, conducted on November 24 and 25, 2003, included collection of samples at ten locations.  Location SW03-003, upstream of the Lewiston Reservoir in Gill Creek was “dry”.

Surface water samples were collected from all 11 locations during the March 2004 sampling event.

For each event, samples were analyzed for VOCs, SVOCs, metals, pesticides, PCBs, bacterial parameters, and miscellaneous water quality parameters as discussed in Section and summarized on Table 4.5-4.  For purposes of discussion, the surface water locations were divided into two groups, “river-sourced” water (water derived directly from the Niagara River as part of Project operations) and “local-sourced” water (surface water present in wetlands and streams near the reservoir).  The “river-sourced” locations include (in order of distance from river): SW03-010, SW03-009, SW03-008, SW03-007, SW03-006 and SW03-011.  The remaining “local-sourced” locations included (in order of distance away from the reservoir); SW03-004, SW03-003, SW03-002, and SW03-001.       Sampling Methods

In accordance with procedures outlined in the Field Sampling Plan (FSP) portion of the Work Plan (URS 2003a), most surface water samples were collected by direct submersion of the sample bottles and capping underwater.  The forebay sample (SW03-007), however, due to forebay surface water inaccessibility, was collected from the forebay deck of the RMNPP with a disposable polyethylene bailer with nylon rope.  Bailer sampling had not been addressed in the FSP.  Samples from the 2 conduits at the pressure relief vents located near Royal Avenue (SW03-008 and SW03-009) were collected through the vent grates using a peristaltic pump with high-density polyethylene and silicone (rotor head) tubing.  The use of peristaltic pumps for sample collection had been described in the FSP for groundwater sample collection, but not for surface water sample collection.       Parameters

Following sample collection, the sample bottles were placed in coolers, iced, and transported via courier to Severn Trent Laboratories of Buffalo, New York, for analysis.  The surface water samples were analyzed for an extensive list of organic, inorganic, and bacteriological parameters.  Parameters and analytical methods are summarized on Table 4.5-4.

4.5.2        Groundwater       Locations

As part of this investigation, it was proposed that groundwater samples be collected at each of the 91 piezometers located among 17 nested groundwater monitoring points (GW03-001 through GW03-017).  The 17 nested groundwater monitoring well locations are shown in Figure 4.2.1-1.  Monitoring locations GW03-001 through GW03-011 are located in the vicinity of the Lewiston Reservoir, and monitoring locations GW03-012 through GW03-017 are located adjacent to the NYPA conduits, along their eastern side.  Multiple piezometers, screened at discrete depth intervals, are present at each monitoring location.  Nested well construction details are provided in Appendix C.

During the September-October sampling event, seven piezometers could not be sampled because (1) they were initially dry (GW03-012A-P1 and GW03-012B-P4), (2) they purged to dryness and never recovered (GW-03-002A-P2, GW03-006B-P3, GW03-011B-P5, and GW03-012C-P7), or (3) they had obstructed risers that did not permit sampling with available equipment (GW03-017A-P3) (see Table 4.5-1.  At six piezometers (GW03-005A-P2, GW03-010A-P3, GW03-013B-P6, GW03-014A-P1, GW03-014A-P2, and GW03-015A-P3), due to poor groundwater yields, only partial sample volumes could be collected (see Table 4.5-1).  Complete sample bottle sets were collected from the other 78 piezometers.

During the November-December sampling event, six piezometers could not be sampled either because they were initially dry (GW03-012A-P1) or they purged to dryness and never recovered (GW03-001A-P2, GW03-006B-P3, GW03-008B-P6, GW03-011B-P5, and GW03-013B-P6) (see Table 4.5-2).  Due to poor groundwater yields, only partial sample volumes could be collected at four piezometers (GW03-002A-P2, GW03-012B-P4, GW03-014A-P2, and GW03-017A-P3) (see Table 4.5-2).  Complete sample bottle sets were collected from the other 81 piezometers.

During the February-March 04 sampling event, eight piezometers could not be sampled either because they were initially dry (GW03-012-P1 and GW03-012-P4) or an insufficient volume of water was present (GW03-002-P2, GW03-006-P3, GW03-008-P6, GW03-011-P5, and GW03-017-P3) or they purged to dryness and never recovered (GW03-004-P1).  Due to poor groundwater yields, only partial sample volumes could be collected at three piezometers (GW03-013-P6, GW03-013-P7, and GW03-015-P3) (see Table 4.5-3).       Sampling Methods

Detailed, step-by-step procedures for the collection of groundwater samples for this project were provided in the FSP.  The following paragraphs provide a general discussion of sampling procedures, noting any deviations from the FSP.

Groundwater samples were collected from the piezometers using dedicated and/or disposable sampling equipment.  Low-flow purging and sampling techniques, in accordance with USEPA Region II recommended procedures (USEPA 1998), were used to obtain representative groundwater samples and to minimize purge water volumes requiring containerization, characterization, and subsequent disposal.  At piezometer locations where depth to groundwater was less than approximately 30 feet below ground surface, purging and groundwater sampling were accomplished using GeoPump 2 peristaltic pumps with dedicated, disposable high-density polyethylene (HDPE) and silicone (rotor head only) tubing.  Tubing associated with the peristaltic pumps was removed at the completion of sampling and disposed of.

At piezometers where depth to groundwater was greater than approximately 30 feet below ground surface, purging and groundwater sampling were accomplished using dedicated Proactive Environmental Products submersible pumps with low-flow controllers and dedicated HDPE tubing.  The dedicated submersible pumps and associated tubing were left in the piezometers following sampling.  The inlet of the submersible pump, or HDPE tubing for the peristaltic pump, was set at the midpoint of the saturated portion of the piezometer screen.

Flow from the purging/sampling pump was routed to a Horiba U-22 multiparameter meter with flow-through cell.  Upon initiation of piezometer purging, the time, pump flow rate, depth to water, and field parameter measurements (pH, temperature, specific conductivity, dissolved oxygen, turbidity, and Eh) were recorded.  The Low Flow Groundwater Purging/ Sampling Log sheets are provided in Appendix I.  Purge water was containerized and later transported to the field staging area for storage and subsequent characterization and disposal.  Once field parameter readings had stabilized within specified tolerances and at least one well volume had been purged, the tubing between the pump and flow cell was cut and the sample bottles were filled from the tubing.  In a few cases, the “purge at least one well volume” criterion was waived for deeper piezometers where field parameters had stabilized but the time necessary to purge one well volume was deemed to be excessive.

During all three sampling events, several piezometers could not be readily sampled with either of the pump types specified in the FSP.  This was due to limited water volume within the piezometer (usually less than 3 feet), poor recharge characteristics (i.e., known to purge to dryness), extreme depths to groundwater (beyond the pumping range of submersible pumps), or a combination of these conditions.  These piezometers were sampled with dedicated, disposable polyethylene bailers with nylon rope (bailer sampling had not been specified in the FSP).  Generally, the piezometers were purged to dryness with the bailer, allowed to recover, and the recovered groundwater was later sampled with the bailer.  During the September-October sampling event, three piezometers were sampled with bailers: GW03-013C-P7, GW03-014A-P2, and GW03-015A-P3.  During the November-December sampling event, six piezometers were sampled with bailers: GW03-002A-P2, GW03-005A-P2, GW03-012B-P4, GW03-013C-P7, GW03-014A-P2, and GW03-015A-P3.  During the February-March 04 sampling event, seven piezometers were sampled with bailers: GW03-005A-P2, GW03-012C-P7, GW03-013B-P6, GW03-013C-P7, GW03-014A-P2, GW03-015A-P2 and GW03-015A-P3

During the September-October sampling event, several methods were attempted to collect a sample from piezometer GW03-017A-P3 (i.e., submersible pump, 1.6-inch OD disposable bailer, and 0.75-inch OD disposable bailer).  Due to an obstruction in the piezometer riser pipe located above the water in the piezometer (preventing use of the submersible pump and the larger disposable bailer), and due also to the extreme depth to water in this piezometer (the water leaked out of the smaller bailer before it could be pulled to the surface), none of these methods was successful.  During the November-December sampling event, an alternative sampling method not specified in the FSP (manual pumping using HDPE tubing with decontaminated stainless-steel check valve) was employed to attempt sampling at GW03-017A-P3.  The yield from this piezometer was poor, however, and only limited sample volume could be collected.  This same approach was attempted during the February-March sampling event, however, not enough water was present to bring sample to the surface.       Parameters

Following sample collection, the sample bottles were placed in coolers, iced, and transported via courier to Severn Trent Laboratories of Buffalo, New York, for analysis.  The groundwater samples were analyzed for an extensive list of organic, inorganic, and bacteriological parameters.  Parameters and analytical methods are summarized on Table 4.5-4.  Based upon elevated total coliform levels in the September-October 2003 sampling event, the December 2003 groundwater samples collected from piezometers GW03-005A-P2, GW03-014C-P7, GW03-015A-P1, GW03-015B-P6, and GW03-017C-P7 were also analyzed for fecal coliform.  In addition, E. coli, which was not analyzed for in the September-October 2003 sampling event, was tested in the subsequent December 2003 and February-March 2004 sampling events.  Piezometer locations where only partial sample volumes could be collected for analysis due to poor groundwater yields were noted in the “Comments” columns of Tables 4.5-1 , 4.5-2 and, 4.5-3.

4.5.3        Quality Control Samples

Quality control (QC) procedures during the groundwater/surface water sampling program included use of trip blanks, matrix spike (MS)/matrix spike duplicate (MSD)/matrix duplicate (MS) samples, and sampling equipment rinsate blanks.  These QC sample types and their collection frequencies were detailed in the Quality Assurance Project Plan (QAPP) portion of (URS 2003a).  Some variations from the QAPP occurred, however, during the sampling events.

The QAPP had specified that one trip blank per day accompany the volatile organic analysis (VOA) sample bottles from the field to the analytical laboratory.  However, multiple sampling crews were working at different locations and samples were being shipped to the laboratory twice per day due to bacteriological parameter method holding times.  Trip blanks therefore accompanied each separate sampling crew’s VOA bottle shipments to the laboratory (up to 6 separate trip blanks per day).

MS/MSD/MD samples were collected at the rate of 1 per 20 groundwater or surface water samples collected, as was specified in the QAPP.  Three equipment rinsate blanks were collected during the September-October sampling event (peristaltic pump and tubing, submersible pump and tubing, and disposable bailer) and two equipment rinsate blanks were collected during the November-December and February-March sampling events (peristaltic pump and tubing, and disposable bailer).  Since the dedicated submersible pumps and tubing were left in the piezometers following the first sampling event, a submersible pump rinsate blank was not collected for the second and third sampling events.

4.6         Well GW03-015 Chloride Monitoring

Based on groundwater sample analytical results from the groundwater monitoring program (as discussed in Sections and, additional study was performed to focus on the seasonal (tourist/non-tourist season) nature of groundwater flow conditions and the effect seasonal variations may have on groundwater flow and contaminant movement in the vicinity of the conduits and well GW03-015.

To assess the degree of change in water quality caused by daily, operationally induced changes in water levels in the CDS, continuous recording ion specific probes with pressure transducers were installed in each piezometer at monitoring location GW03-015.   The probes were deployed between March 25 and April 6, 2004, to monitor the period of one week before the change over from non-tourist to tourist season and one week after.   Details of the study are discussed in the following sections.

4.6.1        Probe Calibration

On March 24, 2004, seven In-Situ® Multi-Parameter (MP) Troll 9000 Series® probes were received.  Each troll was equipped with the capability of measuring water quality parameters including, conductivity, oxidation/reduction potential (ORP), pH, barometric pressure, dissolved oxygen (DO), temperature, and feet of water (height of water column), and chloride concentration.

To ensure proper data collection and quality control, the probes were calibrated prior to deployment.  Calibration solution supplied by In-Situ® was used to calibrate probes for DO, conductivity, pH and ORP.  According to manufacturers specifications, recommended recalibration frequencies range from 2-4 weeks (for DO) to 2-3 months (for conductivity).  Since the trolls were only deployed for a two week period, this calibration was only performed once.  The chloride probe calibration was required daily. 

A more intensive ‘three-point calibration” for the chloride probe was performed once prior to troll deployment.  The three-point calibration utilized two different chloride concentration solutions, which included 355 parts per million (ppm) chloride and 3545 ppm chloride. The three point calibration procedure consisted of the following,

·        Calibrated probe in 355 ppm solution at room temperature until stable readings were obtained.

·        Calibrated probe in 3545 ppm solution at room temperature until stable readings were obtained.

·        Calibrated probe in 3545 ppm solution in an ice bath.  Chloride solution was chilled in ice bath for at least one hour prior to calibration.  Readings were collected until steady and temperature readings were near 0° Celsius (=+ 0.4° Celsius). 

Following the three-point and quick-cal calibrations, each probe was fitted with a small battery operated DO stirrer supplied by In-Situ®.  The purpose of the stirrers was to gently agitate groundwater (similar to laboratory analysis) for more accurate DO readings.  The trolls were deployed in each piezometer at well location GW03-015 on March 26, 2004.  Each troll was programmed to collect data every 15 minutes.  The trolls were placed on average approximately 2-3 feet from the bottom of the piezometer, except piezometer P3.  Due to an insufficient volume of water in P3 (approximately 4-5 feet of water column), the troll was placed directly on the bottom of the well.  However, the design of the DO stirrer unit (which is located directly below the probes) is such that approximately one foot of space is between the DO stirrer and the bottom of the troll.  Because of this, the probes were still located approximately one foot above the bottom of the well.

4.6.2        Field Activities

As mentioned above, the trolls collected data from March 25 through April 6, 2004.  Site visits were performed on a daily basis (including weekends).  Activities performed during each site visit included the following,

·        Upon arrival, static water levels and height of water column readings from trolls were collected simultaneously in order to calibrate troll positions in the piezometers.  These procedures are identical to those described in Sections 4.3.1 and 4.3.3.

·        All data collected between site visits was extracted and downloaded to a computer and saved.  Data was collected by the dataloggers every 15 minutes.

·        The trolls were then removed from each piezometer and the chloride probes were calibrated with a “one-point” calibration.  The difference between the one-point and the three-point calibration is that only one solution at ambient air temperature is used for the one-point calibration.  Based on previous groundwater sample analytical data, probes installed in piezometers P1, P4, P5, P6, and P7 were calibrated with the lower concentration of 355 ppm chloride solution.  Probes installed in piezometers P2 and P3 were calibrated with the higher concentration of 3545 ppm chloride solution. 

·        Following chloride calibration, the trolls were reinstalled into each piezometer and a new test was started.  A second water level/troll position calibration was performed as outlined above. 

·        The trolls were removed on April 6, 2004.       Tourist/Non-Tourist Water Quality

Since location GW03-015 exhibited a change toward greater salinity in the samples collected between fall and winter sampling events, a limited groundwater sampling event was conducted on April 12 and 13, 2004, following the change over to tourist season operational mode and removal of MP 9000 Trolls®. 

Groundwater samples were collected following guidelines outlined in Section  Sample analytical parameters were similar to those described in Section and Table 4.5-4. 

4.6.3        Data Management

Electronic water level data quality assurance measures were performed in accordance with those described in Section 4.3.1. After these calculations were completed, the data was imported into a database. Further quality assurance was completed by graphing each piezometer and comparing to the original spreadsheet and field notes to determine whether reference elevations and depth to water calculations were correct.

Table 4.1.1-1

Parameters Estimated during Calibration, Calibrated USGS 1996 Model


Transmissivity [ft2/d]


Weathered bedrock



horizontal fracture zones


Vertical Hydraulic Conductivity [ft/d]


glacial sediments



weathered bedrock



unweathered bedrock


Average Recharge Rate [ft/d]


urban areas



rural areas


Source: Yager 1996, Table 5


Table 4.1.1-2

Comparison of Measured and Estimated Flow Rates, Calibrated USGS 1996 Model

Discharge Area

Observed Flow [ft3/d]

Calibrated Model Flow [ft3/d]

Crossing of Falls Street Tunnel and NYPA conduits



Remainder of Falls Street Tunnel1



Cayuga Creek2



Bergholtz Creek



Redland Quarry



Perennial springs



1 Does not include flow entering airshaft near 18th Street.  In the 1990’s flow was reduced to less than 0.1 mgd at this location (Pers. Comm. NFWB, 2004).

2 Does not include 70,000 ft3/d entering creek from sewer lift-station near Lockport Road.

Source: Simulated Three-Dimensional Ground-Water Flow in the Lockport Group, A Fractured–Dolomite Aquifer Near Niagara Falls, New York (Yager 1996), Table 7.


Table 4.1.1-3

Water Budget, Calibrated USGS 1996 Model


Rate [ft3/d]

Percentage of Total [%]




Lewiston Reservoir



Niagara River










Rate [ft3/d]

Percentage of Total [%]

Natural Areas

Niagara River Gorge






Niagara River



Niagara Escarpment



Manmade Structures

Falls Street Tunnel at NYPA Conduits

450,000 (3.36 mgd)


Industrial Wells









Source: Yager 1996, Table 6


Table 4.1.2-1

Comparison of Flow Budgets, GMS Model and 2003 USGS Model


Flow Rate 2003 USGS Model1 [ft3/d]

GMS Model2 [ft3/d]

Discrepancy [%]

Flow In:





Constant head




Wells *








Head dependent boundaries








Flow Out:





Constant head












Head dependent boundaries








Maximum discrepancy = 0.056%

1 Source: output file “modflow.out” of the 2003 USGS model.

2 Source: output file of the converted GMS model.

* Injection wells used to simulate the inflow of groundwater from the outside of the model domain.


Table 4.1.2-2

Comparison of Hydraulic Heads at Monitoring Locations, GMS Model and 2003 USGS Model

Observation Point

Observed Head [ft]

Calculated Head

Difference USGS GMS [ft]

USGS Model [ft]

GMS Model [ft]

















































































































































































Table 4.1.2-2 (CONT.)

Comparison of Hydraulic Heads at Monitoring Locations, GMS Model and 2003 USGS Model

Observation Point

Observed Head [ft]

Calculated Head

Difference USGS GMS [ft]

USGS Model [ft]

GMS Model [ft]






















































































































































































Table 4.1.2-2 (CONT.)

Comparison of Hydraulic Heads at Monitoring Locations, GMS Model and 2003 USGS Model

Observation Point

Observed Head [ft]

Calculated Head

Difference USGS GMS [ft]

USGS Model [ft]

GMS Model [ft]