Hydrogeologic Framework, Groundwater Quality, and Simulation of Groundwater Flow at the Fair Lawn Well Field Superfund Site, Bergen County, New Jersey
By Jean C. Lewis-Brown, Donald E. Rice, Robert Rosman, and Nicholas P. Smith
MODEL VERSION/TYPE: MODFLOW-2000, steady state
AREA STUDIED: Bergen County
AQUIFERS SIMULATED: Passaic Formation, glacial deposits and alluvium
MOST RECENT WITHDRAWALS SIMULATED: 2000
MODEL SIZE: 41 layers, 329 rows, 248 columns
MINIMUM GRID SPACING: 25 feet x 25 feet
MODEL ARCHIVE AVAILABLE ON REQUEST: gs-nj-model-request@usgs.gov
REPORT ABSTRACT
Production wells in the Westmoreland well field, Fair Lawn, Bergen County, New Jersey (the "Fair Lawn well field Superfund site"), are contaminated with volatile organic compounds, particularly trichloroethylene, tetrachloroethylene, and 1,1,1-trichloroethane. In 1983, the U.S. Environmental Protection Agency (USEPA) placed the Westmoreland well field on its National Priority List of Superfund sites. In an effort to determine groundwater flow directions, contaminant-plume boundaries, and contributing areas to production wells in Fair Lawn, and to evaluate the effect of present pump-and-treat systems on flowpaths of contaminated groundwater, the U.S. Geological Survey (USGS), in cooperation with the USEPA, developed a conceptual hydrogeologic framework and groundwater flow model of the study area. MODFLOW-2000, the USGS three-dimensional finite-difference model, was used to delineate contributing areas to production wells in Fair Lawn and to compute flowpaths of contaminated groundwater from three potential contaminant sources to the Westmoreland well field. Straddle-packer tests were used to determine the hydrologic framework of, distribution of contaminants in, and hydrologic properties of water-bearing and confining units that make up the fractured-rock aquifer underlying the study area. The study area consists of about 15 square miles in and near Fair Lawn. The area is underlain by 6 to 100 feet of glacial deposits and alluvium that, in turn, are underlain by the Passaic Formation. In the study area, the Passaic Formation consists of brownish-red pebble conglomerate, medium- to coarse-grained feldspathic sandstone, and micaceous siltstone. The bedrock strata strike N. 9° E. and dip 6.5° to the northwest. The bedrock consists of alternating layers of densely fractured rocks and sparsely fractured rocks, forming a fractured-rock aquifer.
Groundwater flow in the fractured-rock aquifer is anisotropic as a result of the interlayering of dipping water-bearing and confining units. Wells of similar depth aligned along the strike of the bedding intersect the same water-bearing units, but wells aligned along the dip of the bedding may intersect different water-bearing units. Consequently, wells aligned along strike are in greater hydraulic connection than wells aligned along dip.
The Borough of Fair Lawn pumps approximately 770 million gallons per year from 13 production wells. Hydrographs from six observation wells ranging in depth from 162 to 505 feet in Fair Lawn show that water levels in much of the study area are affected by pumping. Straddle packers were used to isolate discrete intervals within six open-hole observation wells owned by the Fair Lawn Water Department. Transmissivity, water-quality, and static-water-level data were obtained from the isolated intervals. Measured transmissivity ranged from near 0 to 8,900 feet squared per day. The broad range in measured transmissivity is a result of the heterogeneity of the fractured-rock aquifer.
Eight water-bearing units and eight confining units were identified in the study area on the basis of transmissivity. The water-bearing units range in thickness from 21 to 95 feet; the mean thickness is 50 feet. The confining units range in thickness from 22 to 248 feet; the mean thickness is 83 feet. Water-level and water-quality data indicate effective separation of water-bearing units by the confining units.
Water-quality samples were collected from the six observation wells at 16 depth intervals isolated by the straddle packers in 2000 and 2001. Concentrations of volatile organic compounds generally were low in samples from four of the wells, but were higher in samples from a well in Fair Lawn Industrial Park and in a well in the Westmoreland well field.
The digital groundwater flow model was used to simulate steady-state scenarios representing conditions in the study area in 1991 and 2000. These years were chosen because during the intervening period, pumpage from the Westmoreland well field decreased by more than one-half, and a system of shallow wells (less than 19 feet deep) and trenches was installed at one of the contaminant sources to capture shallow groundwater. Because precipitation was below average in 2000, a "high-recharge" scenario also was simulated to represent 2000 pumpage conditions during more typical groundwater-recharge conditions.
The digital model was used to delineate contributing areas to production well fields in Fair Lawn and contaminant plumes from three contaminant sources. Two of these sources, Fisher Scientific Company and Sandvik, Inc., are known contaminated sites that were previously identified by the N.J. Department of Environmental Protection as potential sources of volatile organic compounds to the Westmoreland well field. The third source, Eastman Kodak Company, is a known contaminated site not previously identified as a potential source of volatile organic compounds to the Westmoreland well field. In 1991, when 130 million gallons of water was pumped from the Westmoreland well field, the contributing area to that well field included nearly all of the Fisher Scientific Company property, a small part (less than 1 percent) of the Sandvik, Inc., property, and about three-quarters of the Eastman Kodak Company property. In 2000, when only 40 million gallons of water was pumped from the Westmoreland well field, the contributing area included only small parts of the Fisher Scientific Company and Eastman Kodak Company properties, and none of the Sandvik, Inc., property. In the 2000 high-recharge scenario, contributing areas to every pumped well were similar to, but smaller than, the contributing areas in the 2000 scenario.
Contaminant plumes from the three potential contaminant sites were delineated using the groundwater flow model. In 1991, most of the plume from the overburden at Fisher Scientific Company discharged to a well in the Westmoreland well field; about half of the plume from the bedrock discharged to deep recovery wells at Fisher Scientific Company and half discharged to two wells in the Westmoreland well field. Only 3 percent of the water from the overburden and 4 percent of the water from the bedrock was not captured by any well. In 1991, 3 percent of the plume from the overburden at Sandvik, Inc., discharged to wells in the Westmoreland well field; 93 percent of the plume from the bedrock discharged to wells in that well field. Nearly all (97 percent) of the water from the overburden and 7 percent of the water from the bedrock at Sandvik, Inc., was not captured by any well and flowed instead to Henderson Brook. In 1991, 73 percent of the plume from the bedrock at Eastman Kodak Company discharged to a well in the Westmoreland well field; the remainder flowed to the Passaic River. No plume from the overburden at Eastman Kodak Company was delineated because simulation results indicate that the overburden was unsaturated.
In 2000, wells in the Westmoreland well field captured less water than in 1991 from all three sites because of the decreased pumpage. Only 6 percent of the plume from the overburden at Fisher Scientific Company discharged to a well in the Westmoreland well field; more than half (55 percent) was captured by the recently installed shallow recovery system and 4 percent was captured by the deep recovery wells at the site. Thirty-four percent of the plume from the bedrock discharged to a well in the Westmoreland well field, and 58 percent was captured by the deep recovery wells at the site. In this simulation, 35 percent of the water from the overburden and 9 percent of the water from the bedrock at Fisher Scientific Company flowed to Henderson Brook and the Passaic River rather than to any well.
In 2000, all of the water originating in the overburden at Sandvik, Inc., discharged to Henderson Brook; none was captured by any well. Twenty-three percent of the plume from the bedrock at Sandvik, Inc., discharged to two wells in the Westmoreland well field; 2 percent discharged to a deep recovery well at Fisher Scientific Company, and the remainder (74 percent) flowed to Henderson Brook and the Passaic River rather than to any well.
In 2000, 9 percent of the water from the bedrock at Eastman Kodak Company discharged to a well in the Westmoreland well field; the remainder discharged to Henderson Brook and the Passaic River.
Hydrogeologic Framework, Groundwater Quality, and Simulation of Groundwater Flow at the Fair Lawn Well Field Superfund Site, Bergen County, New Jersey
By Jean C. Lewis-Brown, Donald E. Rice, Robert Rosman, and Nicholas P. Smith
MODEL VERSION/TYPE: MODFLOW-2000, steady state
AREA STUDIED: Bergen County
AQUIFERS SIMULATED: Passaic Formation, glacial deposits and alluvium
MOST RECENT WITHDRAWALS SIMULATED: 2000
MODEL SIZE: 41 layers, 329 rows, 248 columns
MINIMUM GRID SPACING: 25 feet x 25 feet
MODEL ARCHIVE AVAILABLE ON REQUEST: gs-nj-model-request@usgs.gov
REPORT ABSTRACT
Production wells in the Westmoreland well field, Fair Lawn, Bergen County, New Jersey (the "Fair Lawn well field Superfund site"), are contaminated with volatile organic compounds, particularly trichloroethylene, tetrachloroethylene, and 1,1,1-trichloroethane. In 1983, the U.S. Environmental Protection Agency (USEPA) placed the Westmoreland well field on its National Priority List of Superfund sites. In an effort to determine groundwater flow directions, contaminant-plume boundaries, and contributing areas to production wells in Fair Lawn, and to evaluate the effect of present pump-and-treat systems on flowpaths of contaminated groundwater, the U.S. Geological Survey (USGS), in cooperation with the USEPA, developed a conceptual hydrogeologic framework and groundwater flow model of the study area. MODFLOW-2000, the USGS three-dimensional finite-difference model, was used to delineate contributing areas to production wells in Fair Lawn and to compute flowpaths of contaminated groundwater from three potential contaminant sources to the Westmoreland well field. Straddle-packer tests were used to determine the hydrologic framework of, distribution of contaminants in, and hydrologic properties of water-bearing and confining units that make up the fractured-rock aquifer underlying the study area. The study area consists of about 15 square miles in and near Fair Lawn. The area is underlain by 6 to 100 feet of glacial deposits and alluvium that, in turn, are underlain by the Passaic Formation. In the study area, the Passaic Formation consists of brownish-red pebble conglomerate, medium- to coarse-grained feldspathic sandstone, and micaceous siltstone. The bedrock strata strike N. 9° E. and dip 6.5° to the northwest. The bedrock consists of alternating layers of densely fractured rocks and sparsely fractured rocks, forming a fractured-rock aquifer.
Groundwater flow in the fractured-rock aquifer is anisotropic as a result of the interlayering of dipping water-bearing and confining units. Wells of similar depth aligned along the strike of the bedding intersect the same water-bearing units, but wells aligned along the dip of the bedding may intersect different water-bearing units. Consequently, wells aligned along strike are in greater hydraulic connection than wells aligned along dip.
The Borough of Fair Lawn pumps approximately 770 million gallons per year from 13 production wells. Hydrographs from six observation wells ranging in depth from 162 to 505 feet in Fair Lawn show that water levels in much of the study area are affected by pumping. Straddle packers were used to isolate discrete intervals within six open-hole observation wells owned by the Fair Lawn Water Department. Transmissivity, water-quality, and static-water-level data were obtained from the isolated intervals. Measured transmissivity ranged from near 0 to 8,900 feet squared per day. The broad range in measured transmissivity is a result of the heterogeneity of the fractured-rock aquifer.
Eight water-bearing units and eight confining units were identified in the study area on the basis of transmissivity. The water-bearing units range in thickness from 21 to 95 feet; the mean thickness is 50 feet. The confining units range in thickness from 22 to 248 feet; the mean thickness is 83 feet. Water-level and water-quality data indicate effective separation of water-bearing units by the confining units.
Water-quality samples were collected from the six observation wells at 16 depth intervals isolated by the straddle packers in 2000 and 2001. Concentrations of volatile organic compounds generally were low in samples from four of the wells, but were higher in samples from a well in Fair Lawn Industrial Park and in a well in the Westmoreland well field.
The digital groundwater flow model was used to simulate steady-state scenarios representing conditions in the study area in 1991 and 2000. These years were chosen because during the intervening period, pumpage from the Westmoreland well field decreased by more than one-half, and a system of shallow wells (less than 19 feet deep) and trenches was installed at one of the contaminant sources to capture shallow groundwater. Because precipitation was below average in 2000, a "high-recharge" scenario also was simulated to represent 2000 pumpage conditions during more typical groundwater-recharge conditions.
The digital model was used to delineate contributing areas to production well fields in Fair Lawn and contaminant plumes from three contaminant sources. Two of these sources, Fisher Scientific Company and Sandvik, Inc., are known contaminated sites that were previously identified by the N.J. Department of Environmental Protection as potential sources of volatile organic compounds to the Westmoreland well field. The third source, Eastman Kodak Company, is a known contaminated site not previously identified as a potential source of volatile organic compounds to the Westmoreland well field. In 1991, when 130 million gallons of water was pumped from the Westmoreland well field, the contributing area to that well field included nearly all of the Fisher Scientific Company property, a small part (less than 1 percent) of the Sandvik, Inc., property, and about three-quarters of the Eastman Kodak Company property. In 2000, when only 40 million gallons of water was pumped from the Westmoreland well field, the contributing area included only small parts of the Fisher Scientific Company and Eastman Kodak Company properties, and none of the Sandvik, Inc., property. In the 2000 high-recharge scenario, contributing areas to every pumped well were similar to, but smaller than, the contributing areas in the 2000 scenario.
Contaminant plumes from the three potential contaminant sites were delineated using the groundwater flow model. In 1991, most of the plume from the overburden at Fisher Scientific Company discharged to a well in the Westmoreland well field; about half of the plume from the bedrock discharged to deep recovery wells at Fisher Scientific Company and half discharged to two wells in the Westmoreland well field. Only 3 percent of the water from the overburden and 4 percent of the water from the bedrock was not captured by any well. In 1991, 3 percent of the plume from the overburden at Sandvik, Inc., discharged to wells in the Westmoreland well field; 93 percent of the plume from the bedrock discharged to wells in that well field. Nearly all (97 percent) of the water from the overburden and 7 percent of the water from the bedrock at Sandvik, Inc., was not captured by any well and flowed instead to Henderson Brook. In 1991, 73 percent of the plume from the bedrock at Eastman Kodak Company discharged to a well in the Westmoreland well field; the remainder flowed to the Passaic River. No plume from the overburden at Eastman Kodak Company was delineated because simulation results indicate that the overburden was unsaturated.
In 2000, wells in the Westmoreland well field captured less water than in 1991 from all three sites because of the decreased pumpage. Only 6 percent of the plume from the overburden at Fisher Scientific Company discharged to a well in the Westmoreland well field; more than half (55 percent) was captured by the recently installed shallow recovery system and 4 percent was captured by the deep recovery wells at the site. Thirty-four percent of the plume from the bedrock discharged to a well in the Westmoreland well field, and 58 percent was captured by the deep recovery wells at the site. In this simulation, 35 percent of the water from the overburden and 9 percent of the water from the bedrock at Fisher Scientific Company flowed to Henderson Brook and the Passaic River rather than to any well.
In 2000, all of the water originating in the overburden at Sandvik, Inc., discharged to Henderson Brook; none was captured by any well. Twenty-three percent of the plume from the bedrock at Sandvik, Inc., discharged to two wells in the Westmoreland well field; 2 percent discharged to a deep recovery well at Fisher Scientific Company, and the remainder (74 percent) flowed to Henderson Brook and the Passaic River rather than to any well.
In 2000, 9 percent of the water from the bedrock at Eastman Kodak Company discharged to a well in the Westmoreland well field; the remainder discharged to Henderson Brook and the Passaic River.