This report presents the concentrations of VOCs and PFAS in samples collected in 2018 at the NAWC site. Concentrations of TCE, cis-1,2-dichloroethene (cisDCE), vinyl chloride (VC), PFOS (perfluorooctane sulfonate), PFOA (perfluorooctanoic acid), PFNA (perfluorononanoic acid), and PFBS (perfluorobutane sulfonate) in groundwater and surface-water samples collected during 2018 at NAWC were tabulated and evaluated. The areal and vertical distributions of contamination are presented on maps and cross sections. Concentrations measured in samples collected in 2018 are compared to those reported for the most recent previous sampling date.

The geology of NAWC is described by Lacombe (2000) and Lacombe and Burton (2010). The upper 7–10 feet (ft) of surficial sediments at NAWC are composed primarily of soil and fill material brought in for leveling the site. Bedrock weathering created a layer of unconsolidated saprolite present to depths of 25 to 40 ft below land surface. Below this weathered layer, a fractured sedimentary bedrock aquifer exists, composed primarily of mudstones and siltstones of the Lockatong Formation and sandstones of the Stockton Formation (fig. 1). The Lockatong Formation contains thin, fissile mudstones that appear black from a high organic carbon content, and which are the primary water-bearing units at the site. There are six of these fissile mudstone units mapped at NAWC (fig. 1; appendix 1; Lacombe and Burton, 2010; Fiore and Lacombe, 2020): BlkFis-159, BlkFis-172, BlkFis-190, BlkFis-233, BlkFis-246, and BlkFis-262. Both formations are composed of a series of sloping beds that dip to the northwest and strike northeast-southwest. The Stockton Formation and Lockatong Formation are separated by a fault that strikes about N 65° E and dips at least 70° SW. This fault is a zone of offset discontinuous geologic beds of the Lockatong and Stockton Formations whose attitudes range from flat-lying to overturned. The fault zone is heavily brecciated and is highly weathered, containing considerable amounts of clay that create a low-permeability barrier to groundwater flow between the formations.

Most groundwater flow at NAWC occurs through fractures parallel to the bedding plane along strike, and through dip-aligned fractures between the sloping beds (Lacombe, 2000; Lacombe, 2002; Fiore and Lacombe, 2020). Most of the bedrock deeper than 250 ft is unfractured with a low yield of water (Lacombe and Burton, 2010). When unstressed by the pump-and-treat well pumping, the shallow groundwater flow direction in the unconsolidated overburden and saprolite is generally towards Gold Run Spring located west of NAWC, and to a lesser extent, the West Ditch Spring located within NAWC (Fiore and Lacombe, 2020; appendix 1). Gold Run Spring then flows toward the ancestral west branch of Gold Run (currently culverted beneath Parkway Avenue south of NAWC) and ultimately discharges into the Delaware River (Lacombe, 2000). In the northeastern part of NAWC, Fiore and Lacombe (2020) indicate groundwater in the shallow unconsolidated subsurface may flow off-site toward the east (appendix 1). In the fractured bedrock, the general groundwater flow direction is toward the pump-and-treat wells along bedding plane fractures and dip-aligned fractures, primarily among the fissile mudstone units (Lacombe and Burton, 2010; Fiore and Lacombe, 2020). The main exception is in the eastern part of the Stockton Formation, which flows off-site toward the southeast (appendix 1). The pump-and-treat-system consists of 10 extraction wells; these wells are located throughout the site but are dominantly located in the southwestern part of the site. Pumping of the extraction wells stresses the groundwater gradients and changes the local flow directions at NAWC, thereby acting to hydraulically contain the contaminated groundwater and prevent offsite migration (Lacombe, 2000; Lacombe, 2002; Fiore and Lacombe, 2020).

TCE was initially detected in groundwater at NAWC in the late 1980s (International Technology Corporation, 1994). Subsequent investigations determined that TCE contamination in the groundwater was located in two main areas: (1) Site 1, on the west side of the site, which included the area between Building 40 that contained the heating/cooling system and Building 41 that contained the jet engine test cells; and (2) Site 3, on the northeast side of the site, which included a former wastewater lagoon and a sludge disposal area (fig. 2). Historically, wells in the Site 1 area were found to have maximum TCE concentrations >100,000 micrograms per liter (μg/L), whereas Site 3 was characterized as having maximum TCE concentrations of <1,500 µg/L (International Technology Corporation, 1994). Breakdown products of TCE, specifically cisDCE and VC, were also detected in both areas at concentrations that exceeded the NJDEP Drinking Water maximum contaminant levels (MCLs) for TCE (1 μg/L), cisDCE (70 μg/L), and VC (1 μg/L) (New Jersey Department of Environmental Protection, 2020). NJDEP Surface Water Quality Standards for TCE and VC are 1.0 and 0.082 µg/L, respectively, but there is no regulatory limit for cisDCE in surface water.

The Navy and its contractors established a long-term water quality monitoring program in the late 1990s (EA Engineering, Science, and Technology, Inc., 2000) in which selected wells and storm sewers at the NAWC site are sampled on a quarterly to quadrennial basis. The most recent results from years 2014 through 2017 are presented by Imbrigiotta and Fiore (2021).

The VOC contamination at NAWC evolved over the life of the facility in a similar manner to many sites that overlie fractured rock aquifers. The TCE used at the site was in its pure liquid form, a dense non-aqueous phase liquid (DNAPL). Since DNAPL is denser than water, when it is spilled onto the ground, it seeps into the subsurface and sinks in the aquifer under the influence of gravity (Cohen and Mercer, 1993). DNAPL TCE physically travels downward along sloping bedding partings and vertical fractures until it becomes trapped in dead-end fractures or is blocked by a low permeability layer (Cohen and Mercer, 1993). Once in the aquifer, the DNAPL TCE slowly dissolves, as it is somewhat soluble (1,100 milligrams per liter [mg/L]) (National Institute for Occupational Safety and Health, 2007), and begins traveling with the flowing groundwater. This is the scenario that undoubtedly occurred with the spilled TCE at NAWC.

Once dissolved, the concentration and distribution of TCE is affected by many natural processes (Mackay and others, 1985; Vogel and others, 1987; Wiedemeier and others, 1998). TCE is advectively transported downgradient and eventually discharges into streams, storm sewers, or extraction wells (Lacombe, 2000; Tiedeman and others, 2018). TCE may diffuse into the primary porosity of the rock matrix or sorb onto the organic materials present in the geologic formation (Goode and others, 2014). Because TCE has a high vapor pressure (National Institute for Occupational Safety and Health, 2007), TCE in shallow groundwater may volatilize into the soil gas in the unsaturated zone. TCE may be biodegraded by natural bacteria or by nonnative bacteria that have been injected into the aquifer. Biodegradation dechlorinates TCE to form products including cisDCE, VC, and ethene (Wilson and Wilson, 1985; Barrio-Lage and others, 1986; Wilson and others, 1995; Lorah and Olsen, 1999; Shapiro and others, 2018). Given that TCE was the original VOC contaminant spilled at the site, the presence of cisDCE and VC in the groundwater indicates that anaerobic biodegradation continues to occur in the aquifer at NAWC. Since spillage of TCE stopped when NAWC closed in the late 1990s, the only continuing sources of TCE to the fractured rock aquifer are possible pockets of residual DNAPL in fractures, desorption of TCE from organic materials in the aquifer, and diffusion of TCE from the primary porosity of the rock matrix.

Firefighting foams were reportedly used during the early 1990s along the east side of the former NAWC site during firefighting training exercises. The AFFF likely contained PFOS, PFOA, PFNA, PFBS, and other PFAS compounds (Interstate Technology & Research Council, 2018b). PFOS, PFOA, PFNA, and PFBS are highly soluble in water, have low vapor pressures that limit volatilization, have a high affinity to sorb onto soil organic matter in the geologic substrate (Interstate Technology & Research Council, 2018a), and are mostly resistant to biodegradation, though some may potentially biodegrade in certain environments (Huang and Jaffé, 2019).

Once an aquifer has been contaminated with PFAS compounds, remediation to meet Federal and state water-quality standard levels is difficult. Currently (2023), NJDEP MCLs for PFAS are 13 nanograms per liter (ng/L) for PFNA, 13 ng/L for PFOS, and 14 ng/L for PFOA (New Jersey Department of Environmental Protection, 2020). PFBS is not regulated by NJDEP, so the U.S. Environmental Protection Agency (EPA) Regional Screening Level of 400,000 ng/L is used.). When this sampling was conducted (2018), the EPA did not have MCLs for PFAS compounds but had a proposed health advisory limit (HAL) of 70 ng/L for the individual or combined concentrations of PFOS and PFOA (U.S. Environmental Protection Agency, 2016) There are no surface water-specific regulatory limits for PFAS at the state or Federal level, so PFAS concentrations in surface water are discussed with regards to the NJDEP drinking water MCLs described above (New Jersey Department of Environmental Protection, 2020).

The NAWC well network consists of 130 wells or well zones that are located on and around the former Navy property (fig. 2; appendix 2, table 2.1). Of these wells, 85 were sampled in 2018 (appendix 3, table 3.1). Construction details of all wells at or near the NAWC site are available in Fiore (2019). These wells are used for routine monitoring of water levels (Lacombe, 2000; Lacombe, 2002; Fiore and Lacombe, 2020) as well as scheduled sampling for VOCs and inorganics. Currently (2023), 10 of the wells also serve as extraction wells for the remedial pump-and-treat system operating at the site. For the purpose of discussion in this report, the wells have been assigned to five groups on the basis of the quadrant of the site in which they are located, (1) Northeast (NE), (2) Southeast (SE), (3) Northwest (NW), (4) Southwest (SW), and (5) offsite. Wells in the offsite group are outside the quadrant boundaries on figure 2. Two areas of VOC contamination have been delineated at the NAWC site: a large area in the SW quadrant (Site 1) and a smaller area in the NE quadrant (Site 3). The wells with their respective quadrants and construction information are listed in appendix 2. The wells range from 15 ft to more than 410 ft in depth.

In 2018, 85 wells were sampled at NAWC including 16 wells in the NE quadrant, 16 wells in the SE quadrant, 11 wells in the NW quadrant, 31 wells in the SW quadrant, and 11 offsite wells. These totals include one extraction well in the NE quadrant, three extraction wells in the NW quadrant, and six extraction wells in the SW quadrant. Wells 04BR, 05BR, 16BR, 31S, 41BR, and WDW are reserve extraction wells that are not actively used as pump-and-treat extraction wells and are discussed as monitoring wells in this report. There are 16 designated “research” wells installed for specific research projects conducted at the site: 68BR, 70BR-10, 70BR-72, 71BR, 73BR, 80BR, 83BR, 84BR, 85BR, 86BR, 87BR, 88BR, 89BR, 92BR, 93BR, and 94BR. Wells 70BR-10 and 70BR-72 were constructed after dividing a corehole into two separate vertical zones. Wells 71BR, 73BR, and 80BR contain multiple depth intervals isolated by pneumatic packers. These intervals (71BR-F, 71BR-G, 71BR-H, 73BR-A, 73BR-BC, 73BR-D1, 73BR-D2, 73BR-E, 80BR-A, 80BR-B, 80BR-C, and 80BR-D) are discussed as separate monitoring wells in this report. Monitoring well 36BR is also divided into two depth intervals separated by a pneumatic packer, 36BR-A and 36BR-B, which are also discussed as separate wells in this report.

The surface water system at NAWC was briefly described by Lacombe (2000). Figure 3 shows the location of underground storm sewer lines, manholes, springs, and surface water sites that have been identified and sampled at the NAWC site. The spring that feeds the ancestral branch of Gold Run (hereafter referred to as West Branch Gold Run) occurs near the intersection of Parkway Avenue, Jack Stephan Way, and West Upper Ferry Road (fig. 3). Approximately 3,000 ft to the east of the site perimeter, the West Branch Gold Run culvert combines with the outflow of another culvert that carries storm sewer runoff from an eastern branch of Gold Run; at this location, the flow from these two culverts forms the main stem stream of Gold Run, which runs to the south.

Four stormwater collection systems exist at NAWC (fig. 3): the West Ditch system, Outfall 2 system (OF-2), Outfall 3 system (OF-3), and Outfall 4 system (OF-4). The West Ditch system originates at a spring just south of the cooling towers (WDspring), before being routed into a culvert that discharges into manhole MH-140, and then into an oil-water separator (WDin). Outflow from the separator flows through a culvert and then enters manhole MH-125.9 where it discharges into the Township Line storm sewer that runs east underneath Parkway Avenue. The Township Line is separate from the West Branch Gold Run culvert that also runs under Parkway Avenue. During reconstruction of the West Ditch in 2017, additional connections that discharged into the West Ditch were removed (Tetra Tech, 2018).

The OF-2 system originates north of the cooling towers and runs south towards the Township Line storm sewer. The OF-2 system has several branch connections along it, and ultimately discharges into manhole MH-121.5 where the effluent enters the Township Line storm sewer that runs underneath Parkway Avenue. The OF-3 system originates around a large former hanger building on the airport north of NAWC. The OF-3 system is entirely culverted and has multiple connections along its path, eventually discharging into manhole MH-118.5 and the Township Line storm sewer. The OF-4 system originates in the northeast quadrant of the site as swales in the grassy area located on the airport property. The swales flow into an open-air concrete-lined ditch before joining the OF-4 culvert. The OF-4 system eventually reaches manhole MH-117 where it discharges into the Township Line storm sewer that runs under Parkway Ave. The OF-4 culvert and swales are discussed further in this report with regard to PFAS sources.

The Township Line storm sewer originates along Jack Stephan Way somewhere north of manhole MH-388.9, however, the starting point of this stormwater drainage is unknown. After reaching MH-318.9, the Township Line storm sewer turns southeast and intersects manholes MH-125.9, MH-121.5, MH-118.5, and MH-117, where it receives stormwater originating from the NAWC site. Eventually the Township Line storm sewer reaches manhole GR-OF (approximately 1,650 ft southeast of MH-117) where it discharges into Gold Run at the same location as the West Branch Gold Run culvert, to eventually flow to the Delaware River.

A subset of wells across the NAWC site are sampled each year as part of the U.S. Navy’s long-term monitoring program. The sampling frequency for each well varies according to the schedule detailed in the annual Sampling and Analysis Plan (SAP) produced and agreed upon by the Navy, private contractors, and the USGS (appendix 2; Koman Government Solutions, 2018). In 2018, a total of 85 wells were sampled for VOCs and 10 wells were sampled for PFAS. Samples for inorganic constituents were also analyzed during this sampling event but are not discussed in this report. The surface water and manhole sites were sampled for VOCs on a quarterly basis and for PFAS only once during the spring (appendix 4). Sampling frequency listed in this report may be obsolete at the time of publication of this report.

VOC sampling techniques varied by well according to the protocols laid out in the SAP (Koman Government Solutions, 2018). Some wells were purged before being sampled using a hybrid technique whereby a variable-speed stainless-steel submersible pump was used to first remove 1.5 casing volumes of water from the well. The pumping rate was then lowered to <1 liter per minute (L/min) and field parameters (temperature, pH, specific conductance, dissolved oxygen, redox potential, and turbidity) were monitored using a multiparameter water-quality sonde in a flow-through cell until all parameters reached stable values. Samples were collected unfiltered directly into 40-milliliter VOC vials. Some wells were low-flow purged entirely at a rate of <1 L/min. Regardless of technique, purging and sampling was completed using Teflon-lined polyethylene tubing dedicated to each well. A total of 49 wells were sampled in 2018 using regenerated cellulose dialysis membrane passive samplers following the methods of Imbrigiotta and others (2007), Imbrigiotta and Trotsky (2011), and Imbrigiotta and Harte (2020) (appendix 3). The passive samplers were deployed and allowed to equilibrate for approximately 2 weeks before being retrieved and subsampled. Extraction wells were sampled from a tee in the discharge line of the dedicated submersible pump installed in each well; samples were collected after field parameters had reached stable values.

PFAS samples were collected after purging using a low-flow method that utilizes a peristaltic pump and precleaned high-density polyethylene tubing. Sampling was conducted on a separate date from the VOC sampling to avoid cross-contamination from the Teflon-lined tubing used to collect VOC samples. Purging and sampling was conducted at a flow rate of <1 L/min after field parameters had reached stable values. Unfiltered PFAS samples were collected directly into high density polyethylene bottles.

For the purpose of this report, water in drains and in storm water sewers (sampled at manholes) will be termed “surface water.” The surface waters were sampled for chemical analysis by the Navy contractor on a quarterly basis. Methods of grab-sample collection and analysis, and concentrations were reported to the Navy through contractor reports (Koman Government Solutions, 2018, 2020). Sampling was conducted only if precipitation had not occurred within the preceding 72 hours and after the pump-and-treat system discharging treated effluent into OF-2 was operating continuously for 72 hours prior to being shut down the morning of sampling. This ensured that the surface-water samples were composed primarily of groundwater seepage.

The surface-water sites sampled in 2018 and parameters that were measured at each site are listed in appendix 4 (table 4.1). At some manholes, two grab samples were collected, one from the pipe entering from the north (leaving from the NAWC site), and a second sample at the incoming pipe of the Township Line storm-sewer pipe (flowing from the west). Immediately after Navy contractors conducted the sampling, the USGS measured volumetric discharge (Q) at six sites by directly collecting water into a 2-liter (L) graduated cylinder and measuring the time to collect certain volumes of effluent. The cylinder was placed below the lip of the discharge pipes or at site WDspring, immediately downstream of the weir. Chemical samples and discharge measurements were made four times in 2018: February 27 (winter), June 15 (spring), August 10 (summer), and December 11 (fall). Fluxes (mass per time) were calculated for TCE, cisDCE, VC, PFOA, PFOS, and PFNA. The flux of contaminants was calculated as the discharge multiplied by the concentration in the associated chemical sample. Fluxes could not be calculated for sites where discharge was not measured. When the VOC or PFAS concentration was less than the method detection limit (MDL) and Q was measured, the flux was calculated using the MDL and given a "less than" value.

VOC and PFAS concentrations in samples collected by the Navy contractors were analyzed by a Navy- and NJDEP-certified lab (Koman Government Solutions, 2020). VOC analyses were performed using EPA Method 8260 (U.S. Environmental Protection Agency, 2018a). PFAS analyses were run using a modification of EPA Method 537 (U.S. Environmental Protection Agency, 2018b). Analytical results were sent to the Navy contractors for quality assurance and quality control review. The approved data were forwarded, with permission from the Navy, to the USGS, and are available in a USGS data release (Fiore and Imbrigiotta, 2021).

The areal distribution of TCE, cisDCE, and VC contamination during 2018 were plotted on maps (figs. 4–6) using color-coded symbols to indicate the range in measured concentrations. Table 1 summarizes the number of wells in each quadrant as a function of these ranges in concentrations, and the number of wells having concentrations that exceed the NJDEP MCLs.

Inspection of these maps shows that two areas exist at the site where concentrations of VOCs exceed 1 μg/L. A small area in the NE quadrant (historically called Site 3), and a larger area in the NW and SW quadrants (historically called Site 1) have multiple wells with concentrations that exceed the NJDEP MCLs of 1 µg/L for TCE and VC and 70 µg/L for cisDCE. Most TCE spillage occurred in Site 1; Site 3 once contained a wastewater lagoon and sludge drying beds that provided a secondary source of TCE to the aquifer below them. These areas are delineated on figures 2 and 3 as the pink and green shaded areas. These maps do not differentiate concentrations at different depths in the aquifer.

Maps showing the distribution of the PFOS, PFOA, and PFNA concentrations in the wells at the site are shown in figures 7–9. All PFBS detections were below the EPA Regional Screening Level; therefore, a map was not created for PFBS. Because only the 10 extraction wells were sampled in 2018, the distribution shown on these maps does not fully delineate the distribution of these chemicals across the site. Most of the wells sampled were in the SW (six wells) and NW (three wells) quadrants with only one well sampled in the NE quadrant. For a more in-depth analysis of the previously defined PFAS distribution, see Imbrigiotta and Fiore (2021).

During 2018, only one well had a PFOS concentration greater than its NJDEP MCL of 13 ng/L, well 48BR in the NE quadrant (470 ng/L) and was only found above the detection limit in one other well, 08BR in the NW quadrant (11 ng/L). Two wells had PFOA concentrations greater than or equal to its NJDEP MCL of 14 ng/L, well 48BR (54 ng/L) and well 08BR (14 ng/L). PFOA was only found above the detection limit in four other wells, 22BR (11 ng/L), 15BR (5.2 ng/L), 91BR (4.6 ng/L), and BRP02 (3.3 ng/L), all in the SW quadrant. No wells had PFNA concentrations greater than the NJDEP MCL of 13 ng/L, and PFNA was only detected in 48BR (9.8 ng/L) and 08BR (4.3 ng/L). No wells had PFBS concentrations exceeding the EPA Regional Screening Level of 400,000 ng/L. PFBS was only found above the detection limit in two wells, 48BR (89 ng/L) and 45BR (3.1 ng/L).

Tetra Tech (2018) and Imbrigiotta and Fiore (2021) discuss high concentrations of PFAS in the NE quadrant, with detected concentrations as far east as well 50BR, which indicates that PFAS may be migrating offsite to the east. The high concentrations of PFAS in the storm sewers that drain the NE quadrant (discussed below) and the interpreted groundwater flow directions (Fiore and Lacombe, 2020) indicate that 48BR, the only extraction well on the entire east side of NAWC, is likely not sufficient to capture all of the dissolved PFAS contamination.

To show the extent of VOC contamination at depth below the surface, nine lines of section were drawn across the site (figs. 10–36); the locations of sections A-I are shown on figures 4–9. Sections are dip-aligned and drawn from southeast to northwest. Wells are projected onto cross sections and shown as vertical lines scaled to the depth of each of the wells. Cross sections A-D cover the NE and SE quadrants where Site 3 is located. Cross sections E-I pass through the wells in the Site 1 area. Also shown on the cross sections are the approximate locations and depths of the Lockatong Formation fissile units (black lines), the top of the Stockton Formation layers (red lines), and the bottom of the weathered surface bedrock (gray lines), based on cross sections by Lacombe and Burton (2010) and Fiore and Lacombe (2020). The fault zone separating the Lockatong and Stockton Formations is also shown on the cross sections in figures 10–36. This fault zone strikes northeast to southwest across the site. The general direction of groundwater flow inferred from the measured 2018 groundwater elevations (Fiore and Lacombe, 2020) is also shown.

The distributions of VOCs at the NAWC site are determined by the location of the chemical releases, the migration of DNAPL, and the direction in which the groundwater flows. Groundwater flow direction is controlled by the regional flow patterns, the structure of the bedrock and fracture patterns, and the influence of the extraction well pumping. The general regional groundwater gradient is from the northwest to southeast (Lacombe, 2000), but the groundwater flow in the fractured bedrock at the site has been established as flowing perpendicular to this gradient, toward the southwest, and is controlled by the strike of the bedding partings in the bedrock (Lacombe, 2000; Fiore and Lacombe, 2020). The flow direction at NAWC is also greatly influenced by the cone-of-influence around each of the 10 extraction wells. The horizontal flow of water in the unconsolidated overburden of the site is toward the swales, ditches, and storm sewers (Fiore and Lacombe, 2020). Because no wells offsite to the west or south have TCE concentrations >1 μg/L, this is a good indication that the pump-and-treat system continues to keep the VOC contaminated groundwater from flowing offsite to the west. The presence of only a few affected wells in the Stockton Formation further indicates that the fault acts to hinder southerly flow within the bedrock. Contamination in the Stockton Formation is likely caused by groundwater flow within the shallow overburden from the Lockatong Formation into the Stockton across the fault and then migration downward into the bedrock.

To evaluate the temporal changes in VOC concentrations, the concentration of TCE, cisDCE, and VC in each well sampled in 2018 was compared to the concentration measured in the most recent sample prior to 2018, as well as the overall long-term trends (appendix 3; figs. 37–43). In table 3.1 the relative direction of change in the 2018 sample is indicated as “+” if the concentration increased in 2018, “-” if the concentration decreased in 2018, and “0” if there was no change in concentration or the change could not be determined because one or both sample analyses reported the concentration as being below their respective MDL. The data do not permit determining if the change in concentration was statistically significant, so a change was considered substantial only if the difference was at least double (if an increase), or at least half (if a decrease), of the concentration found in the previous sample. For a more thorough discussion on trends of VOC concentrations over a longer time period prior to 2018, see Imbrigiotta and Fiore (2021).

The only extraction well in that showed a notable change in VOC concentrations between 2017 and 2018 was well 45BR (appendix 3). This well had decreasing TCE and increasing cisDCE and VC concentrations from the previous sample (fig. 37). Since mid-2010, the TCE concentrations and cisDCE and VC concentrations have varied inversely. The increasing cisDCE and VC concentrations trends may be caused by an increase in anaerobic biodegradation in well 45BR itself or within its cone of depression from which it is now drawing in groundwater with higher concentrations of the TCE biodegradation byproducts.

Extraction well 91BR showed a small increase in all VOC compounds between 2017 and 2018 (fig. 38) though not high enough to be considered substantial (appendix 3). TCE and cisDCE have shown increasing trends since its installation in 2014, so an increase in VOC concentrations may be expected as its cone of influence expands and draws in more contaminated groundwater from greater distances. For all other extraction wells, no substantial change in VOC concentrations were noted from 2017 to 2018 (figs. 37–38). The longer-term trends for these extraction wells remain unchanged and were discussed by Imbrigiotta and Fiore (2021).

Inspection of the VOC concentration data in monitoring wells in the SW quadrant showed that TCE increased in concentration in one well (36BR-A) and decreased in six wells (24BR, 40S, 70BR-10, 73BR-A, 74BR, BRP01), cisDCE increased in five wells (30BR, 36BR-A, 60BR, 65BR, 73BR-A) and decreased in three wells (24BR, 70BR-10, BRP01), and VC increased in two wells (40BR, 73BR-A) and decreased in six wells (04BR, 16BR, 24BR, 70BR-10, 74BR, BRP01) (appendix 3, fig. 39). The SW quadrant includes the Site 1 contamination area as well as the locations of the bioaugmentation experiments performed in 2005 (Geosyntec Consultants, 2010) and 2008 (Tiedeman and others, 2018; Shapiro and others, 2018). Wells 30BR, 73BR-A, 74BR, and BRP01 are located in the bioaugmentation experiment areas and their trends in concentrations are likely caused by continued reductive dechlorination occurring in these areas. These wells have all had stable to gradually decreasing VOC concentrations over the longer term (fig. 39). Well 38BR in the 2005 bioaugmentation area, despite showing negligible change from its previous sample, has had long-term trends of decreasing TCE concentrations and increasing cisDCE and VC concentrations that are likely a result of continued biodegradation in this well (fig. 39). Well 16BR had negligible TCE and cisDCE changes from the previous year, but has shown a general decreasing trend in cisDCE and VC since 2012 (fig. 39). Well 70BR-10 is located in the 2008 bioaugmentation area but its shallow depth and generally decreasing VOC concentrations may also be related to volatilization to the unsaturated zone. Well 36BR-A was the injection well for the 2008 bioaugmentation study and TCE concentrations have remained low since the experiment concluded, so the three orders-of-magnitude increase in TCE concentration between 2017 and 2018 was unexpected (fig. 39). This observation may mean that anaerobic biodegradation has slowed in this well (cisDCE and VC were relatively unchanged) or that either desorption of TCE from contaminated aquifer materials or dissolution of DNAPL TCE from dead-end pore spaces has occurred near this well to cause this increase.

Outside the bioaugmentation areas in the SW quadrant, the decreasing trend in TCE and cisDCE concentrations in shallow well 40S (fig. 39) may be caused by volatilization to the unsaturated zone or its close proximity to three different extraction wells. The increasing trends in cisDCE and VC concentrations in wells 40BR, 60BR, and 65BR (fig. 39) are likely caused by the effects of natural biodegradation in the fractured rock aquifer. Although wells 04BR and 24BR showed decreases in all VOCs from 2017 to 2018, cisDCE and VC concentrations were generally stable for the prior 10 years.

Increasing TCE concentrations were found in 02BR (NE quadrant) and 42BR (SE quadrant) and increasing cisDCE concentrations were found in 02BR and 11BR (both NE quadrant) (figs. 40 and 41). Wells 02BR and 11BR exhibit stable or minimal change in VOC concentrations overall, so the changes from the previous sample may not be sustained. The increase in the TCE concentration from well 42BR is from below the MDL to just above it, and as discussed earlier, it may be a sampling artifact. Decreases in TCE concentrations occurred in 31S (SE quadrant), possibly from volatilization or its proximity to extraction well 48BR based on the long-term trend (fig. 41). No wells outside the SW quadrant showed decreasing cisDCE or VC concentrations from 2017 to 2018. The change in concentrations in the NW quadrant and offsite wells were either indeterminant or indicated no change, but generally, most of these wells are below the MDL for all VOCs (figs. 42 and 43).

The presence of VOCs in the surface water at the site indicates that residual TCE contamination may be present in soils and in shallow aquifer materials. Precipitation falling on the ground infiltrates through contaminated unconsolidated surface materials that may discharge desorbed dissolved VOCs into swales and ditches as well as into storm sewers. Seasonally, there may be periods when groundwater levels rise to the point where the saturated zone intersects permeable or broken pipes and culverts; high VOC concentrations seen along the West Ditch (appendix 4; fig. 44) and shallow groundwater levels (Fiore and Lacombe, 2020) support this assertion. Groundwater initially discharges into the West Ditch at WDspring (fig. 3), which has <0.75 μg/L concentrations of all VOCs (appendix 4). However, locations along the West Ditch downstream of WDspring are contaminated with VOCs despite being culverted and not engineered to receive groundwater discharge. A perforated subsurface seepage pipe with a collection sump, referred to as the “Interceptor” site (Koman Government Solutions, 2018), collects groundwater on the east side of the West Ditch in this area. The interceptor sump was found to have substantial concentrations of TCE (6,400 μg/L), cisDCE (1,600 μg/L), and VC (13 μg/L) in 2018 (appendix 4). The high VOC concentrations along the West Ditch downstream from this sump are likely caused by the water table in the Interceptor area intersecting permeable or broken segments of the West Ditch culvert, thereby allowing contaminated groundwater to seep in. This likely occurs at or near MH-140 (fig. 3; appendix 4). Additionally, because the abandoned buildings are open to the weather, rain may wash residual VOCs from the buildings into drains and ultimately the storm-sewer lines. The measurable concentrations of VOCs found at the Roof Drain sampling site (fig. 3; appendix 4) are evidence of this occurring.

In 2018 surface water samples, TCE was quantifiable in 59 samples, cisDCE in 58 samples, and VC in 23 samples (appendix 4). VOCs were present at detectable levels at all surface water sites except WDspring. Locations where notably elevated VOC concentrations were found included MH-118.5-T, MH 121.5-T, MH 125.9-N, WDin, and MH 140-Bottom. The highest surface water TCE concentration (480 μg/L) was found in February in MH140-Bottom (fig. 44). The highest surface water cisDCE concentration (340 μg/L) and VC concentration (13 μg/L) were found in December in MH 121.5-T (figs. 45 and 46). In general, the highest surface water VOC concentrations were found coming from the West Ditch drainage area in the February and December samples, seasons when groundwater levels were highest (Fiore and Lacombe, 2020) and when cold temperatures would have slowed volatilization and biodegradation. Also noteworthy was that in most sample locations, cisDCE concentrations exceeded TCE concentrations (appendix 4; figs. 44 and 45). The only exceptions were in MH140-Bottom, MH-140-In, and the Roof Drain. cisDCE was also elevated in GR-OF (located farthest downstream to the east) in February (18 μg/L). This high concentration of cisDCE appeared to be the result of higher cisDCE entering the Township Line at MH-118.5-N in February (fig. 45). Similarly, elevated VC (1.9 μg/L) in MH-117-T in February was caused by VC entering from the West Ditch system at MH-129.5-N (12 μg/L) and from the OF-3 system at MH-118.5-N (2.8 μg/L) (fig. 46). VC was only sporadically detected, most notably being found in MH-121.5-T (all samples) and MH-125.9-N (February and December). VC concentrations were below detection limits at most surface water sites during the year, likely because of its high volatility (fig. 46). The presence of high concentrations of the degradation products in the surface water indicates that biodegradation is occurring in the contaminated groundwater in the bedrock aquifer and is discharging to the storm-sewer lines.

High groundwater levels result in a greater likelihood for contaminated groundwater discharge and seepage into the storm sewer lines. At all surface-water sampling sites, the highest concentrations of VOCs occurred in the months of February and December, when 2018 NAWC groundwater levels were higher compared to those measured during the June and August sampling events (Fiore and Lacombe, 2020). Manholes MH-388.9, MH-318.9, and MH-125.9-T are located farthest upstream along the Township Line and do not receive direct stormwater runoff from the NAWC site (fig. 3), yet each site had detectable concentrations of TCE and (or) cisDCE above their MCLs at least once in 2018 (appendix 4). Groundwater levels in monitoring wells 33BR and 58BR, located west of MH-388.9, are known to respond to pumping from extraction well 45BR (Lacombe, 2000; Tiedeman and others, 2010), which went offline from February to April and contributed to higher groundwater levels in those wells during this time period (Fiore and Lacombe, 2020). Similarly, pumping volumes were lower overall during December, which also caused higher groundwater levels (Fiore and Lacombe, 2020). Higher groundwater levels in this area would favor TCE- and cisDCE-contaminated groundwater discharge into the West Ditch and potentially the upstream groundwater discharge area that flows into MH-388.9, ultimately resulting in higher VOC concentrations along the entire Township Line storm sewer.

Samples for the PFAS suite of compounds (PFOS, PFOA, and PFNA) were collected at all surface water sites in 2018, except for the Roof Drain, Port 001, and MH-140-In. PFOS, PFOA, and PFNA were detected in all 16 surface-water samples collected in June of 2018 (appendix 4). Site MH-117-N had the highest concentrations of PFOS (2,000 ng/L) and PFOA (100 ng/L). Site WDspring had the highest concentration of PFNA (13 ng/L). Overall, all three PFAS were found to enter the Township Line storm sewer at concentrations exceeding their NJDEP MCLs from the West Ditch system, the OF-2 system, the OF-3 system, and the OF-4 system (fig. 47). PFAS samples were only collected in June, so it was not possible to evaluate the effects of seasonality that may relate to groundwater levels, as was done with VOCs. The fact that the highest overall concentrations of PFAS were sampled in MH-117-N, which connects with the OF-4 system, suggests an important source of the compounds lies in the NE and possibly the SE quadrants, including the northern grassy area located on the Airport property. This was the area of NAWC where AFFF was reportedly discharged to the ground surface in the early 1990s (Tetra Tech, 2018).

Also noteworthy is the geographical relation between the ratio of PFOS to PFOA across the site. At locations where PFOS concentrations were found to exceed the NJDEP MCLs, PFOS concentrations were always higher than the PFOA concentrations (appendix 4; fig. 47). This appears to be exclusively the case along the east side of NAWC, in the area where the reported discharge of AFFF occurred (Tetra Tech, 2018). This higher ratio of PFOS to PFOA in the surface water also occurred in the groundwater on the east side of NAWC.

Along the west side of the NAWC site, the surface-water concentrations of PFOA exceeded those of PFOS (appendix 4; fig. 47). This was the case for samples from WDspring and from manholes MH-388.9 downstream along the Township Line until MH-118.5-N, where the OF-3 system enters the Township Line sewer (figs. 3 and 47). Therefore, although the data on the west side of NAWC are limited, surface water on the west side of the site was characterized by a lower PFOS to PFOA concentration ratio, whereas on the east side of the site it is dominated by a higher PFOS to PFOA concentration ratio. The divide between these areas is near the OF-2 system, as shown in the sample from MH-121.5-N, which demonstrates a negligible difference between concentrations of PFOS and PFOA.

Surface water PFNA concentrations were not found to exceed their NJDEP MCL of 13 ng/L at any surface-water sampling location in 2018. The highest PFNA concentrations were detected in WDspring (13 ng/L), WDspring duplicate (12 ng/L), MH-125.9-N (12 ng/L), and MH-117-N (12 ng/L). PFOA and PFOS both exceeded their NJDEP MCLs of 14 ng/L and 13 ng/L, respectively, at sites MH-121.5-N, MH-118.5-N. MH-118.5-T, MH-1117-N, MH-117-T, and GR-OF.

Daily fluxes of VOCs and PFAS were estimated by multiplying the rate of flow (Q) by the concentration of each VOC or PFAS and normalizing to unit time (day) (appendix 4). Fluxes were calculated for three VOCs (TCE, cisDCE, and VC) and three PFAS (PFOA, PFOS, and PFNA). Fluxes could not be calculated for sites where Q was not measured, for example, if the sampling site was dry and no sample or Q measurement was possible. Fluxes were calculated for samples from WDin, MH-125.9-T, MH-125.9-N, MH 121.5-N, MH-118.5-N, and MH-117-N. The maximum fluxes of the VOC species were found to be 3.6 grams per day (g/d) for TCE at site WDin (December sample), 13 g/d for cisDCE at site MH-118.5-N (February), and 0.31 g/d for VC at site MH-118.5-N (February). The maximum flux measured for PFAS compounds were 160 milligrams per day (mg/d) PFOS, 8.4 mg/d PFOA, and 0.63 mg/d PFNA, which were all measured at MH 118.5-N in the June sample.

These fluxes should be considered rough estimates because the discharge measurements used to measure flow in the storm sewers need to be upgraded and standardized to give more meaningful estimations of the mass of VOCs and PFAS leaving the site through surface waters. Likewise, seasonal sampling of PFAS will be needed to determine ranges in fluxes that may occur as groundwater levels and discharge vary over the year.