USGS National Water Quality Laboratory (NWQL), Denver, Colorado

Tapwater samples were analyzed at the NWQL for 212 pesticides, 32 PFAS, 111 pharmaceuticals, 53 SVOC, and 76 VOC.

Pesticide samples (10 milliliters [mL]) were filtered in the field using a 0.7-micrometer (μm) nominal pore-size glass-fiber syringe filter and analyzed using direct aqueous injection-high performance liquid chromatography and tandem mass spectrometry (DAI HPLC-MS/MS) as detailed in Sandstrom and others (2015). Method detection limits (MDL) are determined using EPA procedures that are outlined in U.S. Environmental Protection Agency (2005). Reporting limits, reported as laboratory reporting limits, were set to twice the MDL (Childress and other, 1999). Concentration results are reported in nanograms per liter (ng/L).

Whole-water samples were analyzed by DAI HPLC-MS/MS for PFAS as described in Kolpin and others (2021). MDL were determined using American Society for Testing and Materials (ASTM) detection and quantitation-estimation and data assessment-utilizing software (DQCALC; ASTM International, 2016). Interim reporting limits (IRL) were set at twice the MDL. Concentration results are reported in ng/L.

Pharmaceutical samples (10 mL) were filtered in the field using a 0.7-μm nominal pore-size glass-fiber syringe filter and analyzed using DAI LC-MS/MS, which is further detailed in Furlong and others (2014). MDL are determined using EPA procedures outlined in U.S. Environmental Protection Agency (2005). IRL were determined as described in Oblinger Childress and others (1999) and allow for quantitation below the corresponding MDL. Concentration results are reported in ng/L.

Whole-water samples were analyzed for SVOC (base/neutral and acid extractable compounds). A 1-liter (L) water sample was extracted prior to analysis by gas chromatography/mass spectrometry (Fishman, 1993). Minimum reporting levels (MRL) were determined by the DQCALC procedure to get the reporting limit detected by DQCALC (RLDQC). The RLDQC is referred to in U.S. Geological Survey (2015) which is further described in ASTM International (2016). Concentration results are reported in μg/L.

Two different whole-water analytical methods were used to determine VOC concentrations in sample water. While both methods utilized electron impact ionization mass spectrometry, 37 heat-purgeable VOC were analyzed in simultaneous full scan/selected ion monitoring mode and 49 ambient purgeable VOC were analyzed in full scan mode (Rose and others, 2016). Reporting limits were determined using the RLDQC method, except for 2-Ethyl-1-hexanol, chloromethane, methylene chloride, and toluene, which were blank-limited and are reported using MRL as described in Oblinger Childress and others (1999). Concentration results are reported in μg/L.

USGS Organic Chemical Research Laboratory (OCRL), Sacramento, California

Whole-water samples were measured for 22 DBP as detailed in Hladik and others (2014). Samples were extracted using an Oasis HLB (hydrophilic and lipophilic balance) solid-phase extraction (SPE) cartridge (Waters Corporation, Milford, MA), and then were analyzed by gas chromatography/mass spectrometry. MDL were determined using methods described in U.S. Environmental Protection Agency, 2005. Concentration results are reported in μg/L.

USGS Redox Chemistry Laboratory (RCL) and Analytical Trace Element Chemistry Laboratory (ATECL), Boulder, Colorado

Filtered (0.45 μm pore-size nylon filter) water samples were analyzed for anions, and whole-water samples were analyzed for major cations and trace elements at the RCL. Lead, arsenic, hexavalent chromium, and total chromium were analyzed by graphite-furnace, atomic-absorption spectrometry, and results are reported in μg/L (Ball and McCleskey, 2003; McCleskey and others, 200320; Hergenreder, 2011). Anions (bromide, chloride, fluoride, nitrate as nitrogen, and sulfate) were analyzed by ion chromatography (U.S. Environmental Protection Agency, 1993; Brinton and others, 1996). Results are reported in milligrams per liter (mg/L). Major cations (calcium, magnesium, potassium, silica, and sodium), phosphorus, and trace elements (aluminum, antimony, barium, beryllium, boron, cadmium, cobalt, copper, iron, lithium, manganese, molybdenum, nickel, selenium, strontium, vanadium, and zinc) were analyzed by inductively coupled plasma-optical emission spectrometry (U.S. Environmental Protection Agency, 2014). Results are reported in mg/L for major ions and μg/L for trace elements (table 1.1). Uranium was analyzed at the ATECL by inductively coupled plasma-mass spectrometry (McCleskey and others, 2019). Uranium results are reported in μg/L.

SGS AXYS Laboratory, British Columbia, Canada

The SGS AXYS analytical method used for PFAS was MLA-110 REV. 02 VER. 11, dated January 13, 2021, and is available upon request from SGS AXYS (Sean Campbell, Environment, Health, and Safety project manager, written commun., 2021). Samples were stored in high-density polyethylene containers that were frozen for up to 90 days. After spiking with isotopically labeled surrogate standards, samples were extracted and cleaned by SPE. Analysis of the sample extract was performed by ultra-high performance liquid chromatography and tandem mass spectrometry (UHPLC-MS/MS) and reversed phase C18 column (18 carbon chains) using a solvent gradient. The column was coupled to a triple-quadrupole mass spectrometer run at unit-mass resolution in the Multiple Reaction Monitoring in negative-electrospray ionization mode. Target compounds were quantified using the isotope dilution/internal standard method, comparing the area of the primary (quantifying) transition product ion of the target compound to the carbon-13-labeled (or deuterium-labeled) standard and correcting for response factors. Results are reported in ng/L.

Northern Lakes Service Inc., Crandon, Wisconsin

The Northern Lakes Service (NLS) utilizes EPA Method 537.1 to analyze for 33 PFAS (U.S. Environmental Protection Agency, 2020). After the addition of isotopically labeled standards, samples were passed through SPE and analyzed using UHPLC-MS/MS. Additional method details are available upon request from NLS (Steve Hefter, laboratory manager, written commun., 2021). MDL (National Archives, 2023) and results are reported in ng/L.

Wisconsin State Laboratory of Hygiene, Environmental Health Division, University of Wisconsin-Madison, Madison, Wisconsin

The Wisconsin State Laboratory of Hygiene (WSLH) utilizes their “PFAS in Water” method, developed at WSLH based on both EPA Method 537.1 and ISO/DIS 21675, to analyze for 33 PFAS (International Organization of Standardization, 2019; U.S. Environmental Protection Agency, 2020). Samples are stored chilled to 6°C for no more than 28 days before SPE. Prior to SPE, isotopically labeled standards are added. Samples are analyzed using HPLC-MS/MS in negative ion electrospray ionization mode. Additional method details are available upon request from WSLH (Alex Schwartz, Environmental Health Division, written commun., 2021). According to the WSLH standard operating procedures, MRL are set at 1 ng/L, or by multiplying the MDL by 2 and rounding up to the nearest calibration standard, whichever value is higher. Values between the MDL and MRL are reported as estimates. Results are reported in ng/L.

USGS Environmental and Public Health Microbiology Laboratory, St. Petersburg, Florida

The tapwater samples were analyzed for total coliforms, Escherichia coli (E. coli), and heterotrophic plate counts (HPC) by the USGS Environmental and Public Health Microbiology Laboratory (EPHML). Total coliforms were measured using Standard Method 9222B (Lipps and others, 2022b), membrane filtration with Endo Media. E. coli was measured using modified membrane-thermotolerant E. coli agar as described in EPA method 1603 (U.S. Environmental Protection Agency, 2014). HPC was measured using membrane filtration as described in Standard Methods 9215D (Lipps and others, 2022a). Results for all methods are reported in colony-forming units per 100 milliliters (CFU/100 mL).

USGS Organic Geochemical Research Laboratory, Lawrence, Kansas

Three 1-liter (L) baked glass-amber bottles filled with unfiltered water were collected at each site (labelled A, B, and C bottles) and extracted and concentrated to a total final volume of 300 microliters (µL). These extracted samples are archived at the OGRL.

The sample masses were calculated from the mass of the bottle before and after extraction. Each 1 L sample was extracted using a 24-port vacuum rack with Teflon tubing attached to 6 mL SPE cartridge adapters. The 200-milligram HLB SPE cartridges (Waters Corp., Milford, Massachusetts) were conditioned with 8 mL of methanol followed by 8 mL of water. The samples were loaded into the cartridges at a flow rate of about 10 mL per minute (min), the bottle and cartridge were rinsed with 8 mL of water, and air was passed through the cartridge for at least 20 minutes. A 24-port vacuum rack equipped with disposable polytetrafluoroethylene (PTFE) liners was used to elute the cartridges with two aliquots of 4 mL of methanol into 15 mL glass conical-bottom centrifuge tubes, and elution colors were recorded. The samples were evaporated to approximately 1 mL at 40°C, and the extracts from the B and C bottles were washed into the A bottle centrifuge tube with two aliquots of 1 mL of methanol. This combined extract was evaporated to 10 µL and then brought to a final volume of 300 µL with liquid chromatography/mass spectrometry (LC-MS)-grade methanol. The sample was transferred to a 12 by 32-millimeter plastic vial with an amber glass 0.1 mL insert and aluminum seal with PTFE/silicone liner lids. Extracts are archived at −20°C.

Inorganic Constituents

When considering the spatial distribution and aquifer comparison of inorganic constituents, little variation is apparent between most sites, but there are two notable exceptions: site 006 (the community-park drinking-water fountain at the northern, upstream tip of the island) and site 001 (UMESC tapwater sourced from the deeper Mount Simon aquifer; table 3 in Romanok and others, 2023a). The tapwater sample from site 006 had some of the highest concentrations of trace elements, including lead (8.5 μg/L), iron (8,600 μg/L), and manganese (104 μg/L). The tapwater sample from site 001 had some of the lowest concentrations of many inorganic constituents, including aluminum, barium, calcium, chloride, magnesium, manganese, nitrate, potassium, strontium, uranium, and iron. Trace elements, such as barium, strontium, and uranium are commonly an order of magnitude lower at site 001 when compared to other sites. The median concentration of barium and strontium from the 10 locations drawing from the alluvial aquifer were 54.5 μg/L (interquartile range=39.5–70.5 μg/L) and 86.5 μg/L (interquartile range=67.2–104 μg/L), respectively, but neither barium nor strontium were detected at site 001 (table 1.5). Similar results are reported for uranium. The median concentration of uranium in samples from the shallow aquifer locations was 0.255 μg/L (interquartile range=0.16–0.562 μg/L), and the median from site 001 was 0.002 μg/L (table 1.5).

Organic Compounds

Of the 74 organic compounds detected (table 1.4), the 4 most frequently detected compounds were didealkylatrazine, metolachlor SA, PFBA, and PFBS, with detections in at least half of the sites (table 1.4). The total number of compounds detected in each sample ranged from 0 to 20 (median number of compounds detected per site of 11), and the sum of concentrations ranged from nd to 2.53 μg/L (median sum of concentrations per site of 0.342 μg/L; fig 2 and table 1.6).

Pesticides were detected in samples from all sites, but the median of the sum of pesticide concentrations from the seven samples collected at site 001 was below the detection limit. The sum of pesticide concentrations per sample ranged from nd to 1.26 μg/L (median per site of 0.086 μg/L; table 1.6). Among the most frequently detected pesticides were metabolites of both metolachlor (metolachlor SA) and atrazine (didealkylatrazine; table 1.4). PFAS were detected at 7 of the 11 locations, with the total number of detections per sample ranging from 0–14 (median per site of 1; table 1.6). The sum of PFAS concentrations per sample ranged from nd to 0.661 μg/L (median sum of concentrations per site of 0.014 μg/L). Sites 002 and 004 had the most PFAS detections per sample with 14 (median of 13) and 7, respectively, and concentration sums of 0.609 (median) and 0.164 μg/L, respectively (fig. 3 and table 1.6) Concentrations of PFOA and PFOS exceeded both the EPA interim HAL and the proposed MCL at sites 002, 003, 004, and 005. Site 007 exceeded the EPA interim HAL for PFOS only.

Pharmaceuticals were detected at 5 of the 11 sites, with total detections per sample ranging from 0–1 (median per site of 0), and the sum of pharmaceutical concentrations per sample ranged from nd–0.036 μg/L (median per site of nd). SVOC were detected in 5 of the 11 sites, with the total number of detected compounds per sample ranging from 0–6 (median per site of 0), and the sum of concentrations per sample ranging from nd–2.2 μg/L (median per site of nd). Sites 004, 007, and 010 had a single SVOC detection at each location; however, site 006 had 6 detected SVOC (table 1.6). VOC were detected in 9 of the 11 sites; however, the median concentration of the 7 samples collected at site 001 was below the detection limit (nd). The total number of VOC detected in all samples ranged from 0–11 (median per site of 1), and the sum of concentrations per sample ranged from nd to 1.9 μg/L (median per site of 0.113 μg/L; table 1.6).

At site 006, VOC and SVOC were the most frequently detected compound groups (11 and 6 compounds detected, respectively) comprising 85 percent of detections. There were no PFAS detected in the sample (table 1.6). Sites 002 and 006 had the most organic compounds detected with 20 each. Residential tapwater sites 007 and 008 on the western edge of the island (fig. 1) had 19 and 13 organic compound detections, respectively. Sites 009, 010, and 011 had 4 or fewer detections of any organic compounds, and no PFAS were detected at any of these locations (table 1.6 and figs. 2 and 3). Site 001, the well that withdraws water from the Mount Simon aquifer, had the fewest number of detected organic compounds and the lowest sum of organic concentrations (tables 1.5 and 1.6).

Inorganic Constituents

Overall, there was little variation in inorganic concentrations in any of the three sites evaluated temporally during October 18–20, 2021. At site 001, water temperatures ranged from 17–22.1°C (median of 18.6°C), specific conductance ranged from 509–519 μS/cm (median of 512 μS/cm), and pH ranged from 7.1–7.8 (median of 7.4; table 1.2). Seventeen inorganic constituents were detected, with 5 (aluminum, chromium, hexavalent chromium, manganese, and phosphorus) only detected in the stagnant first-flush sample on day 1. Iron was only detected in the first-flush (37 μg/L) and post-6h (4 μg/L) samples on day 1 (table 3 in Romanok and others, 2023a). The maximum value for boron occurred during the first-flush sample on day 1 (84 μg/L), with the remaining concentrations for this site ranging from 13–20 μg/L (median for all boron samples at site 001 was 15 μg/L; table 1.2).

At site 002, water temperature ranged from 19.5–25.4°C (median of 24°C), specific conductance ranged from 420–428 μS/cm (median of 426 μS/cm), and pH ranged from 7–7.5 (median of 7.2; table 1.2). Of the 21 inorganic constituents detected, chromium, fluoride, hexavalent chromium, iron, and lead were only detected in the samples collected on the first day at this site and were generally at or near the analytical reporting level. With a few exceptions, all other inorganics were detected at concentrations that remained stable throughout the sampling. For example, boron was detected in every sample with concentrations ranging from 17–18 μg/L (table 1.2). One exception was manganese, with a concentration of 189 μg/L, which was the maximum detected concentration and occurred in the post-flush sample on day 2 (table 3 in Romanok and others, 2023a).

At site 003, water temperature ranged from 17.2–18.1°C (median of 17.8°C), specific conductance ranged from 728 to 731 μS/cm (median of 729 μS/cm), and pH ranged from 7.5–7.8 (median of 7.6; table 1.2). Twenty-four inorganic compounds were detected at site 003. Median concentrations were generally higher at site 003 than at any other site in the study and concentrations were stable throughout the three days of sampling with a few exceptions. For example, chloride was detected in every sample, with concentrations ranging from 75.8–77.4 mg/L (median of 76.3 mg/L; table 1.2).

One exception was between the first-flush and post-flush samples on day 1 for copper and iron, as the concentration in the first-flush sample for copper was 617 μg/L (the maximum detected concentration), and the post-flush was 30 μg/L (the minimum detected concentration). Concentrations generally remained stable throughout the rest of the samples (concentration range was 332–517 μg/L; table 3 in Romanok and others, 2023a). The iron concentration in the first-flush sample on day 1 was 13 μg/L (the minimum detected concentration), while in the post-flush sample on day 1 the concentration was 108 μg/L (the maximum detected concentration), which also remained stable for the rest of the sampling period (concentration range is 48–73 μg/L; table 3 in Romanok and others, 2023a).

Organic Compounds

Six organic compounds were detected at least once in the seven samples collected at site 001 during October 18–20, 2021: 1 pesticide, 1 PFAS, 2 pharmaceuticals, and 2 VOC (table 1.3). On the first day, 2–3 organic compounds were detected from the 3 samplings, with the sum of concentrations per sample ranging from 0.024–1.92 μg/L (median of 0.024 μg/L; table 1.6). These findings included one detection each of isopropyl alcohol (1.9 μg/L in the stagnant first-flush sample), flubendiamide (0.0005 μg/L), and caffeine (E [estimated] 0.005 μg/L) in the post-flush sample, and dehydronifedipine (E0.036 μg/L) in the post-6h sample (table 3 in Romanok and others, 2023a). Perfluorobutyrate (PFBA) was the only compound detected consistently on the first day, with concentrations ranging from nd–E0.025 μg/L (table 1.3). On the following days, only PFBA was detected on day 2 (stagnant first-flush sample), and only methyl acetate was detected on day 3 (post-6h sample). No SVOC were detected at site 001 (table 1.5).

At site 002, there was little variability over the course of the three days of sampling. Twenty-six total compounds were detected at least once in the seven samples. These included 9 pesticides, 14 PFAS, 1 pharmaceutical, and 2 VOC (table 1.3). On the first day, 18–20 compounds (median of 20) were detected in all three samplings; similar results were observed on the second (17 and 20 compounds detected) and third (16 and 20 compounds detected) days (table 1.6). The sum of concentrations per sample ranged from 0.931–1.28 μg/L (median of 1.08 μg/L; table 1.6). In each sample, 2–5 pesticides were detected, with the sum of concentrations per sample ranging from 0.029–0.302 μg/L (median concentration sum was 0.132 μg/L; table 1.6). The most frequently detected pesticides were metabolites of metolachlor, alachlor, and atrazine. No SVOC were detected at this site. Methylbenzotriazole and 1,1-difluoroethane were detected in every sample collected at site 002 (table 1.3), with concentrations ranging from E0.1–E0.141 μg/L (median of E0.123 μg/L) and from 0.208–0.262 μg/L (median of 0.238 μg/L), respectively (table 1.3).

Of the 32 PFAS analyzed, 12–14 (median of 13) were consistently detected at site 002 during the October 18–20, 2021 sampling period: PFBA, PFBS, PFHxS, PFOA, PFOS, perfluoro-n-pentanoate (PFPeA), perfluoro-1-pentanesulfonate (PFPeS), perfluoro-1-hexanesulfonamide (FHx-SA), perfluoro-n-heptanoate (PFHpA), perfluoro-1-heptanesulfonate (PFHpS), perfluoro-n-hexanoate (PFHxA), perfluorobutane sulfonamide (FBSA), 1H,1H,2H,2H-perfluorohexane sulfonate (6:2) (6:2 FTS), and perfluoro-n-nonanoate (PFNA; table 1.6 and table 1.3). This site had the most PFAS detections with the highest individual concentrations for those compounds compared to other locations (table 1.6). Eleven of the 14 PFAS were detected in all of the samples at this location; FBSA, 6:2 FTS, and PFNA were detected in 85.7, 71.4, and 57.1 percent of the samples, respectively (table 1.3). Of the detected PFAS, PFHxS and PFOS had the highest median concentrations with 0.158 μg/L and 0.156 μg/L, respectively (table 1.6). The sum of PFAS concentrations per sample ranged from 0.527–0.661 μg/L (median sum of concentration was 0.609 μg/L) (table 1.6).

Site 003 also showed little variability over the three days of sampling. Of the 19 total compounds detected at this site (6 pesticides, 5 PFAS, 2 pharmaceuticals, 2 SVOC, and 4 VOC, table 1.3), 9–12 compounds were detected at each sampling (median of 9), with the sum of concentrations per sample ranging from 0.165–2.46 μg/L (median sum of concentration was 0.226 μg/L; table 1.6). Of the six detected pesticide compounds, metabolites of metolachlor and atrazine were the most frequently detected (table 1.3). Bis (2-ethylhexyl) phthalate was only detected in the stagnant first-flush sample on day 1 (E2.2 μg/L). Four VOC were detected frequently: 1-Chloro-1,1-difluoroethane (HCFC-142b), 1,4-dichlorobenzene, butanal, and butane. 1,4-dichlorobenzene was detected consistently throughout the three days of sampling, with concentrations per sample ranging from 0.039–0.175 μg/L (median of 0.058 μg/L; table 1.3). Butane was detected in all samples on day 1. Butanal was only detected in the stagnant first-flush sample on each day and in the post-6h sample on day 1. HCFC-142b was not detected on day 1 but was consistently detected on subsequent days (table 3 in Romanok and others, 2023a).

Five PFAS were detected at site 003: PFBA, PFBS, PFHxS (sum of linear and branched isomers), PFOA, and PFOS (sum of linear and branched isomers). PFBS, PFHxS, and PFOA (except for one sample) were detected consistently over the course of three days (table 3 in Romanok and others, 2023a). PFBS concentrations ranged from 0.012–0.014 μg/L (median of 0.012 μg/L), PFHxS ranged from 0.002–0.004 μg/L (median of 0.003 μg/L), and PFOA ranged from nd–E0.005 μg/L (median of 0.005; table 1.3). PFBA and PFOS were each detected in 57 percent of the samples. PFBA ranged from nd–E0.013 μg/L (median of 0.012 μg/L) and PFOS concentrations ranged from nd–0.005 μg/L (median of 0.003 μg/L; table 1.3). The sum of all PFAS concentrations per sample ranged from 0.017–0.039 μg/L (median sum of concentrations was 0.033 μg/L; table 1.6).

For the long-term PFAS sampling at UMESC, 15 of the 33 PFAS were detected at least one time in any sample (table 1.7). Only PFBS (0.0006 μg/L) and FOSA (perfluoro-1-octanesulfonamide; 0.0005 μg/L) were detected (one time each) in the deep well (WR972) at concentrations near the method LOD and below the method quantitation limit (table 3 in Romanok and others, 2023b). The median number of detections ranged from 11.5–12 for the 3 wells in the shallow alluvial aquifer (BG122, BG123, and BG124), with the median sum of concentrations per sample ranging from 0.276–0.676 μg/L (table 1.7). With one exception (PFHpS was only detected in 87.5 percent of samples at site BG123), 10 PFAS were detected in every sample collected from these three wells (table 1.7). These results are consistent with the results collected at sites 001 and 002 in the 3-day temporal sampling during October 18–20, 2021.