Introduction

Groundwater is the primary drinking water source in Delaware. Previous studies have documented the influence of human activities on the quality of this important resource through the detection of low concentrations (<1 microgram per liter [µg/L] or parts per billion [ppb]) of pesticides, volatile organic compounds, and other manmade chemicals in supply wells. These chemicals were detected in source waters of 30 Delaware public water-supply wells (fig. 1) sampled in previous monitoring studies completed in 2000 (Ferrari, 2002), 2008 (Reyes, 2010), and in this 2018 study. All 30 wells sampled are screened in the Columbia surficial aquifer (well-depth range was between 37 and 139 feet below land surface). Although concentrations of these organic compounds were generally low, all water samples had at least one compound detected, and most samples had multiple detections. These results demonstrate the influence of human activities on the quality of this important drinking water resource.

As part of the Source Water Assessment and Protection Program, which was created by Congress as part of the Safe Drinking-Water Act Amendments of 1996, State of Delaware officials have a continued interest in monitoring and in observing broad changes in the water quality of source waters since previous sampling events (2000 and 2008) detected low-level contamination in supply well samples. As part of this new sampling effort, per- and polyfluoroalkyl substances (PFAS) were included. PFAS, which are a large group of manmade chemicals that include perfluorooctanoic acid (PFOA), perfluorooctane sulfanate (PFOS), and hexafluoropropylene oxide dimer acid (HFPO-DA or Gen X), have come to the attention of researchers and risk managers because of their persistence, toxicity, and number of adverse effects in humans and laboratory animals (U.S. Environmental Protection Agency, 2018). Although there are many chemicals included in the PFAS group and the U.S. Environmental Protection Agency (EPA) has not established a maximum contaminant level (MCL) for all of them, the EPA did establish a lifetime drinking-water health advisory level (HAL) of 70 parts per trillion (ppt; or 0.07 ppb) for PFOA and PFOS in May 2016, as these have been the most extensively produced and studied PFAS (U.S. Environmental Protection Agency, 2016a). Both chemicals are potentially of concern to human health in the Coastal Plain aquifer and warrant continued study of their and other PFAS distribution in groundwater. Although combined concentrations of PFOA and PFOS were compared to the EPA’s HAL, these comparisons were made for informational purposes and not for evidence of compliance or noncompliance with Federal regulations. The EPA HAL is a health-based reference level for public drinking water as supplied to customers and is not applied to source (raw) water. Groundwater-age data were collected as part of this study by averaging the laboratory sulfur hexafluoride (SF₆) ages estimated from each of two samples collected from each water-supply well and were compared for sites sampled in 2000, 2008, and 2018 to document any groundwater-age changes.

Description of Study Area

The study area falls entirely in the Delaware portion of the Atlantic Coastal Plain Physiographic Province and all wells sampled are screened in the Columbia aquifer (Benson and others, 1986; Reyes, 2010). The Columbia aquifer consists primarily of sands and gravels of fluvial and marginal marine origin and is an important drinking water resource in the Atlantic Coastal Plain of Delaware. Since the Columbia aquifer is largely unconfined, it is susceptible to contamination from spills, unmanaged wastes, and the application of agricultural chemicals on or near the land surface.

Agriculture is the predominant land use in Delaware with corn, soybeans, and small grains as the major crops. In 2017, approximately 37 percent of the available land area was used for farming (U.S. Department of Agriculture, 2017). Most of the agricultural activity in Delaware takes place in Kent and Sussex Counties and south of the Chesapeake and Delaware (C&D) Canal in New Castle County. The primary land uses surrounding this relatively shallow public-supply well network in the Coastal Plain of Delaware include agriculture, and low-density urban and suburban areas associated with small towns and communities (fig. 1).

PFAS General Description

What are PFAS?

Per- and polyfluoroalkyl substances (PFAS) are a large group (more than 4,000) of man-made, highly persistent chemicals that have been in use since the 1940s. PFAS are found in a wide range of consumer and industrial products such as cookware, pizza boxes, carpets, stain repellants, and firefighting foams, as well as other common products (fig. 2). PFAS are released into the air, soil, and water by PFAS manufacturing and processing facilities (such as metal plating operations), historical use in firefighting foams, and regular everyday use by consumers. Owing to their widespread use, persistence in the environment, and long half-lives, most people in the United States have been exposed to PFAS. There is evidence that because of their long half-lives and toxicity, continued exposure above threshold levels for specific PFAS may lead to adverse health effects (U.S. Environmental Protection Agency, 2016a; Agency for Toxic Substances and Disease Registry, 2018). PFOA and PFOS have been the most extensively produced and studied of these chemicals. Both chemicals are very persistent in the environment and human body, and can bioaccumulate in living organisms (U.S. Environmental Protection Agency, 2017).

There are a variety of ways in which people can be exposed to PFAS and at different levels of exposure, for instance

• Food ingestion and food packaging

• Daily exposure with commonly used items such as, but not limited to, carpets, cookware, rain gear, and so forth

• Occupational exposure

• Drinking water

Study Methods

This study was developed to obtain information on the spatial occurrence and distribution of PFAS in the unconfined aquifer in the Delaware Coastal Plain. Source water samples collected for this purpose were analyzed for PFAS and gaseous SF₆ to determine groundwater ages.

PFAS Results

All 30 wells were analyzed for 18 PFAS using the modified EPA Method 537 (U.S. Environmental Protection Agency, 2009; table 2). More than half (16) of the sampled wells had one or more PFAS detections, and as many as eight PFAS were detected in two wells (Ri23-15 and Oi25-18; table 3). The spatial distribution of PFAS detections across Delaware is shown in figure 3. Wells with higher numbers of compounds detected (five or more) were located in New Castle and Sussex Counties. Forty-four percent of the analyzed PFAS (8 of 18) were detected in the study well network, with individual compound concentrations detected raging from E5.0 to 130 ppt. The four most frequently detected PFAS were PFOA with 47 percent detection, PFHxA with 33 percent detection, and PFOS and PFHxS with 27 percent detection each (table 3).

Most of the PFAS detections were in oxic environments (dissolved oxygen levels greater than or equal to 1 mg/L), although higher concentrations of PFAS compounds (greater than or equal to 100 ppt) were observed in low-oxygen environments (fig. 4).

Two wells (Cd52-40 in New Castle County and Ri23-15 in Sussex County), out of the 14 wells with combined PFOS and PFAS detections, were above the EPA drinking-water HAL of 70 ppt (fig. 5). Both wells are treated to remove PFAS from their system.

Groundwater Age Results

Water that infiltrates the landscape and percolates downward toward the water table becomes recharge to the aquifer system. As additional recharge continues to enter the aquifer, older recharge is pushed deeper by the newer recharge, resulting in a trend of increasing groundwater age with depth. Groundwater age represents the average age of water withdrawn from the well, as well screens typically span several feet of aquifer sediment, thus integrating waters of various ages. Dissolved-gas analyses also were performed on samples of each well as part of the age-dating analysis to determine the average recharge temperature of the water in the aquifer (appendix 2; table 2.1). Recharge dates for groundwater samples were estimated based on measured concentrations of SF₆, which is a common chemical refrigerant that has been released into the environment. The reported age in this study was calculated by averaging the laboratory SF₆ ages estimated from each of two samples collected from each well (appendix 3; table 3.1).

The average groundwater age among the 30 wells sampled in 2018 ranged from 8.2 to 45.8 years, with a median of 25.7 years. Refer to table 1 for well-construction and groundwater-age data. The same relations from previous study years were observed; groundwater ages are positively correlated (R²=0.6024, p-value<0.0004) with well depth (fig. 6A) and negatively correlated (R²= –0.4293, p-value <0.018) with dissolved oxygen concentrations (fig. 6B) in the Columbia aquifer. The relation of groundwater-age data and total PFAS concentrations is shown in figure 7. This groundwater age-concentration relation supports the fact that PFAS are as result of anthropogenic activities and that these contaminants are relatively novel.

Comparison of the average modeled groundwater sample-age data (fig. 8A) from the 23 wells sampled in 2018 indicates that the average groundwater age was significantly different than the 2008 sample-age data; 2018 sample ages were, on average, 5 years older than 2008 sample ages. A similar pattern was found during cycle two (2008) of this study, where the 2008 average modeled groundwater age was, on average, 7 years older than the samples collected in 2000. The distribution of the groundwater sample ages among the 17 trend wells and during the three study cycles (fig. 8B) indicates that sample-age medians were statistically different than zero (p-value<0.00124). Well-water sample-age data show a slight increase in groundwater sample age. Groundwater age represents a mixture of older and young groundwater because well screens in public-supply wells are relatively long and pumping enhances the mixing of waters of disparate groundwater ages (Dunkle and others, 1993). The specific reason for the differences in groundwater ages is unknown, although Gholam and others (2006) indicated that this age difference could reflect a higher proportion of water drawn from slow-moving storage, possibly as a result of pumping for long periods of time, or a pumping increase that caused more water from older, deeper, or more distant sources to be drawn into the well.

Summary

This study’s results mark a baseline for per- and polyfluorinated alkyl substances (PFAS) concentration in the Columbia aquifer. Many findings from this study suggest the need to continue active monitoring for these anthropogenic contaminants. There was a widespread distribution of PFAS detections throughout Delaware, as well as a significant number of detections per wells and the variety of compounds detected, including the higher concentrations and high number of detections in younger waters and the results of the U.S. Environmental Protection Agency’s health advisory level exceedances for perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) in source water samples. These findings show the susceptibility of this aquifer for these and other anthropogenic contaminants which may significantly affect the drinking water sources in the Columbia aquifer as a result of the aquifer’s characteristics (surficial aquifer; good soil drainage; and sandy, transmissive aquifer sediments) and the persistence of these compounds. Comparison of the average modeled groundwater ages indicates that the 2018 samples were significantly older than the 2008 and 2000 samples. Among the resampled wells, the average groundwater age difference was 7 years from 2000 to 2008 and 5 years from 2008 to 2018.