Occurrence and concentrations of traditional and emerging contaminants in onsite wastewater systems and water supply wells in eastern North Carolina, USA

Onsite wastewater treatment systems (OWTSs) are commonly used in rural areas for treatment and dispersal of wastewater. OWTSs include a septic tank which separates solids and liquids and provides an environment for the anaerobic digestion of organic matter, thus reducing the total suspended solids and biochemical oxygen demand of the waste (Lusk et al. 2017). Septic tank effluent is piped to drain field trenches where the effluent infiltrates soil. Various physical, chemical, and biological processes may occur in the soil to remove pollutants in the wastewater. However, groundwater contamination may result if OWTSs are not properly designed, installed, and maintained (Lusk et al. 2017). After installation, OWTSs are operated and maintained by the home/property owner. Thus, it is the owner's responsibility to evaluate the functionality of their OWS, but many people may not be aware of the maintenance recommendations such as routinely pumping the tank to remove solids, resulting in little maintenance of the systems (Noss & Billa 1988). Also, many communities in North Carolina (NC) that rely on OWTSs also use private groundwater wells for their drinking water supply (Naylor et al. 2018). Like with OWTSs, once a well is installed it is the responsibility of the goods owner to operate and maintain the well (Lee & Murphy 2020). Well owners may not be aware of the harmful effects that pollutants in groundwater supplies may cause. Water from private groundwater wells in NC is not tested routinely unless the property owner requests and pays for the sample analysis (Naylor et al. 2018). These proactive water quality assessments may be cost-prohibitive for impoverished communities, many of which include a predominance of racial minorities, thus contributing to environmental justice (Gavino-Lopez et al. 2022) and health disparities concerns (Stillo & MacDonald 2017).

There are many different pollutants that can be present in groundwater supply wells that may result in negative health outcomes for water consumers. Examples include contaminants associated with wastewater that have traditionally been tested for by public water providers such as pathogens (or pathogen indicators) like Escherichia coli (E. coli) (Humphrey et al. 2011; Schneeberger et al. 2015) and compounds such as nitrate (Humphrey et al. 2010; Wigginton et al. 2018). Testing is important because the consumption of water or food with pathogenic E. coli may cause severe gastroenteritis, diarrhea, and sometimes death (Kosek et al. 2003; Estrada-Garcia et al. 2009). Human and animal wastes are sources of elevated concentrations of E. coli and other bacteria (Hynds et al. 2014). Studies have shown a significant correlation between fecal indicator bacteria in water resources and OWTS density (Borchardt et al. 2003; Humphrey et al. 2018). Therefore, effective waste management strategies including policies regarding septic system density and site condition requirements (vertical separation distances) for system installations are important (Humphrey et al. 2011, 2018; Cox et al. 2019). Also, proper siting, construction and sealing of wells are vital for reducing the likelihood of groundwater contamination with E. coli (Lee & Murphy 2020). The maximum contaminant level (MCL) for E. coli in water supplies is zero colony-forming units per 100 mL, thus any E. coli observed is considered a hazard (US EPA 2023). Total coliform bacteria have also been used as a potential indication of contamination of water supplies (Theng-Theng et al. 2007; Rosso et al. 2012). In a study of 16 water supply wells in South Bass Island Ohio, USA, Theng-Theng et al. (2007) reported that all the water samples from the wells were positive for total coliform and E. coli, while seven wells also tested positive for enterococci and Arcobacter. The study was conducted following multiple reports of gastroenteritis by visitors to the area. Theng-Theng et al. (2007) attributed the water contamination to malfunctioning wastewater treatment facilities and septic systems following heavy rainfall. Rosso et al. (2012) evaluated the presence of total coliform in newer (post 2008) and older (pre-2008) wells in central NC. Rosso et al. (2012) reported that 29% (10 of 35 wells) of older wells and 31% (11 of 35 wells) of newer wells sampled for the study were positive for total coliform. Stillo & MacDonald (2017) tested 171 drinking water wells in central NC and found 29% were positive for total coliform and 6.4% were positive for E. coli. These findings highlight the need for monitoring of drinking water supplies.

Elevated nitrate-nitrogen concentrations can cause methemoglobinemia or ‘blue baby syndrome’ in infants (Sadeq et al. 2008), and some evidence suggests that chronic consumption of nitrate can increase the risk of various cancers (Ward et al. 2005). The MCL for nitrate-nitrogen is 10 mg L⁻¹ (US EPA 2023). There are many different potential sources of nitrate in the environment including human waste, animal waste, fertilizers, automobile emissions, and natural decomposition of organic matter (Havlin et al. 1999). Prior research has shown a significant correlation between OWTS density and nitrate concentrations in groundwater (Naylor et al. 2018) and surface water (Hoghooghi et al. 2016). Humphrey et al. (2010) reported that nitrate concentrations in groundwater near OWTS exceeded 10 mg L⁻¹ in more than half of the 16 OWTS evaluated in a study conducted in the NC Coastal Plain. Naylor et al. (2018), summarizing nitrate data from private drinking water wells across NC, found that seven of the top 10 counties with the highest mean nitrate concentrations were in the Inner Coastal Plain region. Groundwater from wells in those counties had a mean nitrate concentration of 3.4 mg L⁻¹ and there was an average of over 25,000 OWTS in those 7 counties. Therefore, communities that rely heavily on OWTS and private wells, such as those in Eastern NC, may be at high risk with regard to elevated nitrate concentrations in groundwater.

Other pollutants that have more recently been screened for in water supplies include synthetic chemicals such as perfluorooctane sulfonate (PFOS), perfluorooctanoic acid (PFOA), and hexafluoropropylene oxide dimer acid and its ammonium salt (GenX) which may be considered contaminants of emerging concern (US EPA 2022). The PFOS, PFOA, and GenX chemicals are used in various products including cookware, fire foams, paints and varnishes, detergents, and contact paper (Coperchinin et al. 2017). Exposure to high concentrations of PFOA and PFOS may lead to decreased fertility, developmental delays in children, low birth weight, increased risk of some cancers, reduced immune system function, and higher risk of obesity (EPA 2022). Production with PFOS and PFOA at many industrial plants ended more than a decade ago. However, these compounds are persistent in the environment and can accumulate, thus continuing to cause public health threats (Pritchett et al. 2019) even long after production ceased. In fact, PFOS and PFOA have recently been found in water, soil, and air in many countries across the world (Pan et al. 2018; EPA 2022). Production using GenX chemicals instead of PFOA was initiated to better protect public health as GenX was thought to be less of a health risk (Choi et al. 2018). However, animal toxicity studies of GenX exposure have shown detrimental health effects related to the liver, kidney, and immune system, and development of cancers in the test animals (EPA 2022). Recent work has also shown that exposure to GenX may also negatively influence amphibian growth (Barragan et al. 2023) and fish immune system and liver functions (Guillette et al. 2020). Thus, GenX exposure may pose risks to humans and wildlife. The US EPA (2022) has issued a lifetime health advisory (HA) of 10 parts per trillion (ppt) for GenX chemicals. Communities in Eastern NC near an industrial chemical plant that produces GenX have been exposed to chemicals via legacy and current wastewater discharges and atmospheric emissions (Sun et al. 2016; McCord et al. 2018; Pritchett et al. 2019). For example, Cahoon (2019) reported that the industrial plant discharged effluent to the Cape Fear River for many years that had GenX concentrations that exceeded 500 ppt. Sun et al. (2016) and McCord et al. (2018) reported GenX concentrations in the finished water of a water treatment plant that exceeded 450 ppt. The raw water intake for the plant was in the Cape Fear River and downstream from an industrial manufacturer of GenX (located in Fayetteville within Cumberland County, NC). Pritchett et al. (2019), summarizing a study by NC health officials, reported that 207 drinking water wells within a 12 km radius of the industrial plant in Cumberland County had GenX concentrations that exceeded 140 ppt. Therefore, prior research has shown contamination of groundwater and surface waters near the industrial plant. Emissions of GenX into the atmosphere from the industrial plant have also impacted nearby communities. NC DEQ (2018) estimated that the annual air emissions of GenX from the industrial plant in Fayetteville NC could have sometimes exceeded 1,000 kg and that some rain samples collected within 10 km of the plant had Gen X concentrations between 45 and 60 ppt. Winds typically move in a northeasterly direction in NC due to the Gulf Stream (NOAA 2023), but wind patterns change based on atmospheric conditions. Research by Galloway et al. (2020) in Ohio and West Virginia near an industrial plant that produces GenX found that 38% of soil samples collected downwind of the plant had detectable concentrations of GenX and one sample contained 8.14 ppt. Roostaei et al. (2021) used a machine-learned Bayesian network model with groundwater and air quality data and other geographical/geological information from near the industrial plant in Fayetteville, NC and concluded that one of the most important factors regarding risk for GenX contamination in groundwater wells was the historic atmospheric deposition rate of GenX and the distance and orientation from the plant. Brandsma et al. (2019) reported concentrations of GenX (1 to 27 ppt) in or on all grass and leaf samples analyzed within 3 km of an industrial manufacturing plant. These studies have shown that atmospheric emissions of GenX from industrial plants may influence air, soil, and water quality in the surrounding areas. Exposure to GenX may also occur due to consumption after migration of the chemical from cookware and food packaging articles (Choi et al. 2018). Migration of GenX from household items into the waste stream (e.g., washing cookware, garbage grinders) may result in the discharge of GenX to subsoil and groundwater near OWTS. A study by Semerad et al. (2020) in the Czech Republic reported detectable concentrations of GenX in about 20% of sewage sludge samples with concentrations ranging from 0.3 to 1.2 ppt. However, the potential contribution of GenX to subsoil and groundwater via OWTS is unknown.

This project aims to provide field-based data on the presence and concentrations of traditional and emerging contaminants in wastewater from OWTS and groundwater wells in low-income communities of Eastern NC. To the authors' knowledge, this is the first study where wastewater samples from septic tanks and groundwater samples from water supply wells were collected and analyzed for GenX along with E. coli, nitrate, and total coliform.

Concentrations of nitrogen, E. coli, total coliform, and GenX in wastewater samples were compared to groundwater samples to determine if wastewater was influencing groundwater. Groundwater concentrations of GenX within 11 km of the industrial plant were compared to concentrations beyond 11 km of the industrial plant to determine if emissions from the plant may be an influencing factor with regard to GenX concentrations. It was anticipated that nitrate, E. coli, or total coliform in drinking water wells on the properties would more likely be related to OWTS discharge of effluent rather than industrial plant emissions. Comparisons of E. coli, total coliform, and nitrate within and beyond 11 km of the industrial plant were also made in case groundwater contamination at sites was influenced by faulty construction or maintenance of the wells that could facilitate the transport of all the contaminants (GenX, E. coli, total coliform, and nitrate) to water supply aquifers. Using parcel data in ArcGIS Pro, centroid points were created for the industrial plant and the study sites. The ‘Measure’ tool was used to determine the distance between centroid points of the home sites and plant. An Anderson-Darling test of normality was performed on each data set. When data exhibited a normal distribution, t-tests were used to determine if differences in concentrations between comparison groups were significantly different (p < 0.05). If data were not normally distributed, Mann Whitney non-parametric tests were used to determine significance. Spearman correlations were performed to determine if there were statistically significant associations between GenX concentrations in water supplies and distance from the industrial plant. Minitab 20 statistical software was used for the tests.

Also, GenX concentrations were significantly greater (p = 0.021) in water supplies located within 11 km (mean = 16.74 ppt) of the industrial plant relative to water supplies further away (mean = 0.69 ppt) (Figure 6). These findings agree with a recent study by Roostaei et al. (2021) that concluded that elevated concentrations of GenX in water supply wells were greatly influenced by atmospheric deposition rates of GenX and distance and orientation (relative to prevailing winds) relative to the industrial plant. Roostaei et al. (2021) found that wells sampled southwest or northeast of the industrial plant and within 11 km had a higher risk of exceedance of the HA for GenX. Pritchett et al. (2019) reported that 25% of groundwater wells sampled within a 12 km radius of the industrial plant in Fayetteville NC exceeded the HA for GenX. These results imply that water supplies especially within 11 km of the industrial plant pose the most risk of contamination with GenX likely because of historic atmospheric emissions sources. The industrial plant installed a packed bed scrubber and thermal oxidizer to improve air emissions in late 2019 which should reduce future atmospheric emissions (NC DEQ 2024).

Wastewater sampled from the septic tanks at the 18 sites had significantly higher (p < 0.001) concentrations of TDN and ammonium relative to groundwater (Table 2). This was expected as wastewater is enriched with organic matter from fecal material and food wastes generated and discharged from the homes (Lusk et al. 2017). The specific conductance and turbidity of wastewater were higher relative to the water supply at each site (Table 3). More specifically, wastewater samples had a mean turbidity and specific conductance of 263 FNU and 927 μS cm⁻¹ respectively, while water supplies were significantly (p < 0.05) lower at 2.8 FNU and 117 μS cm⁻¹. Wastewater has high concentrations of ions and solids which influence the conductance and clarity (Lusk et al. 2017; Humphrey et al. 2018). The pH of wastewater at the sites ranged from 5.5 at Site 4 to 7.47 at Site 18 (Table 3). The mean pH of wastewater (6.54) was elevated relative to the mean pH of the water supplies (5.36). The mean temperature of wastewater was also slightly elevated (18.4 °C) relative to the water supplies (17.7 °C) (Table 3).

Each of the septic tanks sampled at the volunteer sites were pumped to remove the solids and information regarding suggested routine maintenance of the OWTS was communicated to the owners. The soil types for the 18 sites were mostly dominated by sandy-textured, well-drained soils (Table 1). However, the soils at Sites 6 (Roanoke) and 12 (Leon) had indications of seasonal high wetness conditions (grey mottles) within 0.3 m of the OWTS trench bottoms and would not be suitable for a conventional style system using today's regulations in NC, unless water table monitoring revealed groundwater was deeper than soil morphological indicators suggested. During the time of the site visits though, groundwater was not within 0.6 m of the trenches at any of the sites. Four of the 18 OWTSs (22%) evaluated were exhibiting active signs of malfunction. The malfunctions included a leaking septic tank at Site 18 as indicated by a wastewater level approximately 0.3 m below the inlet pipe of the tank; encroachment of freeboard volume in septic tanks at Sites 12 and 7; and roots growing in a septic tank at Site 4 (Table 1). However, the 4 malfunctioning OWTSs did not seem to have a negative influence on well water quality as none of the water supplies at those sites contained detectable concentrations of GenX, E. coli, or total coliform, and the nitrate concentration was below 2 mg L⁻¹ at each one. The OWTSs at Sites 7 and 5 were exhibiting evidence of unequal distribution of septic tank effluent as the color and growth of grass were noticeably different over one trench relative to the others. The GenX concentration in the water supply at Site 5 was 31.7 ppt and thus exceeded the HA while wastewater at the site had a lower concentration of 25 ppt. Given the lower concentration of GenX in wastewater relative to groundwater, it could not be determined that wastewater from the OWTS was a source of GenX in groundwater at that location. Wastewater at Site 7 had a GenX concentration of 99.6 ppt, but GenX was not detected in groundwater, so GenX contributions from the OWTS to groundwater supply could not be confirmed at the site. Two of the OWTSs (Site 2 and 17) had recently received new drain field trenches due to hydraulic malfunctions within the past few years, but GenX was not observed in the water supply at those sites. Excessive solids layers were observed in the septic tanks at Sites 8 and 16, demonstrating a lack of routine maintenance, but GenX was not detected in the water supplies at those sites either. Therefore, most of the water quality data suggest that OWTSs were not influencing the water supply wells with regard to contaminants. Water supply wells are typically installed in deep, confined aquifers in Eastern NC (Lautier 2006) which may have reduced the likelihood of contamination from OWTS discharges to the shallow subsurface.

NC did not have a state-wide well program until 2008, thus water testing was not required even during the initial startup in many communities. Well water is now tested prior to first use for common contaminants such as nitrate, E. coli, and coliform bacteria. However, this research has shown that well water that has lower than the MCL concentrations of nitrate and indicator bacteria may still be a health hazard due to contaminants of emerging concern, such as synthetic chemicals like GenX. It is important that private drinking water wells and community water supplies in this region near the industrial plant are routinely monitored for these chemicals and filtration systems and/or alternative sources of water provided where needed to protect human health. The industrial plant in the community evaluated in this study has agreed to test water supplies in the area and based on the results, has provided alternative sources of water and/or provided advanced water treatment systems to some affected residences (NC DEQ 2024). The industrial plant now sends wastewater to a hazardous waste landfill rather than discharging effluent to the Cape Fear River (Cahoon 2019), and to reduce atmospheric emissions, they have installed a packed bed scrubber and thermal oxidizer (NC DEQ 2024). These efforts should help reduce future loadings of GenX to the environment, but contamination of groundwater is still present.

Despite four of 18 OWTSs showing active signs of malfunction, groundwater from the water supply wells did not seem to be influenced with regards to E. coli or nitrate at those sites. Removal mechanisms for E. coli including filtration, sorption, and die off and for nitrate including denitrification, anammox, and immobilization (Lusk et al. 2017; Wigginton et al. 2018) may have been active in the soil beneath the studied OWTSs, thus resulting in little to no negative health impacts for the occupants of the homes with regards to these pollutants. Persistent contaminants such as GenX were observed in wastewater at each of the 18 homes likely due to migration of GenX from cookware and food packaging. Therefore, the discharge of wastewater to the subsurface from homes contributes GenX to soil and potentially groundwater near those OWTSs. The encouragement by the State of NC to use shallower aquifers and surface waters for a water supply to reduce the trends of dewatering, saltwater intrusion, and declining water levels for the deeper aquifers (NC DENR 1998) may increase the risk of GenX exposure if shallow groundwater contains high concentrations of the chemical.

The goal of this study was to gain a better understanding of the presence and concentrations of traditional and emerging contaminants in OWTSs and water supply wells in some economically distressed communities of Eastern NC. Based on the results from sampling wastewater and water supplies from 18 sites, it was shown that 22% of water supply samples contained concentrations of GenX that exceeded the HA. None of the water supplies sampled had E. coli, while one of the 18 sites was positive for total coliform (1 MPN 100 mL⁻¹). Nitrate concentrations in the water supplies were all below 10 mg L⁻¹ and most were below 0.8 mg L⁻¹. The main threat to human health with regards to contamination of the water supplies was with the synthetic chemical GenX, which was elevated above the HA for most evaluated sites within 11 km of the industrial plant. GenX discharges via OWTS effluent to the subsoil were observed at all 18 sites. More research is suggested to assess the occurrence and concentration of GenX in soil and shallow groundwater near OWTSs. Monitoring for persistent compounds in groundwater is suggested especially for communities using OWTS and that rely on surficial or shallow groundwater for a water supply.

All relevant data are included in the paper or its Supplementary Information.