The goal of this study was to gain a better understanding of the PO4-P treatment efficiency of onsite wastewater systems (OWS) installed in nutrient-sensitive watersheds of the North Carolina Piedmont. Four OWS including two conventional and two single-pass sand filter (SF) systems were evaluated at sites with clay-rich soils. Piezometers were installed near all of the OWS, and down-gradient from the conventional OWS for groundwater collection and characterization. Septic tanks, groundwater, SF effluent, and surface waters were sampled each season during 2015 (five times) and analyzed for PO4-P and Cl concentrations and for various environmental parameters. The conventional and SF OWS reduced PO4-P concentrations by an average of 99% and 90%, respectively, before discharge to surface waters. Mass-load reductions of PO4-P were also greater for the conventional OWS (mean 95%), relative to SF (83%) systems. The effluents discharged by SF OWS were influencing surface water quality. Additional treatment of the effluent from single-pass SF with reactive media is suggested, along with monitoring of the final effluent for PO4-P concentrations. This research provides important information that is absent from the published literature concerning PO4-P contributions to water resources from OWS in clay soils.

INTRODUCTION

Onsite wastewater systems and phosphorus-enriched waters

Onsite wastewater systems (OWS) are a common method of wastewater treatment in many countries including the USA (United States Environmental Protection Agency 2002), Canada (Harman et al. 1996), and Ireland (Gill et al. 2009). Wastewater treated by these OWS contains high concentrations of various environmental contaminants including eutrophication-stimulating nutrients like phosphorus (Zanini et al. 1998; United States Environmental Protection Agency 2002; Robertson 2008). Septic tank effluent has total phosphorus concentrations that typically range from 2 to 12 mg/L (United States Environmental Protection Agency 2002) most of which is reactive orthophosphate (PO4-P) (Robertson 2008; Humphrey et al. 2015). Concentrations of PO4-P several orders of magnitude lower than septic effluent can stimulate primary productivity in some waters (National Oceanic and Atmospheric Association 1996). For example, in Florida the total phosphorus threshold for streams ranges from 0.06 to 0.49 mg/L, depending on geographic region (Florida Department of Environmental Protection 2013). Excess PO4-P loadings have caused algal blooms, eutrophic conditions, fish kills and/or water use impairment for water resources in China and Japan (Conley et al. 2009), throughout the European Union (Arnscheidt et al. 2007; Holman et al. 2008), and in the USA (Valiela & Costa 1988; Reay 2004; Conley et al. 2009), making PO4-P management a global environmental issue.

In the central Piedmont and eastern regions of North Carolina, major water resources including Falls Lake, Jordan Lake, High Rock Lake, the Tar River and some coastal waters are impaired due to excess nutrient loadings (Conley et al. 2009; Mallin & McIver 2012; North Carolina Division of Water Resources 2015). Regulations have been and are being enacted to lessen the phosphorus delivered from agricultural and urban runoff, and from wastewater treatment plants in many of these nutrient-sensitive water bodies (North Carolina Division of Water Resources 2015). Research has shown that in some regions, effluent from OWS may be a significant source of PO4-P in groundwater (Robertson 2008; Humphrey & O'Driscoll 2011; Humphrey et al. 2014, 2015) and surface waters (Valiela & Costa 1988; Arnscheidt et al. 2007; Mallin & McIver 2012; Humphrey et al. 2015). However, most of the prior research regarding OWS and PO4-P transport in North Carolina has been focused in the coastal plain (Mallin & McIver 2012; Humphrey et al. 2014, 2015) where groundwater is relatively shallow and the soils are sandy. Groundwater depths are typically greater in the Piedmont relative to the coastal plain, and Piedmont soils have a higher clay content (Daniels et al. 1999). These factors (clay content and vadose zone thickness) may greatly influence PO4-P removal beneath OWS (Karathanasis et al. 2006; Robertson 2008; Humphrey & O'Driscoll 2011), so PO4-P treatment by OWS in the coastal plain and Piedmont may differ. More information is needed regarding the contributions of PO4-P from OWS to water resources to determine if mitigatory strategies should include reducing PO4-P contributions from OWS, in particular for the Piedmont regions where most of the state's population is centered and OWS are commonly used.

Onsite wastewater system technologies and phosphorus treatment

There are approximately two million active OWS in North Carolina and most are conventional style (Hoover 2004). Conventional-style OWS typically have four basic components including a septic tank, drainfield trenches, soil beneath the trenches, and setback distances from the OWS trenches to surface waters (Figure 1(a)). The septic tank provides primary treatment of wastewater via settling and retention of solids (United States Environmental Protection Agency 2002; Appling et al. 2013). Liquid effluent leaves the tank and enters the drainfield trenches via pipe. Septic tank effluent is stored in the trenches until it infiltrates the soil. The soil beneath the drainfield trenches should be aerobic, thus allowing for unsaturated flow conditions, and enabling treatment of phosphorus and other wastewater contaminants (Postma et al. 1992; Zanini et al. 1998; Reay 2004; Del Rosario et al. 2014; O'Driscoll et al. 2014). Phosphorus treatment processes in soil may include mineral precipitation, adsorption, and immobilization (Postma et al. 1992; Harman et al. 1996; Robertson et al. 1998; Zanini et al. 1998; Westholm 2006; Humphrey et al. 2014). Precipitation occurs when PO4-P forms solid minerals with Al, Ca, or Fe cations in the soil or filter media. Precipitation reactions are influenced by the concentration of cations, redox potential and pH of the subsurface environment (Robertson et al. 1998; Zanini et al. 1998; Humphrey et al. 2015). Adsorption occurs when PO4-P species bind to surface cations attached to soil particles (Westholm 2006; Robertson 2008). Immobilization of PO4-P may occur via uptake by riparian vegetation down-gradient from OWS (Osmond et al. 2002). Setback distances (typically ≥15 m) from the OWS to streams allow additional time and opportunity for further treatment of wastewater contaminants and for dilution and dispersion in the groundwater system (Robertson et al. 1998; Humphrey et al. 2014).
Figure 1

(a) Conventional OWS and (b) SF onsite system.

Figure 1

(a) Conventional OWS and (b) SF onsite system.

In locations where useable soil depth is not sufficient for a subsurface discharge OWS, then surface discharge systems such as sand filters (SF) may be permitted. SF are commonly used in portions of the Falls Lake Watershed where expansive clay mineralogy does not allow for subsurface dispersal of septic tank effluent (North Carolina Department of Environmental Quality 2009). SF systems include a septic tank, effluent dispersal pipe at the top of the filter, sand media within the filter, and effluent discharge pipe at the bottom of the filter (Figure 1(b)). The sand in the filter provides the media and environment for septic tank effluent treatment (Nielsen et al. 1993). The SF effluent then discharges into a surface water such as a stream via the discharge pipe.

Conventional- and SF-style OWS are commonly used in the Piedmont of North Carolina, but there is a lack of published research with regards to their effectiveness in treating PO4-P. The goal of this project was to help fill that research need by evaluating the influence that OWS have on contributing PO4-P to groundwater and surface waters in the Piedmont of North Carolina, where problems with excess nutrients in water resources continue to be a major environmental issue.

METHODS

Site selection

Four volunteered sites with OWS in Wake (one site) and Durham counties (three sites) in North Carolina, USA, were instrumented with piezometers for this study (Figure 2). These sites were chosen because portions of Wake and Durham counties drain to the nutrient-sensitive waters of Falls Lake and Jordan Lake. Also, Durham and Wake counties are two of North Carolina's largest counties and they rely heavily on OWS. For example, more than 4,900 operation permits for new OWS were issued between 2006 and 2010 (North Carolina Department of Health and Human Services 2015) in the two counties. Characteristics of the OWS, soils, and geology for the sites are shown in Tables 13.
Table 1

OWS characteristics for the five sites including system type, tank capacity, drainfield or SF area, system age, and number of occupants in home/building served by the system

LocationSystem typeTank capacity (L)Drainfield/Filter area (m2)System age (yrs)Occupants
Site 100 Conventional 3,780 90 18 
Site 200 Conventional 3,400 30 85 
Site 300 SF 3,400 30 47/1 
Site 400 SF 3,780 36 42 
LocationSystem typeTank capacity (L)Drainfield/Filter area (m2)System age (yrs)Occupants
Site 100 Conventional 3,780 90 18 
Site 200 Conventional 3,400 30 85 
Site 300 SF 3,400 30 47/1 
Site 400 SF 3,780 36 42 
Table 2

Soil characteristics of the five sites including particle size distribution, pH, effective cation exchange capacity (ECEC), and humic matter percentage

LocationUSDA soil seriesSubsoil clay/silt/sand %Subsoil pHSubsoil ECEC (cmol/kg)Subsoil humic matter %
Site 100 Cecil 65/9/26 5.5 4.1 0.04 
Site 200 Georgeville 42/37/21 6.3 6.7 0.09 
Site 300 Chewacla 20/49/31 5.6 17.5 0.04 
Site 400 Chewacla 25/23/52 0.04 
LocationUSDA soil seriesSubsoil clay/silt/sand %Subsoil pHSubsoil ECEC (cmol/kg)Subsoil humic matter %
Site 100 Cecil 65/9/26 5.5 4.1 0.04 
Site 200 Georgeville 42/37/21 6.3 6.7 0.09 
Site 300 Chewacla 20/49/31 5.6 17.5 0.04 
Site 400 Chewacla 25/23/52 0.04 
Table 3

Geological characteristics of the five sites including elevation, slope, groundwater depth, hydraulic gradient, and geologic parent material

LocationElevation (m) near systemLand % slopeGroundwater depth (cm) near OWSHydraulic gradientHydraulic conductivity (cm/day)Geologic parent material
Site 100 114 11 420 0.007 204 Felsic Crystalline 
Site 200 144 190 0.053 152 Slate Belt 
Site 300 95 60 NED Triassic Basin 
Site 400 91 90 0.014 Triassic Basin 
LocationElevation (m) near systemLand % slopeGroundwater depth (cm) near OWSHydraulic gradientHydraulic conductivity (cm/day)Geologic parent material
Site 100 114 11 420 0.007 204 Felsic Crystalline 
Site 200 144 190 0.053 152 Slate Belt 
Site 300 95 60 NED Triassic Basin 
Site 400 91 90 0.014 Triassic Basin 

NED: Not enough data.

Figure 2

Maps showing the four research sites in Durham and Wake counties, North Carolina, USA. Sites 100 and 200 are conventional systems with piezometers installed near the drainfield and 15 and 35 m down-gradient from the systems. Sites 300 and 400 are SF systems with piezometers installed near the SF and away from the filters.

Figure 2

Maps showing the four research sites in Durham and Wake counties, North Carolina, USA. Sites 100 and 200 are conventional systems with piezometers installed near the drainfield and 15 and 35 m down-gradient from the systems. Sites 300 and 400 are SF systems with piezometers installed near the SF and away from the filters.

Groundwater monitoring network

Piezometers were used for monitoring groundwater at each volunteered site. Piezometers were nested at different depths to determine if PO4-P concentrations were elevated in certain sections of the water column down-gradient from the conventional-style OWS. Boreholes were created for piezometer installations using soil augers and/or a truck-mounted Geoprobe. The boreholes were drilled to depths below the water table. The soils from the boreholes were laid on a tarp to characterize the profiles and for comparison to the USDA mapped series for the sites. Soil samples were collected from depths beneath the trench/filter bottoms and analyzed for physical and chemical properties including particle size distribution, pH, ECEC, and humic matter content. Particle size distribution was determined by the hydrometer method and the NC Agronomics laboratory in Raleigh, NC, performed the analyses of pH, ECEC, and humic matter percentage (Table 2).

Piezometers were constructed using either 3.18 cm or 5.08 cm PVC pipe connected to a well screen. Once the piezometers were installed in the boreholes, sand was poured in the annular space surrounding the well screen. Bentonite and borehole soil was used to fill the remainder of the annular space to the top of the casing. Piezometers were installed near (<1.5 m) the drainfields of the conventional OWS and at varying distances (5 to 35 m) down-gradient towards surface waters. There were 22 piezometers installed at the conventional-style OWS sites (11 at Site 100 and 11 at Site 200). Piezometers (five total) were also installed at Site 300 (two piezometers) and Site 400 (three piezometers) near the SF and in at least one location away from the filter. It was anticipated that most effluent would discharge through the outlet pipe of the SF, however, piezometers were installed near the SF to determine if some tank effluent infiltrates the soil in the SF and influences groundwater quality. Valve boxes were placed over piezometers and installed level with the ground surface to help protect the piezometers and to limit their intrusiveness. Piezometers and valve boxes were assigned unique labels.

A laser level and receiving rod were used to determine the relative elevation of the piezometer casings. Distances were recorded from piezometer to piezometer and also from piezometers to other landmarks visible on the aerial photographs of the sites. Site maps were constructed using the survey information (Figure 2). The elevations, groundwater depths, and piezometer locations were used with the three-point solution method (Domenico & Schwartz 1998) to determine groundwater flow direction and hydraulic gradient. Slug tests (Domenico & Schwartz 1998) were conducted to determine the hydraulic conductivity (K) of the soils for the subsurface discharge OWS. Geological data from the sites are displayed in Table 3.

Water quality monitoring

Nutrient, physical and chemical analyses of water samples

Water sampling (groundwater, septic effluent, SF effluent, and surface water) at the sites was conducted five times over the course of the year and at least once during each season (March, May, July, September, and November 2015). Depth to groundwater was measured first at each piezometer using a Solinst Temperature, Water Level, and Conductivity Meter (Solinst Canada Ltd, Georgetown, ON, Canada), then a new disposable bailer was lowered into the piezometer to purge the water and to allow for recharge to occur. Next, groundwater samples were withdrawn from the piezometer and transferred to sample bottles and a calibration cup for environmental parameter determination. Environmental readings including pH, temperature, specific conductance (SC), dissolved oxygen (DO), and the oxidation-reduction potential (ORP) were measured in the field using a YSI-556 Multiprobe Meter (Yellow Springs Incorporated, Yellow Springs, OH, USA). Samples were also collected from the septic tanks and adjacent surface water at each site and environmental readings were performed on those samples with the YSI 556. Sample bottles were labeled with the sampling point identification and placed in ice-filled coolers for transport back to the laboratory. The samples were filtered using a 0.7 μm pore size, and analyzed for PO4-P and Cl concentrations in the East Carolina University Environmental Research Laboratory using a SmartChem170/200. Detection limits were 0.0014 mg/L for PO4-P, and 0.05 mg/L for Cl.

Onsite wastewater system phosphate treatment efficiency determination

Concentrations of PO4-P in septic tank effluent were compared to groundwater concentrations near and down-gradient from OWS trenches and surface waters at Sites 100 and 200 (conventional OWS). The PO4-P concentrations from the septic tanks, SF effluent, groundwater near the SF, and surface waters were compared at the SF sites (Sites 300 and 400). These comparisons were made to evaluate whether the OWS were influencing water resources. Mann–Whitney tests were used to determine if there were statistically significant (p < 0.05) differences between comparison groups. The statistical program Minitab 16 was used for the analyses.

The treatment efficiency of each OWS was determined by comparing the concentrations of PO4-P in septic tank effluent to groundwater down-gradient from the drainfields or at the outlet of the SF. The Cl/PO4-P ratios of wastewater, and groundwater down-gradient from the OWS were analyzed to determine if concentration reductions were due to dilution or other processes. Chloride has been used as a conservative tracer to help analyze nutrient concentration reduction processes in several prior studies (Bradshaw & Radcliffe 2013; Del Rosario et al. 2014; O'Driscoll et al. 2014). If there is no change in the Cl/PO4-P ratio from septic tank effluent to groundwater, then dilution is the most likely cause for any decreases in PO4-P concentrations. A two-component mixing model (O'Driscoll et al. 2014) using the septic effluent and background groundwater Cl concentration was used to determine the percentage wastewater and background groundwater at specific down-gradient sampling locations. The mixing model is shown in Equation (1).
formula
1
With an estimate of wastewater and background groundwater fraction for each sampling location, a predicted concentration of PO4-P can be calculated for each piezometer and compared to the observed PO4-P concentration. The predicted value is based on the assumption that all reductions in PO4-P concentrations are because of dilution. The difference in predicted and observed PO4-P concentrations provides an estimated reduction of PO4-P mass via adsorption, precipitation, and/or immobilization.

RESULTS AND DISCUSSION

Phosphate treatment

Site 100 (conventional onsite wastewater system)

Septic effluent PO4-P concentrations were significantly higher than groundwater PO4-P concentrations near the drainfield (p = 0.0015), 15 m down-gradient (p = 0.0122), and 35 m down-gradient (p = 0.0025) from the OWS (Figure 3). The mean PO4-P concentrations of groundwater near the drainfield and down-gradient from the OWS were all less than 0.02 mg/L and more than 99% lower than septic effluent (6.60 mg/L) (Figure 3). No significant differences (p > 0.05 for all comparisons) were observed between groundwater PO4-P concentrations near the drainfield relative to groundwater 15 m and 35 m down-gradient from the OWS, and in the lake. The Cl/PO4-P ratios show two orders of magnitude difference between septic tank effluent and groundwater down-gradient from the OWS, thus confirming that significant attenuation of PO4-P mass occurred at this site (Table 4). More specifically, the mixing model results show that greater than 98% of the PO4-P mass is likely lost along the groundwater flow path (Table 4). Therefore, the OWS was very efficient at reducing the concentration and mass of PO4-P at this site. PO4-P treatment via adsorption was most likely enhanced by the relatively thick vadose zone (>3 m) and high clay content (65%) of the soil beneath the drainfield trenches (Karathanasis et al. 2006; Gill et al. 2009; Humphrey et al. 2015) and the abundance of red (4YR 4/6) iron coatings on the Cecil (Clayey, kaolonitic, thermic Typic Kanhapludults) series soil (Daniels et al. 1999).
Table 4

Ratios of PO4-P to Cl in wastewater and in groundwater near the drainfield (DF) and distances down-gradient (15 to 35 m)

Site 100Cl (mg/L)Fraction WWFraction GWPredicted PO4-P (mg/L)Observed PO4-P (mg/L)Cl/PO4-P ratioPO4-P mass reduction
Tank 42.3  6.06  
Drainfield 12.4 0.23 0.77 1.408 0.006 3,114 0.996 
15 m 14.8 0.29 0.71 1.770 0.009 1,637 0.995 
35 m 7.2 0.08 0.91 0.502 0.009 840 0.982 
Background 3.6  0.019   
Site 200Cl (mg/L)WWGWPredicted PO4-P (mg/L)Observed PO4-P (mg/L)Cl/PO4-P ratioPO4-P mass reduction
Tank 50.7  7.78 6.5  
Drainfield 26.3 0.47 0.53 3.661 0.061 520 0.983 
5 m 7.4 0.06 0.94 0.474 0.024 314 0.949 
15 m 6.9 0.05 0.95 0.396 0.029 241 0.926 
35 m 7.0 0.05 0.95 0.396 0.025 386 0.936 
Background 4.6  0.008   
Site 100Cl (mg/L)Fraction WWFraction GWPredicted PO4-P (mg/L)Observed PO4-P (mg/L)Cl/PO4-P ratioPO4-P mass reduction
Tank 42.3  6.06  
Drainfield 12.4 0.23 0.77 1.408 0.006 3,114 0.996 
15 m 14.8 0.29 0.71 1.770 0.009 1,637 0.995 
35 m 7.2 0.08 0.91 0.502 0.009 840 0.982 
Background 3.6  0.019   
Site 200Cl (mg/L)WWGWPredicted PO4-P (mg/L)Observed PO4-P (mg/L)Cl/PO4-P ratioPO4-P mass reduction
Tank 50.7  7.78 6.5  
Drainfield 26.3 0.47 0.53 3.661 0.061 520 0.983 
5 m 7.4 0.06 0.94 0.474 0.024 314 0.949 
15 m 6.9 0.05 0.95 0.396 0.029 241 0.926 
35 m 7.0 0.05 0.95 0.396 0.025 386 0.936 
Background 4.6  0.008   

Mix model parameters for Site 100 and 200 including septic effluent and background groundwater Cl and PO4-P data. Mass reduction of PO4-P calculated using a two-component mixing model with septic effluent and background groundwater Cl as the two sources of Cl and PO4-P.

Figure 3

Phosphate concentrations at Site 100 and Site 200 monitoring locations including in groundwater near the drainfield and up to 35 m down-gradient from the systems, and in surface waters.

Figure 3

Phosphate concentrations at Site 100 and Site 200 monitoring locations including in groundwater near the drainfield and up to 35 m down-gradient from the systems, and in surface waters.

Specific conductance was highest for septic effluent (617 μS/cm) and was elevated beneath the OWS drainfield trenches (138 μS/cm) and down-gradient from the OWS (84 to 122 μS/cm) relative to water from the background well (74 μS/cm) and lake (49 μS/cm) (Table 5). Therefore, effluent from the OWS was influencing the conductance of groundwater. Several prior studies have shown that specific conductance of wastewater is elevated (Appling et al. 2013; Del Rosario et al. 2014; O'Driscoll et al. 2014) and thus may be a good indicator of wastewater-impacted groundwater. The elevated specific conductance of groundwater down-gradient from the OWS helps provide evidence that wastewater-impacted groundwater was sampled.

Table 5

Mean environmental readings including depth to water (DTW), temperature (Temp), specific conductance (SC), DO, pH, and ORP

 DTW (cm)Temp (C°)SC (μS/cm)DO (mg/L)pHORP (mV)
Site 100 
 Tank  20.5 617 1.5 6.9 −325 
 Drainfield 409 17.5 138 3.5 5.9 34 
 15 m 180 17.7 84 5.4 45 
 35 m 50 20 122 −94 
 Lake 30 19.8 49 5.9 6.5 −49 
 Background  20.6 74 6.1 6.2 −20 
Site 200 
 Tank  20.5 765 1.4 6.5 −302 
 Drainfield 160 18.1 280 3.2 5.9 45 
 5 m 120 16.9 120 4.5 5.4 155 
 15 m 100 17.8 108 5.7 5.7 104 
 35 m 50 17.3 174 2.9 5.9 −42 
 Creek  17.5 169 6.2 6.6 
 Well  18.9 225 4.6 6.2 86 
Site 300 
 Tank  17 1,021 1.5 6.8 −238 
 Filter  23 240 4.5 5.9 51 
 Overflow  17.7 369 1.4 7.1 −77 
 GW near SF 63 18.2 229 2.6 6.5 16 
 Background 85 17.7 174 2.6 6.2 31 
 Creek  12.1 165 7.1 6.9 108 
 Well  17.1 458 3.4 6.9 152 
Site 400 
 Tank  19 1,019 1.5 7.1 −134 
 Filter  15.3 5,165 5.5 7.2 595 
 Ditch  16.7 527 7.3 6.9 133 
 GW near SF 86 18.1 54 3.9 6.6 −36 
 Background 95 16.6 250 3.6 6.3 −8 
 Creek  19.9 87 7.2 −17 
 Well  18.1 91 2.3 7.3 175 
 DTW (cm)Temp (C°)SC (μS/cm)DO (mg/L)pHORP (mV)
Site 100 
 Tank  20.5 617 1.5 6.9 −325 
 Drainfield 409 17.5 138 3.5 5.9 34 
 15 m 180 17.7 84 5.4 45 
 35 m 50 20 122 −94 
 Lake 30 19.8 49 5.9 6.5 −49 
 Background  20.6 74 6.1 6.2 −20 
Site 200 
 Tank  20.5 765 1.4 6.5 −302 
 Drainfield 160 18.1 280 3.2 5.9 45 
 5 m 120 16.9 120 4.5 5.4 155 
 15 m 100 17.8 108 5.7 5.7 104 
 35 m 50 17.3 174 2.9 5.9 −42 
 Creek  17.5 169 6.2 6.6 
 Well  18.9 225 4.6 6.2 86 
Site 300 
 Tank  17 1,021 1.5 6.8 −238 
 Filter  23 240 4.5 5.9 51 
 Overflow  17.7 369 1.4 7.1 −77 
 GW near SF 63 18.2 229 2.6 6.5 16 
 Background 85 17.7 174 2.6 6.2 31 
 Creek  12.1 165 7.1 6.9 108 
 Well  17.1 458 3.4 6.9 152 
Site 400 
 Tank  19 1,019 1.5 7.1 −134 
 Filter  15.3 5,165 5.5 7.2 595 
 Ditch  16.7 527 7.3 6.9 133 
 GW near SF 86 18.1 54 3.9 6.6 −36 
 Background 95 16.6 250 3.6 6.3 −8 
 Creek  19.9 87 7.2 −17 
 Well  18.1 91 2.3 7.3 175 

Site 200 (conventional onsite wastewater system)

Groundwater PO4-P concentrations declined with increasing distance from the OWS, and all sampling locations had PO4-P concentrations that were significantly (p < 0.05) lower than septic effluent (7.78 mg/L) (Figure 3). For example, piezometers 200 and 201 s near the drainfield had a mean groundwater PO4-P concentration of 0.061 mg/L or 99% lower than septic effluent. Groundwater 15 m and 35 m down-gradient from the OWS had mean PO4-P concentrations of 0.029 mg/L and 0.025 mg/L, respectively (Figure 3). While PO4-P concentrations in groundwater near the drainfield were not significantly different than in groundwater 15 m down-gradient (p = 0.1590), statistically significant differences were observed 35 m down-gradient (p = 0.0129) relative to near the drainfield. Therefore the OWS was influencing the PO4-P concentrations in groundwater near the system (less than 35 m down-gradient). Piezometers 207 and 208 were outside the groundwater flow path of the OWS, and were considered the background sampling locations. These piezometers had a mean PO4-P concentration of 0.008 mg/L, and were significantly lower (p = 0.0084) than concentrations near the drainfield. The creek had a mean PO4-P concentration of 0.016 mg/L, but the concentrations were not significantly different than in groundwater 35 m down-gradient from the OWS (p = 0.5744).

The Cl/PO4-P ratios for all the groundwater sampling locations down-gradient from the OWS were elevated by more than one order of magnitude (range: 241 to 520) relative to septic effluent (6.5), indicating significant attenuation of the mass of PO4-P (Table 4). Based on the mixing model results, all groundwater sampling locations showed mass reductions of PO4-P greater than 92% relative to septic tank effluent. Adsorption of PO4-P was likely influenced by the iron oxide coatings on the clay minerals (Humphrey & O'Driscoll 2011) of the Georgeville (Clayey, Kaolonitic, thermic Typic Hapludults) series soils (Daniels et al. 1999).

Similar to site 200, the mean specific conductance of septic effluent (765 μS/cm) and groundwater near the drainfield (280 μS/cm) was elevated relative to other sampling locations (108 to 174 μS/cm), indicating that effluent was influencing the conductance of groundwater beneath the OWS (Table 5).

Groundwater PO4-P concentrations near the drainfield and down-gradient from the OWS at Site 200 were elevated in comparison to similar locations at Site 100. The differences in groundwater PO4-P concentrations between the two sites may be related to differences in mean wastewater strength (Site 100: 6.06 mg/L PO4-P; Site 200: 7.78 mg/L PO4-P), years of wastewater discharge to the subsurface (Site 100: 18 yrs; Site 200: 85 yrs), vadose zone thickness beneath the trenches (Site 100: 3 m; Site 200: 1 m), and clay content (Site 100: 65%; Site 200: 42%) (Karathanasis et al. 2006; Gill et al. 2009; Humphrey et al. 2015). The PO4-P treatment efficiencies of these systems did not change much season to season. Overall, both OWS were very efficient at reducing the concentrations (>99%) and mass (92–98%) of PO4-P. These findings are in agreement with prior research regarding the PO4-P concentration reductions in clayey soils (Karathanasis et al. 2006; Gill et al. 2009; Humphrey et al. 2015).

Groundwater near the OWS drainfields at Site 100 and Site 200 had similar mean (5.9 for both) and ORP (34 and 45 mV) (Table 5). These environmental conditions would be conducive to precipitation of the minerals variscite (AlPO4·2H2O) or vivianite (Fe3(PO4)2·8H2O), thus removing PO4-P from solution, and reducing the mass of PO4-P entering groundwater beneath the OWS (Zanini et al. 1998; Humphrey et al. 2014).

Site 300 (sand filter onsite wastewater system)

Site 300 was served by a surface discharge SF. The SF had been repaired a few months prior to initiation of our study and the septic contractor did not correctly connect a discharge pipe for filter effluent conveyance to the creek. As a result, the SF was functioning more like a conventional-style subsurface discharge OWS in the early months (March–June 2015) of the project. During the first two sampling events in March and May, effluent from the filter was upwelling through the soil to the ground surface. The surfacing effluent (overflow) was sampled during this period for PO4-P and environmental readings analyses. A discharge pipe for the filter was correctly connected in June and filter effluent was sampled from the pipe afterwards (July, September, and November).

The mean PO4-P concentration for septic tank effluent at Site 300 was 5.66 mg/L. Wastewater that was surfacing during the period prior to the installation of the filter discharge pipe had a mean PO4-P concentration of 0.736 mg/L, which was 87% lower than septic effluent (Figure 4). Groundwater adjacent to the SF at piezometer 301 had a mean PO4-P concentration identical to groundwater at piezometer 302 on the opposite side of the creek and away from the filter (both 0.010 mg/L) (Figure 4). Filter effluent had a mean PO4-P concentration of 0.108 mg/L, which is a 98% reduction in PO4-P concentration relative to septic tank effluent. The Cl/PO4-P ratios indicate there was mass removal of PO4-P via the SF as the Cl/PO4-P ratios were lowest for septic tank effluent (16.6), followed by the overflow (31.0), SF effluent (210.2), and groundwater near the SF (970.8). Using the two-component mixing model with Cl and PO4-P concentration data from piezometer background groundwater and septic effluent, it was determined that 89% of the filter effluent was groundwater and 11% of the effluent was wastewater. There was an estimated 83% mass reduction of PO4-P from the septic tank to filter outlet based on the model. The creek had a mean PO4-P concentration of 0.053 mg/L. Several other SF systems also discharged effluent into the same creek, contributing to the relatively high concentration.
Figure 4

Phosphate concentrations at Site 300 and Site 400 monitoring locations including water supply wells, surfacing effluent above the filter (300 Overflow), SF effluent (Filter), groundwater near the filter (301 and 401), groundwater away from the filter (302, 402, and 403) and the adjacent creeks.

Figure 4

Phosphate concentrations at Site 300 and Site 400 monitoring locations including water supply wells, surfacing effluent above the filter (300 Overflow), SF effluent (Filter), groundwater near the filter (301 and 401), groundwater away from the filter (302, 402, and 403) and the adjacent creeks.

The removal efficiency of the SF was likely influenced by the age of the media (<1 yr), and relatively low flows (two occupants). Prior studies have shown that as the adsorptive capacity of the media exhausts over time, the P treatment efficiency declines (Nielsen et al. 1993; Tonan et al. 2015), so PO4-P filter effluent concentrations may increase in future years. The mean pH (5.9) and ORP (51 mV) of filter effluent (Table 5) would be conducive to precipitation of the minerals variscite (AlPO4·2H2O) or vivianite (Fe3(PO4)2·8H2O) if aluminum and/or iron cations were present in the filter media (Zanini et al. 1998; Humphrey et al. 2014).

Septic effluent had the highest mean specific conductance (1,021 μS/cm), lowest mean ORP (−238 mV), and lowest mean DO (1.5 mg/L) of the sampling locations (Table 5). The overflow had physical and chemical properties most similar to septic effluent (specific conductance: 369 μS/cm; ORP: −77 mV; DO: 1.4 mg/L) (Table 5). The mean specific conductance of SF effluent (240 μS/cm) and groundwater near the SF (229 μS/cm) were also elevated relative to the creek (165 μS/cm) and background groundwater (174 μS/cm). This indicates that the OWS was influencing groundwater and surface water electrical properties.

Site 400 (sand filter onsite wastewater system)

Septic effluent at Site 400 had a mean PO4-P concentration of 5.683 mg/L and was elevated relative to all other sampling locations (Figure 4). Effluent from the SF had a mean PO4-P concentration (1.069 mg/L) that was 81% lower than septic tank effluent and the ditch receiving the SF discharge had a mean PO4-P concentration of 0.814 mg/L or 86% lower than the mean septic effluent concentration. SF effluent concentrations of PO4-P were highest during the wet periods of March and May, and were lowest during July and September. Therefore, the system was most efficient during warmer periods. Groundwater adjacent to the SF at piezometer 401 had a mean PO4 concentration of 0.250 mg/L (96% reduction) which was significantly elevated relative to groundwater away from the filter at piezometers 402 (0.033 mg/L; p = 0.0369) and 403 (0.004 mg/L; p = 0.0358). Because this OWS used chlorination for filter effluent disinfection, ratios of Cl to PO4-P and the mixing model could not be used. Specific conductance of the filter effluent was much higher than other sampling locations because of the chlorination process (Table 5). The OWS at this site was influencing groundwater and surface water PO4-P concentrations, and the specific conductance of surface water.

CONCLUSIONS

Data from Sites 100–200 (conventional OWS) indicated that PO4-P concentrations were reduced by an average of >99% in groundwater 35 m+ down-gradient from the OWS, relative to concentrations in septic tank effluent. Based on mixing model results, there was an average PO4-P load reduction of 92% to 98%, respectively, at Sites 200 and 100, and differences in PO4-P concentrations in groundwater 35 m down-gradient from the OWS, in background groundwater, and in adjacent surface waters were not statistically significant. Therefore, the conventional OWS in this study (with >40 m setback to surface waters) were not contributing significant loads of PO4-P to surface waters. In contrast, the mean PO4-P concentrations in SF effluent for Sites 300 and 400 were 0.108 to 1.069 mg/L, respectively. These concentrations were elevated relative to concentrations of PO4-P in the creeks that receive the SF discharge, and relative to groundwater 35 m down-gradient from the OWS at Sites 100 and 200. While the mean concentration reductions for the SF systems at Sites 300 and 400 were 80% and 84%, respectively, the SF effluent was still enriched with PO4-P, and the effluent was discharged directly to surface waters.

The conventional OWS were more efficient at reducing the concentration and mass of PO4-P in comparison to the SF. Water quality could be improved by further treating the SF effluent before discharge to surface waters by either extending the existing SF and incorporating reactive media such as wollastonite or slag material, or creating a constructed wetland built with reactive media to receive SF effluent (Westholm 2006).

While nitrogen concentrations in SF effluent are monitored for new systems, there is no requirement for PO4-P monitoring. Phosphorus has been identified as a pollutant of concern for Falls Lake, Jordan Lake, and High Rock in the Piedmont of North Carolina, where many SF are in operation (North Carolina Division of Water Resources 2015). Given the findings of this study with regards to PO4-P concentrations in SF effluent, it is suggested that PO4-P be included in the routine analyses of SF and other media filter effluent to provide a broader perspective on the contributions of these OWS to surface waters.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the efforts of Jonathan Harris, Jim Watson, John Woods, Christa Sanderford, and Colleen Rochelle with field and/or laboratory work. Funding was provided by the North Carolina Department of Health and Human Services.

REFERENCES

Arnscheidt
J.
Jordan
P.
Li
S.
McCormick
S.
McFaul
R.
McGrogan
H. J.
Neal
M.
Sims
J. T.
2007
Defining the sources of low-flow phosphorus transfers in complex catchments
.
Science of the Total Environment
382
,
1
13
.
Bradshaw
J. K.
Radcliffe
D. E.
2013
Nitrogen fate and transport in a conventional onsite wastewater treatment system installed in a clay soil: experimental results
.
Vadose Zone Journal
12
(
3
),
doi:10.2136/vzj2012.0149.
Conley
D. J.
Paerl
H. W.
Howarth
R. W.
Boesch
D. F.
Seitzinger
S. P.
Havens
K. E.
Lancelot
C.
Likens
G. E.
2009
Controlling eutrophication: nitrogen and phosphorus
.
Science
323
,
1014
1015
.
Daniels
R. B.
Buol
S. W.
Kleiss
H. J.
Ditzler
C. A.
1999
Soil Systems in North Carolina. Technical Bulletin 314
,
Department of Soil Science, North Carolina State University
,
Raleigh, NC, USA
.
Del Rosario
K. L.
Humphrey
C. P.
Mitra
S.
O'Driscoll
M.
2014
Nitrogen and carbon dynamics beneath on-site wastewater treatment systems in Pitt County, North Carolina
.
Journal of Water Science and Technology
69
(
3
),
663
671
.
Domenico
P. A.
Schwartz
W.
1998
Physical and Chemical Hydrogeology
. 2nd edn,
Wiley
,
New York
, p.
115
.
Florida Department of Environmental Protection
2013
Implementation of Florida's Numeric Nutrient Standards
.
Available from:
.
Gill
L. W.
O'Luanaigh
N.
Johnston
P. M.
Misstear
B. D. R.
O'Suilleabhain
C.
2009
Nutrient loading on subsoils from on-site wastewater effluent, comparing septic tank and secondary treatment systems
.
Water Research
43
,
2739
2749
.
Harman
J.
Robertson
W. D.
Cherry
J. A.
Zanini
L.
1996
Impacts on a sand aquifer from an old septic system: nitrate and phosphate
.
Groundwater
34
,
1105
1114
.
Holman
I. P.
Whelan
M. J.
Howden
N. J. K.
Bellamy
P. H.
Willby
N. J.
Rivas-Casado
M.
McConvey
P.
2008
Phosphorus in groundwater – an overlooked contributor to eutrophication
.
Hydrological Processes
22
,
5121
5127
.
Hoover
M. T.
2004
Soil Facts: Septic Systems and Their Maintenance. AG-439-13
,
North Carolina Cooperative Extension Service
,
NC, USA
.
Humphrey
C. P.
O'Driscoll
M. A.
2011
Biogeochemistry of groundwater beneath on-site wastewater systems in a coastal watershed
.
Universal Journal of Environmental Research and Technology
1
(
3
),
320
328
.
Humphrey
C. P.
O'Driscoll
M. A.
Deal
N.
Lindbo
D.
2014
Fate and transport of phosphate from an on-site wastewater system in Beaufort County, North Carolina
.
Journal of Environmental Health
76
(
6
),
28
34
.
Humphrey
C.
Anderson-Evans
E.
O'Driscoll
M.
Manda
A.
Iverson
G.
2015
Comparison of phosphorus concentrations in coastal plain watersheds served by onsite wastewater treatment systems and a municipal sewer treatment system
.
Water Air and Soil Pollution
226
(
1
),
doi: 10.1007/s11270-014-2259-4.
Karathanasis
A. D.
Mueller
T. G.
Boone
B.
Thompson
Y. L.
2006
Nutrient removal from septic effluents as affected by soil thickness and texture
.
Journal of Water and Health
4
(
2
),
177
195
.
National Oceanic and Atmospheric Administration
1996
NOAA's Estuarine Eutrophication Survey. Volume 1: South Atlantic Region
.
Office of Ocean Resources Conservation Assessment
,
Silver Spring, MD
,
USA
, p.
50
.
Nielsen
J.
Lynggaard-Jensen
A.
Hasling
A.
1993
Purification efficiency of Danish biological sand filter systems
.
Water Science and Technology
28
(
10
),
89
97
.
North Carolina Department of Health and Human Services
2015
Onsite Water Protection Branch: Annual Activity Reports
.
(accessed 10 December 2015)
.
North Carolina Division of Water Resources
2015
Nutrient Strategies
.
(accessed 10 December 2015)
.
North Carolina Department of Environmental Quality
2009
Lick Creek Watershed Restoration Plan
. (
accessed 4 January 2016
).
O'Driscoll
M. A.
Humphrey,
C. P.
Jr
Deal
N. E.
Lindbo
D. L.
Zarate-Bermudez
M. A.
2014
Meteorological influences on nitrogen dynamics of a coastal onsite wastewater treatment system
.
Journal of Environmental Quality
43
(
6
),
1873
1885
.
Osmond
D. L.
Gilliam
J. W.
Evans
R. O.
2002
Riparian Buffers and Controlled Drainage to Reduce Agricultural Nonpoint Source Pollution
.
North Carolina Agricultural Research Service Technical Bulletin 318
,
North Carolina State University
,
Raleigh, NC, USA
.
Postma
F. B.
Gold
A. J.
Loomis
G. W.
1992
Nutrient and microbial movement from seasonally-used septic systems
.
Journal of Environmental Health
55
(
2
),
5
10
.
Robertson
W. D.
Schiff
S. L.
Ptacek
C. J.
1998
Review of phosphate mobility and persistence in ten septic system plumes
.
Groundwater
36
,
1000
1010
.
United States Environmental Protection Agency
2002
Onsite Wastewater Treatment Systems Manual
.
EPA/625/R-00/008
,
Office of Water, Office of Research and Development
,
Washington, DC
,
USA
.
Zanini
L.
Robertson
W. D.
Ptacek
C. J.
Schiff
S. L.
1998
Phosphorus characterization in sediments impacted by septic effluent at four sites in central Canada
.
Journal of Contaminant Hydrology
33
,
405
429
.