Approximately 50% of North Carolininans do not have access to public sewerage system. Instead, these people utilize on-site septic systems for treatmetn and disposal of their household wastewaters. One problem, however, is that conventional septic systems with gravity distrbution using 60 to 90 cm (24 to 36 in) deep trenches will function properly only on a limited number of soils in the state. Future growth in rural and suburban North Carolina will be contingent upon proper use of alternative septic systems since federal grants are no longer available for building public sewerage treatment plants.

In 1987, the North Carolina General Assembly funded a four year study to assess the usage of alternative on-site technologies throughout the state, evaluate the effectiveness of teh most commonly used alternative systems and further improve the potential usefulness of alternative on-site technologies.

The statewide distribution of alternative systems was assessed through a questionaire survey. The effectiveness of the most commonly used alternative systems was analyzed by selecting a county (from the questionaire results) that utilized thousands of alternative systems and conducting an intensive field investigation of a random, stratified sample of 340 systems in the spring of 1988. Hoover and Ammozegar (1989) reproted that the alternative systems studied (low pressure pipe and sand mound systems) were not functioning as intended, in large part because of a lack of homeowner maintenance. As a result of that study, state rules concerning the siting, design, installation, and maintenance of alternative systems have been improved and will require certified operators to perform regular maintenance of alternative systems beginning in 1992.

The third task was investigated in the Piedmont physiographic province in the central part of the state. This paper will focus upon the part of that research effort conducted in th eTriassic Basin Soil System. Triassic basins occur in North Carolina as elongated basins with their long axes trending northeast to southwest. These basins contain poorly consolidated sedimentary rocks including shales, siltstones, snadstones, and conglomerates (Wilson et al. 1980). soils formed in these sedimentary rocks typically have clayey, slowly permeable subsoils with moderate to very high shrink-swell potentials (Table 1). Most of these soils are not suited for conventional septic systems due to their slow permeability, high shrink-swell potential, perched water tables, and shallow depth to Cr (paralithic contact or rock-like material) or R horizons (hard rock).

Each of the soil series in Table 1 encompasses a range of characteristics and sutabilities for on-site wastewater treatment and disposal. Pinkston soils, for example, are not typically considered suitable for on-site subsurface disposal systems because of moderately deep bedrock. However, these soils are occasionally used. The Mayodan soils also are occasionally sutiable while the Granville soils are usually suitable for on- site systems. The White Store soils have a thin sandy surface overlying the B-horizon and are, at best, suited only to single- family spray irrigation systems. The Creedmoor soils have a thicker sandy surface, and a slightly deeper perched water table than White Store soils. Also, the upper B horizon in most Creedmoor soils is less clayey and more permeable than in White Store soils.

Since the Creedmoor soils are not considered suitable for conventional systems, there is a significant need to assess the potential effectiveness of alternative subsurface treatment and disposal systems in these soils. The objective of the study reported here was to evaluate the effectiveness of sand filter, low pressure pipe systems utilized on these clayey, mixed, thermic Aquic Hapludults and on other soils on similar landscapes in the Triassic Basin Soil System. Specific subobjectives included (1) evaluating the operation and funcitoning of system components (e.g., tanks, pumps, floats, etc.), (2) assessing the treatment effectiveness of buried, pressure-dosed sand filters and (3) determining the hydraulic capacity and wastewater treatment potential of these soils.

Study Sites

Sand filter, low pressure pipe systems were studied at two sites located in Chatham County, NC, in the Deep River Triassic Basin that extends from Oxford, NC, to South Carolina (Fig. 1). Two study sites were within 25 km (15 mi.) of each other.

Site Characterization: Soil morphology was described form auger borings throughout each site and from a backhoe-dug pit near the septic system at eact site. Standard USDA morphological criteria (Soil Conservation Service 1981) were used.

Saturated hydraulic conductivity (Ks) was measured in-situ and from undisturbed soil cores obtained form major soil horizons at Site No. 1 (Weymann 1989). The core samples were prepared for analysis as per Amoozegar (1988) and Ks measured using the constant head technique. Samples with very low conductivities were measured using the falling head technique (Klute and Dirksen 1986). In-situ Ks measurements were made using the compact constant head permeameter and the Glover solution (Amoozegar 1989a, 1989b).

System Design at Site No. 1: This system included a 3800 L (1000 gal) septic tank (referred to as 1st), a Tyson low splitter (The use of trade names in this publication does not imply endorsement by the North Carolina Agricultural Research Service or North Carolina Cooperative Extension Service of the products named, nor criticism of similar ones not mentioned.), two 3800 L (1000 gal.) dosing tanks, a pressure-dosed, buried sand filter, and two similar, side-by-side low pressure pipe (LPP) drainfields. One drainfield received septic tank effluent and the other received sand filter effluent. The system was designed for a 3 bedroom house and became operational in August, 1988. Two people lived in the home during the study period.

Septic tank effluent was partitioned so that approximately one- half the wastewater flowed into Pump Tank No. 1 (1PT1) that dosed to the sand filer and the other half flowed into Pump Tank No. 2 (1PT2) that dosed to one LPP subfield. The sand filter was designed to receive only one-half the daily flow at a design loading rate of 45 Ld^-1m^-2 (1.1 gpd/ft²). It was 2.7 by 5.4 m (9 by 18 ft) with a 0.76 m (2.5 ft) thick layer of coarse snd. It was pressure dosed using 3.2 cm (1.25 in) diameter PVC laterals on 0.9 m (3 ft) centers with 40 mm (5/32 in) holes. the entire sand filter was placed in a plastic-lined (6-mil) excavation, covered with topsoil, and vegetated. Effluent that passed through the sand filter drained to the Sand Filter Pump Tank (1SFPT) from which it was dosed to the second LPP subfield.

These LPP drainfields were designed according to Cogger et al. (1982) except for trench width, hole size, and earthen dam placement. Trenches were 25 cm (10 in) wide. Laterals were 3.2 cm (1.25 in) diameter PVC with 40 and 36 mm (5/32 and 9/64 in) holes. Earthen dams were only used at the manifold end of lateral trenches. The design loading rate was 5.3 Ld ^-1m^-2 (0.13 gpd/ft²) on a drainfield area basis (as per the calculation method of Cogger et al. 1982). Each drainfield contained eight laterals on 1.5 m (5 ft) centers.

System Design at Site No. 2: This system included a 3800 L (1000 gal.) septic tank, a 3800 L (1000 gal) dosing tank, a pressure-dosed, buried sand filter, and four small LPP subfields (called Subfields A, B, C and D). the system was designed for a 2 bedroom house and becaome operational in October, 1988. Two people lived in the house during the study period.

All of the septic tank effluent was dosed from the Pump Tank (2PT) to the sand filter at a design loading rate of 45 Ld ^-1m^-2 (1.1 gpd/ft²). The 2.7 by 7.3 m (9 by 24 ft) filter was similar to that at Site No. 1. Effluent that passed through the sand filter drained to the Sand Filter Pump Tank (2SFPT) and was dosed to two of the four subfields. A pressure-activated K-rain¹ valve alternately directed the sand filter effluent to LPP Subfields A and C during one dosing event, then to LPP Subfields B and D the next dosing event. Construction details for these subfields were similar to Site No. 1 except the design loading rates were 3.3, 3.7, 2.9, and 4.5 Ld^-1m^-2 (0.08, 0.069, 0.07 and 0.11 gpd/ft²) on a drainfield area basis for Subfields A, B, C, and D, respectively.

System Monitoring

The performance of the pressure distribution networks including the pumps, electrical controls, pipe networks, and pump tanks were monitored continuously. This paper will summarize data collected through December, 1990. Impulse counters, hard wired into pump control systems, and Stevens¹ (TypeF, Model 68) continuous water level recorders on each pump0 tank allowed evaluation of frequency of pump cycling and the volume of dosing events. From these data the extent of water infiltration into tanks and wastewater exfiltration out of tanks were evaluated. Data form recording rain gauges were used to evaluate the relationships between rainfall events, infiltration, and wastewater treatment efficiency.

All pressure distribution systems had turnups that came above the ground surface at the distal ends of the laterals. Clear PVC pipes were periodically attached to these turnups, the pump activated, and the operating pressure head in the pipe network determined.

The wastewater delivery rate was calculated as the L min^{- 1} (gpm) rate that the pump delivered wastewater to the drainfield when the pressure head was set to the design specifications. The pump run time was the total time required for the pump to deliver one dosing volume to the drainfield. These data and water level recorder charts were used to assess the extent of clogging of the perforations in the laterals and clogging of gate valves used to regulate pressure.

The ability of the soil to absorb the effluent was evaluated by measuring weekly the level of wastewater ponding in 17 slotted observation wells in each of the two LPP drainfields at Site No. 1 and from 11, 12, 7, and 5 observations wells in the trenches of Subfields A, B, C, and D, respectively, at Site No. 2. Sampling/monitoring wells were installed 0.3, 0.6, 0.9, 1.5 and 3 m (1, 2, 3, 5, and 10 ft) directly downslope of (1) the LPP system receiving sand filter effluent, (2) the LPP system receiving septic tank effluent, and (3) a part of the landscape receiving both at Site No. 1. These wells were installed 1.15 m (45 in) below the ground surface (into the Cr horizon) and were slotted throughout the lower E, Bt, and Cr horizons.

During a 13 month study period in 1988-89 after the system at Site No. 1 became operational, Weymann (1989) measured soil water content and water potential within and around the drainfields using a Troxler¹ neutron probe and tensiometry, respectively.

Wastewater/Water Sampling and Analyses

The treatemtn efficiency of teh sand filter units at both sites and of the soil treatment system at Site No. 1 were determined by analyzing wastewater/water samples collected once to four times monthly. Sample collection began in April, 1989 after both sand filters had matured. Grab samples were collected from tanks and wells with a non-contaminating hand vacuum bilge pump or a peristalic pump and transported on ice to the laboratory as per U.S. EPA (1982). Sampling wells were purged the day prior to each sampling event. This paper summarizes water quality data through 1990.

Wastewater samples were analyzed for pH, suspended solids, total Kjeldahl nitrogen (TKN-N), ammonium-nitrogen (NH₄-N), nitrate-nitrogen (NO₃-N), total phosphorus, orthophosphate, chlorides, and periodically for chemical oxygen demand, biological oxygen demand, and fecal coliform densities. Well water samples were analyzed for pH, suspended solids, ammonium-nitrogen, nitrate-nitrogen, orthophosphates, chlorides, and periodically for fecal coliform densities. Organic-nitrogne (Organic-N) was calculated as the difference between TKN-N and NH₄-N concentrations. Total nitrogen (TN) was calculated as the sum of NO₃-N and TKN-N concentrations.

Chemcial analyses were form standard techniquesdescribed by APHA (1985) or long-standing techniques utilized in our laboratory. Suspended solids and pH were measured using standard methods (APHA 1985). Total phosphorus was determined using persulfate digestion without filtration followed by colorimetric determination using the ascorbic acid method (APHA 1985). Orthophosphate was determined using the same colorimetric technique on undigested samples that had particulate matter removed by centrifugation. Chemical oxygen demand was determined by the closed reflux colorimetric method using commercially- purcheased COD ampules (APHA 1985). Total Kjeldahl nitrogen was measured using a semi-micro Kjeldahl method described by APHA (1985) and substituting a mixed catalyst (zirconium dioxide and cupric sulfate) for mercury in the digestion step (Glowa 1974). Nitrate-nitrogen (including nitrite-nitrogen) was determined colorimetrically using a salicylic acid technique (Cataldo et al. 1975). Ammonium-nitrogen was determined colorimetrically using a phenol-hypochlorite technique (Cataldo et al. 1974; Smith 1980). Chlorides were determined utilizing a conductimetric titration procedure developed in our laboratory (Crabtree 1989).

The treatment effectiveness of the sand filter units was evaluated by comparing the concentrations of constituents entering and leaving the filters. Also, mass loadings were claculated by summing, over time, the measured concentration times the flow volume for each part of the system. Flow was assumed to occur at the constituent concentration measured during the nearest wastewater sampling date.

The membrane filter technique was used to determine the density of fecal coliform colonies per 100 ml of sample (APHA1985). These analyses were conducted immediately following sample collection and transportation to the laboratory (U.S. EPA 1982). Each sample was analysed using three replication at each of three dilutions. The average for the most appropriate dilution was used for analyses. Since fecal coliform density data typically have a log-normal distribution, fecal coliform averages that will be presented are the anti-log of the average of the log transformed data (APHA 1985).

Operation of System Components

In general the mechanical and electrical system comp9onents performed quite well. The pumps, floats, and electrical controls did not require maintenance during the study period. Also, the pump impulse counters accurately recorded pump dosing frequency. There were, however, periodic operational problems due to leaking tanks, partial clogging of some distribution systems and broken pipes.

Tanks: There was a 36% higher flow and a 30% lower Cl concentration in the sand filter pump tank (1SFPT) than in 1PT1 at Site No. 1 (Table 2). However, the total daily Cl content passing through the tanks was equivalent. These data indicated that water with a low Cl concentration leaked into the sand filter and/or sand filter pump tank at Site No. 1. Also, fluctuations of the Cl concetration ratio 1SFPT/1PT1 throughout the study period indicated that infiltrating water frequently diluted the Cl concentration in 1SFPT (Fig. 2). while infiltration of rainwater falling directly on the sand filter may have been significant, it could not account for the increased flow in 1SFPT. the most intensive rainfall event in the study period (9 to 10 cm (3.5 to 4 in) rainfall) in Week 43, 1990 resulted in 8 dosing events over a 24 hour time period and 4 additional doses the next 24 hours at 1SFPT. In contrast, a 5- day rainy period in Weeks 49-50, 1989 included many small showers totalling less than 7.5 cm (3 in) of precipitation that caused numerous dosing events at 1SFPT during the seven days following the 5-day rainy period. During this time there were increasingly long intervals between dosing events until the dosing frequency approached a 0.5 to 1.0 doses per day frequency due to wastewater production in the home at the time. Therefore, other waters, such as shallow flowing perched waters, probably accounted for a substantial amount of the infiltration into pump tanks and/or the sand filter.

Table 2. Average Daily Flow Through Tanks and AVerage Chloride Concentration.

Please address any questions to Dr. David Lindbo.

This page (http://www.ces.ncsu.edu/plymouth/septic/95piluk.html) created by
Vera MacConnell, Research Technician, I on April 16, 1998.
Last Updated on June 3, 1999.