Reflections on the ASAE National On-Site Wastewater Treatment Conference, 1998

Catherine Taylor, Don Jones, and Joseph Yahner

Note: This document was prepared for the Indiana State Department of Health. The objective was to discuss information from the 1998 ASAE conference, not to extensively review literature on the subjects discussed. Please go to source documents for a more information. This document discusses only a portion of the material presented at the conference. A copy of the conference proceedings can be obtained from:

American Society of Agricultural Engineers, 2950 Niles Road, St. Joseph, Michigan, 49085-9659. Phone: 616-429-9659, Fax: 616-429-3852, email:

Absorption Fields

The ability of on-site system absorption fields to receive typical septic tank effluent was evaluated over a 10 year period using a mass balance approach (Keys et al., 1998). Trenches were loaded at 4.1 cm/day (1.01 gpd/ft2) from 1987 to 1993 (at which point they were in or close to failure) and then at 1.6 cm/day (0.39 ft2) from 1993 to 1997. They found that at 4.1 cm/day, the system failed in 7 years, while at 1.6 cm/day, the system lasted 11 years (with extrapolation). They calculated that equilibrium would be achieved at minimum loading rate of 0.05 cm/day (0.01 gpd/ft2) and a maximum loading rate of (0.11 gpd/ft2). This study in sand found that system life was dependent upon biomat layer development. The author noted that the study should be replicated under different conditions to confirm observations.

This study raises questions about the validity of increasing the soil loading rate on more permeable soils, because the failure of trenches receiving septic tank effluent is dependent on the biological mass loading, not the soil permeability. It can be theorized from the study that if more permeable soils are going to be loaded at a higher rate, pretreatment should be used to decrease the mass loading to ensure system longevity. If a three-bedroom home (450 gpd) were loaded at 0.11 gpd/ft2, the system would require 4090 ft2 of absorptive area or approximately 14 trenches 100 by 3 feet. If trenches are 7.5 ft apart (rule 410 IAC 6-8.1), the absorptive area (without dispersal area) would require 0.32 acres. One potential difficulty with this study was that the lower loading rate was applied to trenches that had already received a higher loading rate. However, while this may change the numbers somewhat, the fundamental principle holds. This may provide an insight into the reported failures of several at-grade systems. The phenomena would be more readily visible on mound and at-grade systems because in the case of failure, less lateral overflow area is available prior to effluent surfacing. Soil permeabilities also tend to be less on soils requiring mound and at-grade systems decreasing the buffering capacity of the system. However, the mechanisms seem to be complex as other studies have shown that the dynamics of biomat formation vary with differing soils and distribution mechanisms (Hargett et al., 1981, Converse et al., 1974, and Anderson, 1981).

In Florida, a study was conducted to determine the average age of an on-site system at the time of failure (Sherman, et al., 1998). Two counties were studied in depth, in addition to a general statewide survey. They determined that the average age of an on-site system at the time of failure was 18.01, 18.35, and 18.53 years in Sarasota and Marion Counties, and statewide, respectively. The authors pointed out that this was the average age of failure, not the average lifespan, as only repair permits were analyzed.

The functioning of different types of trenches receiving either recirculating sand filter (RSF) or septic tank effluent (STE) was evaluated by Loudon et al. (1998). The trenches were as follows: standard gravel, half pipe, chamber, wrapped pipe, shallow, sand lined, and a trench that was smeared in at a wet time of year. In an effort to be "true to life", gravel with a noticeable amount of fines was used (ASAE On-Site Wastewater Symposium, 1998). The half pipe and a gravel trench received RSF effluent, while the rest received STE. The soil was a well drained loam with slow permeability. The Bt horizon averaged 42+6% sand, 30+5% silt, and 27+3% clay throughout the absorption field site. Trenches have been loaded since 1994. In 1997, average acceptance of STE was 1.68 cm/day (0.41 gpd/ft2) for the gravel, 1.23 cm/day (0.30 gpd/ft2) for the chamber, 2.1 cm/day (0.51 gpd/ft2) for the shallow gravity, and 1.54 cm/day (0.38 gpd/ft2) for the wrapped pipe. Similar trends have been reported by the Wisconsin Small Scale Waste Management Project when comparing gravel and chamber systems (noted March 19, 1998 on The trench installed under wet conditions and intentionally smeared accepted 1.26 cm/day (0.31 gpd/ft2) STE in 1997. Initial wastewater acceptance was lower than the other trenches receiving STE, however it improved to similar levels within approximately two years. The gravel trench that received RSF effluent, averaging 3.6 BOD5, 4.1 TSS, 21.0 NO3, and 0.97 NH4, accepted an average of 17.29 cm/day (4.25 gpd/ft2). (This effluent quality is becoming known as "standard" for recirculating sand filters (ASAE On-Site Wastewater Symposium, 1998)). The only statistical difference found was between the trenches receiving STE and RSF effluent. Based on these results and long term experience, Loudon recommended a three to five fold decrease in absorption field sizing following recirculating sand filter effluent (ASAE On-Site Wastewater Symposium, 1998). He stated that he felt this was a conservative recommendation, as in research trenches have been accepting approximately eight to ten times the amount recommended for septic tank effluent for long periods of time (>10 years) in a variety of difficult soils. This recommendation is consistent with those of Ball (1995), Bruen and Piluk, (1994), and Roy and Dupe (1994).

to top

Aerobic Pretreatment Systems

Converse et al. (1998) evaluated effluent from 20 single pass sand filters (ISF), 21 aerobic treatment units (ATU1) with filter fabric for solid retention and 10 aerobic treatment units (ATU2) without filter fabric preceded by a septic tank. Sand filter effluent had 4 and 6 mg/L BOD5, 18 and 25 mg/L TSS, 3.4 and 10 mg/L TKN, 2.3 and 9 mg/L NH3, 28 and 29 mg/L NO3, and 3 and 280 mg/L fecal coliform col/100 mL, median and average values respectively. ATU1 pump chamber effluent characteristics were 4.4 and 10.8 mg/L BOD5, 20 and 30 mg/L TSS, 3.3 and 8 mg/L TKN, 1.0 and 5.2 mg/L NH3, 24 and 28 mg/L NO3, and 530 and 10,000 fecal coliform col/100 mL, median and average respectively. Similar results were found for two systems (ATU1) evaluated for 9.4 and 7.2 years as BOD5 averaged 3.2 mg/L with a median of 2.8 mg/L and TSS averaged 21 mg/L and had a median of 14 mg/L. Systems had received maintenance every 6 months. The second type of aerobic treatment unit provided somewhat less treatment with parameter median and averages of 27 and 36 mg/L BOD5, 40 and 44 mg/L TSS, 40 and 39 mg/L TKN, 28 and 30 mg/L NH3, 0.9 and 5.2 mg/L NO3, and 2.4E4 and 1.5E5 fecal coliform col./100 mL, respectively. It was found that single pass sand filters provided the highest quality effluent, particularly with respect to fecal coliforms, followed by aerobic treatment units with filter fabric, and finally aerobic treatment units without filter fabric. Additionally, greater treatment variability was found in the ATU effluent than the sand filter effluent. Temperature had a greater affect on treatment in ATU systems than the sand filter systems. Decreased separation distances and absorption field downsizing was recommended for pretreated effluent with consistently low BOD5, TSS and fecal coliform counts.

Sievers (1998) also reported high quality effluent from a single pass sand filter in Missouri. Effluent averages were 3 mg/L BOD5, 3 mg/L TSS, 0.48 mg/L NH3, 27 mg/L NO3, and 72.8 fecal coliforms/100mL. The trenches, which were located in subsurface soil due to prior removal of the top 12 to 14 inches, were successfully loaded at 3.6 gal/ft2 (nine times that recommended for septic tank effluent). Pump operation costs were $0.70 per month. Sievers emphasized the advantage of dosing small amounts of effluent to all treatment components. Soil carbon analysis indicated that no biomat had formed, a result consistent with observations by Ball, Loudon, Converse, and others.

Both single pass and recirculating sand filters continue to show promise for on-site wastewater treatment. However, one barrier to their use has been the availability of appropriately sized sand media (0.3 mm - 1.0 mm with less than 4% fines). In Oregon, sand, recycled crushed glass, crushed limestone, polyethylene pellets, and foam (Waterloo Biofilter® media) were evaluated as potential alternative media (Weaver et al., 1998). When sand, glass, and limestone were loaded in a cylindrical downsized single pass filter at 1.87 gal/day, resulting parameters averaged BOD5 <3.5 mg/L, TSS <4 mg/L, TKN <4 mg/L, NH3 <4 mg/L, and NO3 <114 mg/L. Foam loaded at 9.82 gal/day resulted in averages of BOD5 6 and 7 mg/L, TSS 14 and 19 mg/L, TKN 6 and 5 mg/L, NH3 6 and 3 mg/L, and NO3 52 and 57 mg/L for hard and soft foam respectively. Plastic pellets loaded at 5.98 gal/day produced effluent averaging 15 mg/L BOD5, 3 mg/L TSS, 20 mg/L TKN, 16 mg/L NH3, and 38 mg/L NO3. The authors concluded that all of the media produced highly treated effluent in single pass sand filters with a media depth of 24 inches.

Constructed Wetlands

Thom et al. (1998) looked at the long-term treatment provided in residential constructed wetlands in Kentucky, a state with over 4,000 constructed wetlands in service, with the following results.


mg/L (% reduction)

Two cell system evaluated seven years Single cell system evaluated six years Single cell system evaluated four years Weighted average % removal
BOD5 48 (75%) 40.5 (79%) 60.2 (73%) 76%
TSS 488 21.1 (69%) 38.8 (75%) 70%

(single cell systems)

Fecal coliforms (cfu/100 mL) 6.1E5 (99%) 9.1E3 (99%) 3.3E5 (99%) >99%
TN 21.2 (54%) 23.5 (67%) 18.6 (63%) 60%
NH4 15.3 (62%) 19,6 (71%) 18.0 (53%) 65%
NO3 0.5 1.3 0  
TP 4.2 (47%) 3.1 (67%) 2.0 (73%) 57%

The two-cell system had wood shavings placed on the gravel bed in addition the plants. Wood shaving decomposition was thought to be responsible for the increase in TSS. Treatment in constructed wetlands was found to be variable. Reductions in wastewater parameters were significant, but in this study wastewater was not being treated to levels perceived necessary for significant absorption field downsizing.

Similar variability was found in three residential constructed wetlands that were monitored for two years in Southern Kansas (Mankin and Powell, 1998). Average reduction in wastewater parameters was 82-85% BOD5, 23-90% TSS, 71-92% fecal coliforms, 40-60% NH4, and 44-72% TP. On average, two of the three systems exceeded the 30 mg/L goal for BOD5 and TSS (the third system was around 30 mg/L TSS) and all systems exceeded the goal of 20-2000 fecal col./100 mL. Average performance did not describe the increase in fecal coliform counts at "numerous" sampling dates and periods when TSS and BOD5 removal was less than 50%. It was noted that septic effluent strength was unusually high, and the need for proper septic tank constructed was emphasized. Constructed wetland construction costs ranged from $2.50/ft to $3.00/ft. The authors stated that "… a need exists for modifications in design to allow more consistent, reliable treatment. These modifications are essential in order to place wetlands on a par with some other established alternative treatment systems, such as sand filters." In light of these findings, the authors recommended sizing the constructed wetlands to retain 100% of the system flow. Recommendations for a three bedroom home include: a septic tank effluent filter, a 500 ft2 x 1.5 ft deep lined cell with a 5-day retention time, then a 500 ft2 x 1.5 ft deep unlined sand filled cell with a 5-day retention time, and finally 2 ft deep open water cell with 1500-2000 ft2 of surface area.

White and Shirk (1998) researched the performance and design of constructed wetlands in two Southern Alabama communities. The two cell constructed wetlands designed to the TVA specifications produced first cell effluent with an average of 8,620 fecal coliform col./100 mL, 26 mg/L BOD5, 19 mg/L TSS, and 29 mg/L NH3. The authors noted that the TVA guidelines recommend a standard sized (14 ft. by 20.5 ft.) second unlined dispersal cell without regard to soil characteristics. In wet, silty or clayey soils, the second cell did not provide adequate disposal area and resulted in ponding in the cooler months. White and Shirk recommended sizing the dispersal area in accordance with 50-70% of the standard recommendations for septic tank effluent. Two days was found to be an adequate hydraulic retention time for BOD5, TSS, and fecal coliform removal, but inadequate for nitrogen removal. Length to width ratios of 3:1 appeared adequate. Average cost for the residential constructed wetland was $5,700. It may be of importance to consider that the increase in treatment efficiency for these wetlands may be linked to the higher temperatures in Southern Alabama.

to top

Peat Filters

Peat filters have been shown to produce effluent with low BOD5 and TSS levels, and moderate fecal coliform levels in a variety of environments. Peat filters have historically produced effluent ranging between 5-20 mg/L BOD5, 5-15 mg/L TSS, 5 mg/L NO3, <17 mg/L NH3, and a 99.9% reduction in fecal coliforms in the North Eastern United States. The filters offered the greatest treatment in the winter at 0-10oC. This was thought to be due to growth of fungi at these temperatures. The long term performance of peat filters was studied in Alabama (O'Driscoll et al., 1998). In an Alabama study in 1994, 20 peat filters (PurafloTM) installed in September 1993 averaged 18 mg/L BOD5, 1.2 mg/L NH3, 25 mg/L NO3, and 57,665 fecal coliform col/100 mL with treatment improving at the end of the 12-month monitoring period. In 1997 a follow up study began concentrating on four of the original systems. Wastewater parameters averaged 4 mg/L BOD5, 13 mg/L TSS, 11 mg/L NH3, 7 mg/L NO3 and 23,769 fecal col/100 mL. It was noted that none of the homeowner had renewed their two-year service contract that came with the filter and that effluent filters were recommended for adequate treatment. The homeowners had performed little maintenance. Regular maintenance includes periodic surface raking to break up any biomat development in addition to peat replacement every eight years (Talbot et al., 1998). In Quebec, a peat filter company (Ecoflo®) includes a service contract in the price of the product for the eight years. All distributors are trained in installation and maintenance practices by the company and required to attend refresher training sessions every year. Ecoflo® periodically audits installations. This particular peat filter is estimated to cost between 25-50% more than conventional systems, a number that includes maintenance.

Septic Tanks

Septic tanks are an essential component to almost all types of on-site wastewater treatment systems. Yet, researchers and regulatory agencies around the country are finding that often tanks are not water tight and/or suitably constructed to allow use with conventional or alternative technologies. North Carolina developed recommendation guidelines to ensure adequate septic tank construction (D'Amato and Devkota, 1998). For concrete tanks, they recommend the use of concrete with a 28-day compressive strength of 3,500 psi (ideally 4,000 psi). Concrete with a sump of 10 to 15 cm should be used. They found that often the concrete is mixed appropriately, but then water is added to increase the flexibility of the concrete. This decreases its strength and reduces its ability to hold water. If it is necessary to increase plasticity, admixtures (plasticizers) should be added instead of water. Consolidate the concrete using a mechanical vibrator. Water to concrete ratios of 0.5 or less and moist curing for 7 days has been found to produce watertight tanks. Tanks should harden in a warm moist environment 24 - 48 hours before removal from its mold. Tanks that harden too fast will develop cracks and decreased strength. D'Amato and Devkota recommended covering the tank with plastic or periodically watering the tanks down in hot dry environments. North Carolina requires that tanks are reinforced with 6 inch by 6 inch, 10/10 fine gauge wire fabric. One-inch mastic, butyl rubber seals the seams. A detailed inspection, testing, and troubleshooting guide is also presented.


The most recent literature on on-site systems resembles previous literature (Taylor et al., 1997). Initial observations of pretreatment technologies were reinforced in the 1998 ASAE symposium with greater experience in more environments. General observations are as follows:

  1. Single pass and recirculating sand filters show promise for providing consistently high quality effluent.
  2. Sand filter effluent does not form a biomat layer and therefore may allow downsizing of three to five times the sizing for conventional septic tank effluent. This may not be possible in soils with a high water table or with saturated hydraulic conductivities less than what would be required. Information on saturated hydraulic conductivities in Indiana can be found in the earlier report (Taylor et. al, 1997).
  3. The primary advantage to recirculating sand filters is their ability to remove nitrogen. The recirculation usually results in 60-70% removal.
  4. Crushed glass, limestone, polyethylene pellets, and foam have been shown as adequate replacements for sand in sand filters in areas where sand meeting the necessary size requirement is difficult to obtain. Media sizing is extremely important to proper functioning.
  5. Constructed wetlands can produce high quality effluent, but treatment levels are often variable. This is expected, as constructed wetlands are a natural system with greater dependence on the environment. In light of this, downsizing recommendations generally range between 30% and 50%. It is important to note that many of the studies reported results from systems with approximately a two-day retention time. A five-day retention time may produce better results.
  6. Nitrogen removal has not been effective in constructed wetlands with less than a six or seven day retention time. A seven to ten day retention time may produce adequate nitrogen removal, however this is generally cost prohibitive. Potentially, placing a small trickling filter immediately proceeding the wetland would produce greater nitrogen removal with less expense.
  7. Aerobic treatment units can be a viable alternative for sites limited in space. Units that incorporate a filter fabric component have been shown to provide greater treatment than those that do not. ATU systems are the most sensitive to changes in flows and require the most frequent maintenance of the systems studied. They can be used effectively if the homeowner and company are committed to regular, long-term maintenance.
  8. Peat filters may be a viable technology for Indiana if limitations are recognized. Their low BOD5 and TSS may not form a biomat, which may allow use on slowly permeable soils and other difficult sites. Further research is required to test this hypothesis. Peat filters would not be a good option for sites with a seasonally high water table due to risk of ground water contamination with fecal coliforms. Like all pretreatment systems, regular maintenance is required. Peat filters are unique in that the peat has a limited lifespan. Peat must be replaced every eight to ten years to assure proper treatment.
  9. Researchers evaluating sand filters, constructed wetlands, and peat filters have discussed the importance of dosing the pretreatment unit with small doses of effluent.
  10. Septic tanks are a key component to proper functioning of most pretreatment technologies. Proper design and testing are critical. Additionally, effluent filters are being used and recommended in most projects as a simple, cost effective way to improve treatment.
  11. Evidence has indicated that variations in trench type (gravel, chamber, half pipe, wrapped pipe and sand lined) have provided little difference (not statistically significant) in infiltration and failure rates when no pretreatment technologies are used.

to top


Anderson, J.L., R.E. Machmeier, and M.J. Hansel, 1981. Long-Term Acceptance Rates of Soils for Wastewater. On-Site Sewage Treatment, Proceeding of the Third National Symposium on Individual and Small Community Sewage Treatment. ASAE.

Ball, H.L., 1995. Sand Filters and Shallow Drainfields. Orenco Systems, Inc., Sutherlin, Oregon.

Bruen, M.G. and R.J. Piluk. 1994. Performance and Costs of On-Site Recirculating Sand Filters. On-Site Wastewater Treatment. Volume 7. Proceedings of the Seventh International Symposium on Individual and Small Community Sewage Systems. ASAE. Michigan.

Converse, J.C., J.L. Anderson, W.A. Ziebell, and J Bouma, 1974. Pressure Distribution to Improve Soil Absorption Systems. Home Sewage Disposal. Proceedings of the National Home Sewage Disposal Symposium. ASAE.

Converse, J.C., M.M. Converse, 1998. Pump Chamber Effluent Quality Following Aerobic Units and Sand Filters Serving Residences. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

D'Amato, V.A., I.C. Devkota, 1998. Development of Prefabricated Septic and Pump Tank Construction and Installation Standards for North Carolina. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Hargett, D.L., E.J. Tyler, and R. L. Siegrist, 1981. Soil Infiltration Capacity as Affected by Septic Tank Effluent Application Strategies. On-Site Sewage Treatment, Proceeding of the Third National Symposium on Individual and Small Community Sewage Treatment. ASAE.

Keys, J.R., E.J. Tyler, and J.C. Converse, 1998. Predicting Life for Wastewater Absorption Systems. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Loudon, T.L, G.S. Salthouse, and D.L. Mokma, 1998. Wastewater Quality and Trench System Design Effects on Soil Acceptance Rates. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Mankin, K.R., and G. M. Powell, 1998. Onsite Rock-Plant Filter Monitoring and Evaluation in Kansas. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

O'Driscoll, J-P., K.D. White, D.W. Salter, and L. Garner, 1998. Long Term Performance of Peat Biofilters for Onsite Wastewater Treatment. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Roy, C. and J.P. Dupe. 1994. A Recirculating Gravel Filter for Cold Climates. On-Site Wastewater Treatment. Volume 7. Proceedings of the Seventh International Symposium on Individual and Small Community Sewage Systems. ASAE. Michigan.

Sherman, K.M., R.W. Varnadore, and R.W. Forbes, 1998. Examining Failures of Onsite Sewage Treatment Systems in Florida. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Sievers, D.M., 1998. Pressurized Intermittent Sand filter with Shallow Disposal Field for a Single Residence in Boone County, Missouri. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Talbot, P., H. Ouellet, and G. Laliberte, 1998. Development of a New On-Site Wastewater Treatment Technology in the Evolving Context of the Last Decade. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Taylor, C.H., J.E. Yahner, and D.D. Jones, 1997. An Evaluation of On-Site Technology in Indiana: A Report to the Indiana State Department of Health. Purdue University.

Thom, W.O., Y.T. Wang, and J.S. Dinger, 1998. Long-Term Results of Residential Constructed Wetlands. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

Weaver, C.P., B.S. Gaddy, and H.L. Ball, 1998. Effects of Media Variations on Intermittent Sand Filter Performance. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

White, K.D., and C.M. Shirk, 1998. Performance and Design Recommendations for On-Site Wastewater Treatment Using Constructed Wetlands. In On-Site Wastewater Treatment, Proceedings of the Eighth National Symposium on Individual and Small Community Sewage Systems. ASAE.

to top

Back to Purdue On-Site Articles