Environmental impact of irrigating lawns with treated domestic wastewater.

Three lawn plots equipped for measuring and sampling runoff were used to observe the environmental impact resulting from daily irrigation with treated septic tank effluent. "Pop-up" sprinklers were used to apply septic tank effluent treated by an on-site recirculating sand filter to one plot and potable water to another. The third plot received no irrigation. All lawn-plot runoff during a nine-month period was measured, sampled and analyzed for quality. Coliform bacterial exposure risk was assessed for each plot on a quarterly basis.Greater quantities of each monitored pollutant (i.e., nitrogen, COD and coliform bacteria) moved off-site in runoff from the potable-water-irrigated plot than from either of the other two plots. Fecal and total coliform bacteria were found to be present on all lawn plots. The risk in summer of encountering a coliform bacterium as a result of contacting grass in the plot irrigated with treated wastewater appeared to be no greater than that associated with contacting grass in a conventionally managed lawn.Observation of single family residential units in the United States suggests that there are at least two specific requirements on almost everyone's list of essentials for an acceptable residence. One requirement is a means for discharging domestic wastewater and the other is an outside area containing living plants (i.e., a lawn and/or ornamental garden).In areas not served by sewer systems, soil plays an important role in satisfying both requirements. On-site domestic wastewater renovation systems have traditionally relied on the physical, chemical and biological properties of soil to remove waste components as wastewater passes through on its way to a groundwater reservoir. Plants growing in lawns and gardens also rely on soil for physical support as well as a reservoir from which water and essential nutrients can be drawn.Unfortunately, many sites where people desire and/or need to live have thin soils that require special efforts to meet the essential requirements of a lawn and facilities for waste water discharge. Effective waste water treatment systems that do not rely upon natural soil for cleaning wastewater, such as recirculating sand filters (1), can be implemented on such sites. However, they do not totally satisfy the requirement of a facility for wastewater discharge, since a means of discharging effluent from such systems must be provided. The requirement for a lawn on such sites can become a benefit at this point, because, on thin soils, water is often the factor limiting the establishment and maintenance of lawn plants. Discharge of treated wastewater through a lawn irrigation system, therefore, provides the means for satisfying two essential requirements that often limit use of sites with thin soils for residential development.However, concerns about potential negative impacts on the environment always accompany suggestions of lawn irrigation with treated wastewater. Concerns about potential health risks associated with human lawn contact and transport of contaminated runoff onto adjoining property and into streams serve to prevent widespread implementation of lawn irrigation with treated wastewater.Some of the concerns can be addressed by requiring subsurface irrigation. Treated wastewater subsurface irrigation systems are being implemented to some extent in some western states (2).The state of North Carolina permits the use of surface sprinklers for on-site irrigation of treated domestic wastewater and has issued rules and regulations governing such systems (3). Because of concerns about negative environmental impacts, buffer zones of 122 meters (400 feet) between inhabited residences and irrigation systems are required.A better understanding of risks coupled with efforts to develop design and operating criteria that will minimize risk may help identify a technology that can be readily accepted by society. The objective of the project described herein was to observe the impact of lawn runoff quality and bacterial exposure risk produced by daily lawn irrigation with treated domestic wastewater.Procedure summaryThree lawn plots were delineated and equipped for measuring and sampling runoff. Pop-up irrigation sprinklers set to discharge in a semi-circular pattern applied water daily to two of the three plots. The third plot was not irrigated. Irrigation water for one of the irrigated plots was produced by a prototype on-site domestic wastewater treatment system which used a partially saturated recirculating sand filter to treat septic tank effluent. The other irrigated plot was watered with potable water from the local water utility.All runoff from each lawn plot during the nine-month study period (February to November) was measured and sampled. Samples were analyzed for chemical oxygen demand (COD), nitrogen, total coliform bacteria, and fecal coliform bacteria. Risk of human exposure to coliform bacteria from grass in each of the lawn plots was also evaluated once during spring, summer and autumn (i.e., a total of three assessments were made).Lawn plotsThree lawn plots, located 26 meters (85 feet) on center, were established in an area across a hedgerow from the established lawn surrounding a single family residence on the plant science unit of the Tennessee Agricultural Experiment Station's Knoxville Branch Station. All three TABULAR DATA OMITTED plots were within a common soil mapping unit and occupied almost identical landscape positions. The area in which the plots were established had a slope of approximately 10 percent and was populated with a mixed culture of vegetation (dominated by tall rescue). The area had been regularly mowed and not used for either plot work or agricultural production in a number of years. Established vegetation within each plot was not disturbed except as necessary for installation of irrigation sprinklers.The unfenced perimeter of each plot was defined by earthen embankments arranged in the pattern of an isosceles triangle as shown in Figure 1. The embankments were a low ridge of soil formed by a moldboard plow. The embankments served the dual purpose of diverting runoff from surrounding areas and directing runoff from within a plot to an H-type flume installed at the down-slope apex of the triangular perimeter. Each flume was equipped with a continuous chart-type stage or depth-of flow recorder.All discharge from each flume dropped into a two liter (0.52 gallon) composite sample-collection container. In the few instances when the sample container became filled during a storm event, runoff continued to enter the container and produce an overflow of equal volume. Overflow from the sample container was discharged onto the soil outside the lawn plots. Both the flumes and collection containers were covered in order to prevent precipitation from diluting runoff samples.IrrigationApparatus: Two of the three lawn plots (i.e., the test plot and the irrigated control plot) were each equipped with a single pop-up sprinkler installed 1.5 meters (5 feet) off the center of the base of the triangular perimeter embankment. Each sprinkler was a Toro Series 300-12-02 shrub hi-pop with a rated discharge of 5.3 liters/ minute (1.4 gpm) and a spray radius of 6.7 meters (22 feet) at an operating pressure of 206.8 kPa (30 psi). Each sprinkler was installed so that it was 20 cm (8 inches) below ground level when not in operation and 10 cm (4 inches) above ground during operation. A section of 10-cm (4-inch) diameter PVC pipe surrounded each sprinkler and provided a passage through which the sprinkler traveled as it "popped up" and retracted. A circular lid hinged to the wall of the PVC pipe closed on the top of the pipe as the sprinkler retracted and helped retain heat from the soil. Retained heat helped prevent freezing of water droplets in the small sprinkler discharge ports.Water to the sprinkler in the test plot was supplied by a pump installed in the collection sump of a prototype on-site wastewater treatment system. The pump was started at 4:00 a.m. each day by an electro-mechanical timer. The pump continued to operate and deliver water to the sprinkler until all accumulated treated wastewater had been discharged.The same control system that started and stopped the wastewater irrigation pump also controlled a valve which allowed potable water from the local water utility system to flow to the sprinkler on the irrigated control plot. Because of a common control system, flow to each irrigation sprinkler persisted for exactly the same length of time. A manual valve in the supply line to the potable water sprinkler was initially adjusted to produce a flow rate equal to that in the line to the wastewater sprinkler. A flow meter in the line to each sprinkler verified equal application rates on each irrigated plot throughout the project.Water: Water applied to the test plot was septic tank effluent that had been processed by a prototype on-site domestic wastewater treatment system. The active portion of the prototype system was a partially saturated recirculating sand filter. The partially saturated filter provided an aerobic zone for nitrification followed by an anaerobic zone for denitrification. With such an arrangement, approximately 80 percent of the nitrogen in effluent from a septic tank serving a single family residence was removed prior to its irrigation onto the test plot. Approximately 90 percent of the COD and more than 95 percent of the coliform bacteria were also removed from septic tank effluent by the prototype system. Quality of the water applied to the test plot is summarized in Table 1.Water applied to the irrigated control plot was supplied by the local municipal water system which provides drinking quality water to a community of more than 300,000 people. Quality of the water entering the distribution system is carefully monitored and controlled to meet established drinking water standards.The quantity of water applied to each irrigated lawn plot was controlled by the amount of wastewater processed by the prototype treatment system. The mean and standard deviation for the application rate was 309 liters/day (82 gpd) and 99 liters/ day (26 gpd), respectively.Sample collectionRunoff: After each precipitation event which produced runoff, contents of the composite-sample-collection container from each lawn plot were transferred to a sterile sample bottle and transported to the TABULAR DATA OMITTED laboratory for analysis. The volume of each runoff event was calculated from the stage recorder chart. There was not runoff associated with daily irrigation events. All runoff was produced by rainstorms.Bacterial exposure: Since people periodically come in contact with lawn surfaces, an assessment of potential health risk associated with lawn irrigation of treated wastewater was of interest. Coliform bacteria are commonly used as an indicator of the presence of pathogens and can, therefore, be used to help assess health risk. Two types of assessment for exposure to both fecal and total coliform bacteria were performed on April 27, August 2 and November 9, for each of the plots. One type of assessment was for exposure to coliform bacteria by contact with the lawn surface (i.e., surface contact). The other assessment was for exposure to coliform bacteria by a more intense contact with lawn grass as by ingestion (i.e., oral contact).The procedure used for obtaining a sample for the surface contact exposure assessment was modified from one cited by Scherer and Mitchell (6). A petri dish containing a membrane filter disc on top of an absorbent pad soaked with the appropriate bacteria growth media was pressed onto the lawn surface three times for an effective contact area of 52 |cm.sup.2~ (8.1 square inches).In addition to the contact samples, plant material samples were collected by cutting blades of grass with sterile scissors and placing them into a sterile bag for future laboratory enumeration of coliform bacteria. Determination of the number of coliform bacteria on the plant material permitted an assessment of the exposure risk associated with oral contact with the grass.Bacterial exposure samples were taken during a four-hour period beginning at the conclusion of an irrigation event. Samples were taken every five minutes for the first hour, every 15 minutes for the next two hours, and every 30 minutes for the remainder of the test period.Sample analysisRunoff: Each runoff sample was analyzed for chemical oxygen demand (COD), total nitrogen, mineral nitrogen, total coliform bacteria and fecal coliform bacteria. Each analysis was done in accordance with a method described in Standard Methods for the Examination of Water and Wastewater (4).Bacterial exposure: Press-plates used for assessing the risk of exposure to fecal and total coliform bacteria by surface contact with grass that had been irrigated with treated septic tank effluent were prepared prior to the study and analyzed after exposure as described in Section 909 of Standard Methods for the Examination of Water and Wastewater (4).Grass samples collected from each of the plots were used for assessing the risk of exposure to fecal and coliform bacteria resulting from oral contact with grass that had been irrigated with treated septic tank effluent. The grass analysis involved cutting the grass to approximately 2.5 cm (1 inch)-long pieces and suspending the material in a buffer solution which was shaken for five minutes using a wrist action shaker. After shaking, the buffer solution was examined by the membrane filter technique, as described in Section 909, Standard Methods for the Examination of Water and Wastewater (4).Data analysisA rank mean method was used to statistically compare environmental impact data collected from each of the three lawn plots. This type of nonparametric method was chosen because it is well suited to data sets with large variations in magnitude of observed values. Large variations in magnitude commonly occur when enumerating bacteria from environmental samples. Bacterial exposure data for this study contained several zero observations or observations in which the number of bacteria were below the detectable limits of the enumeration procedure. The rank mean procedure was used to evaluate runoff water quality, plant growth and bacterial exposure.The rank mean method ranks all of the data from the lowest to highest and assigns an integer value starting with 0 to the lowest value (5). The data are then analyzed by means of nonparametric statistics (statistical analysis performed on the rank values assigned to the data instead of on the actual data) to determine if there are significant differences. Using the rank means, standard statistical methods such as Analysis of Variance (ANOVA) may be performed. Comparisons for this project were between data sets from the test plot (irrigated with treated wastewater), irrigated control plot (irrigated with potable water), and dry control plot (no irrigation).RunoffThe study period coincided with a relatively dry period for east Tennessee. There were only 15 precipitation events during the study period that produced runoff from any of the lawn plots. Total precipitation for all of the 15 events was 36 cm (14 inches).Application of treated wastewater to the test plot did not result in a reduction in runoff quality. Concentration of each observed measure of quality in runoff from the test plot was similar to or less than the measure's concentration in runoff from one or both of the control plots. Even though the quality of the water running off the lawn plots was similar, a smaller quantity of each monitored measure of quality was carried off the test plot in runoff. As can be seen from Table 3, the test plot had lower mean values for quantity of all quality measures except in the case of total coliform bacteria. Total coliforms from the test plot were higher than from the dry control plot, but still considerably lower than from the irrigated control plot.The lower quantity of each quality measure transported off the test plot resulted primarily from a lower runoff volume. A complete explanation for the lower runoff volume from the test plot is not available. However, some conditions known to reduce runoff and enhance infiltration did exist on the test plot. Vegetation on the test plot was much more dense and was growing at a faster rate. A more dense vegetation provides a greater capacity for intercepting falling water and impeding its overland flow, thus enhancing the potential for it to infiltrate the soil. A faster vegetation growth rate results in greater transpiration which removes water from soil and, thus, increases its potential for infiltrating water.Rank means were computed and used to check for statistically significant differences in runoff quality among the three lawn plots. Table 4 shows the rank mean and probability values for the various comparisons. An evaluation of the comparisons of rank mean values reveals no disadvantage, with respect to any measure of quality, to the application of treated septic tank effluent onto a lawn. In all cases where there were significant differences, quality-measure values associated with a control plot were higher. (In all cases where statistically significant differences are declared, probability of observed differences occurring by chance was determined to be 0.05 or less. However, all probability values are reported and readers are free to make their own decisions about significance of observed differences.)There is a possibility that the composite sample collected in this study did not represent the quality of all runoff from a given plot, since any runoff in excess of the sample container's capacity dropped into the container and displaced an equal volume which was not saved. However, it is reasonable to expect that even if a composite sample more representative of the total runoff volume was used, the comparison among the plots would have demonstrated similar results. To illustrate this point, let us make the reasonable assumption that water coming off the plots in the latter stages of a runoff event would contain lower quantities of contaminants than that coming off earlier. Such an assumption leads to the conclusion that a lawn plot from which runoff was greatest and the composite sampler TABULAR DATA OMITTED capacity was exceeded more times would erroneously demonstrate less off-site movement of pollutants since the composite sample would contain more of the cleaner water coming off during the latter portion of runoff events. However, the data showed that runoff from the irrigated control plot (with six events where the sample container overflowed) had higher quantities of pollutants removed in the runoff water than either the dry control or test plots (with two and three events, respectively, where the sample container overflowed). This suggests that, even if a composite sample more representative of all runoff was used, the irrigated control plot would have been identified as discharging greater quantities of contaminants.Bacterial exposureOral contact: The risk of encountering either fecal or total coliform bacteria by oral contact with grass was compared among the three lawn plots by comparing the number of bacteria removed and enumerated from grass samples taken from each plot. The assumption with this procedure is that risk of encountering bacterial by oral contact increases as the number of bacteria per unit of grass increases. Table 5 summarizes the comparison made using the rank mean procedure described earlier. For the summer observation period, no increased exposure risk was found for the test plot. The number of total coliform bacteria found in samples from the test plot were not significantly different from those found in samples from each control plot. The number of fecal coliform bacteria were significantly greater from grass in each of the control plots than from grass in the test plot. The only condition for which a clear disadvantage for the test plot was indicated was for exposure to total coliform bacteria during the spring. During all observation periods for fecal coliforms and during summer and autumn for total coliforms, there was either no difference between the test and one control plot, or one of the control plots had significantly higher numbers than the test plot.Surface contact: The risk of encountering a coliform bacterium by surface contact with grass was compared among the three lawn plots by comparing the number of bacterial colonies that developed on the exposed press plates. Table 6 summarizes the comparisons made using the rank mean procedure described earlier. For the summer observation period, there was no difference between the test and dry control plots with respect to the number of fecal or total coliform bacteria found on the press plates. However, the number of bacteria on the plates from the test plot significantly exceeded those on plates from each control plot during each of the other observation periods.Discussion: The oral contact exposure data from the grass samples suggested no clear disadvantage for the test lawn plot (with the exception of the total coliform data during the spring sampling period), without regard to the season. However, the surface contact exposure data from the press plates suggested that the potential for encountering coliform bacteria was greater from the test plot during the spring and fall months but not during the warmer summer months when comparing the test and dry control plots. This observation may ultimately prove to be important in society's process of deciding whether to adopt lawn irrigation with treated wastewater, since the warmer summer months are probably when more humans would frequently contact the grass.TABULAR DATA OMITTEDRestraint is advised, however, in embracing the notion that risk of exposure to pathogens is no greater during the summer from contact with lawns irrigated with treated wastewater than from contact with conventionally managed lawns. Data collected during this study demonstrate that fecal coliform bacteria are naturally present in lawns, and that a person may contact them whether or not wastewater is added to the lawn. However, as is the case with most lawns, the control pots in this study had no obvious source of fecal coliform bacteria from humans. Contact with a fecal coliform bacteria on a control plot does not, therefore, necessarily indicate the same degree of health risk as contact with a fecal coliform bacteria on the test plot where there is a possible human source, since human pathogens are more likely to be associated with coliform bacteria associated with human feces than with animal feces.Results from this study offer encouragement that lawn irrigation with treated wastewater may pose acceptable health risks, and that further efforts to develop such practices are warranted. However, a more conclusive procedure for assessing health risk must be identified and/or developed.ConclusionsResults from this study indicate no reason to expect that daily irrigation of a lawn with treated domestic wastewater will have a detrimental impact on off-site movement of pollutants in runoff. Results also offer encouragement that lawn irrigation with treated domestic wastewater may pose acceptable health risks. All evidence to date suggests that efforts to develop such practice should continue.TABULAR DATA OMITTEDReferences1. Hines, M. and R.E. Favreau (1975), Recirculating Sand Filter: an Alternative to Traditional Sewage Absorption Systems. Proceedings of the National Home Sewage Disposal Symposium, ASAE pub. PROC- 175, St. Joseph, MI. Pp. 130-136.2. Sanjines, A. and B.L. Carlile (1991), Treated Wastewater Subsurface Irrigation Systems, GEOFLOW, Inc., San Francisco, CA.3. Rubin, A.R. (1988), North Carolinians Match Treatment Systems to Soil, Site Conditions, Small Flows, West Virginia University, Morgantown, WV. June, p. 10.4. American Public Health Association (1985), Standard Methods for the Examination of Water and Wastewater, 16th edition, Washington, D.C.5. SAS Institute Inc. (1985), SAS Procedures Guide for Personal Computers, Cary, NC. P. 373.6. Scherer, B.P. and D.T. Mitchell (1982), Individual Household Surface Disposal of Treated Wastewater Without Chlorination. Proceedings of the Third National Symposium on Individual and Small Community Sewage Treatment. ASAE Publication 1-82, St. Joseph, MI. Pp. 207-214.C. Roland Mote, Ph.D., P.E., Agricultural Engineering Dept., The University of Tennessee, P.O. Box 1071, Knoxville, TN 37901-1071.