In 1993, Olmsted County installed an advanced individual sewage treatment system (ISTS) at the caretaker's residence at a County-owned park to demonstrate waste load reduction and downsizing of drainfield area. The subject system has two components--1) a composting/vermiculture blackwater system and a 2) aerobic graywater treatment system. The separation of the high strength human feces and urine from other wastewater generated by the household is the key feature of this system. Kitchen food wastes are processed in the BMRC (Bio-Matter Re-Sequencing Converter), while sink, bath, dishwasher, and clothes washing wastewaters are handled in the aerobic graywater system. As part of the demonstration, monitoring was conducted for a range of chemical and physical wastewater parameters, as well as for water, electric, and drainfield use. Test results indicate that the system is operating as expected with regards to effluent nitrogen, organics, and fecal coliform levels, as well as electrical use, airflow, and drain-field use.
Summary and Conclusions
Based on the monitoring data and on characteristics of typical untreated domestic wastewater, the system is calculated to have achieved a 90% or greater reduction in loadings of total nitrogen, BOD, and suspended solids to the drainfield. Fecal coliform levels are calculated to be up to 1,000 fold less than effluent from conventional septic tank effluent. Measured water use, measured pollutant loadings, and the lack of ponding in drainfield trenches, are consistent with the original design of a 40% downsized soil treatment area. Drainfield use in the first four years of operation is consistent with a projected life expectancy of at least 20-25 years. The volume of compost accumulated since commissioning the unit is consistent with the vendors' projection. This material will need to be removed from the unit within the next year and a new charge of bulking material will be necessary. Best management practices for finished compost will be determined based on further compost testing. Currently, the handling of this material is covered by the septage and biosolid rules MN Statutes 7040.
Description of the Wastewater Treatment System and Process
The Equaris Corporation system, chosen for the subject demonstration, is shown schematically in Figure 1. To maintain an aerobic composting process, the compost mass must be kept moist but not saturated with water. Hence special low-flow toilets are integral to the design and proper operation of the overall system. The toilet on the upper level is a Neponâ unit that uses about 200 ml of water and a drop of soap to make a foam slurry that flushes wastes via gravity to the BMRC. The lower level toilet is a vacuum flush SeaLandâ unit that utilizes approximately 500 ml per flush and is carried by small diameter pipe to the BMRC. The kitchen garbage disposal contained in a separate sink (again to minimize water addition) discharges via gravity to the BMRC. The BMRC and graywater tanks are both located in a basement utility room.
The BMRC has a total volume of about 3,600 liters and can accumulate about 2,000 liters of compost. The tank is initially charged with about 300 kg of pine bedding as a bulking agent and carbon source (human wastes are typically deficient in carbon relative to nitrogen for proper composting). Additional pine bedding is added periodically to maintain aerobic composting conditions. Rotating agitators distribute fresh wastes over the compost mass. Two types of worms are used to aid in the decomposition of the organic material. Warm household air is drawn through ducts imbedded in the compost mass to maintain temperature and aerobic conditions; exhaust air is vented to the outside. Bulk food scraps and other organic wastes can be added manually to the tank. Screw augers at the bottom of the compost tank provide for periodic compost removal.
Liquid draining from the BMRC is collected at a sump in the front of the tank and is periodically resprayed over the compost to maintain proper moisture. Water not evaporated or resprayed (compost leachate) is pumped intermittently to the aerobic graywater treatment tank.
The graywater tank has a design capacity of about 1,000 liters per day and is about the same size as the BMRC. The tank is divided into three compartments for staged treatment. The first compartment provides settling and surge protection, while the second and third chambers provide for aeration and clarification, respectively. A Zabelâ screen filter is installed in the outlet from the tank to provide additional solids retention. Plastic wiffle-type balls in the aeration and clarification tanks provide a large surface area for biofloc growth. The overflow from the graywater system (treated graywater) is discharged intermittently to an outside dosing tank (3,800 liters). Dose tank effluent is metered to a drainfield consisting of five trenches, each approximately nine meters in length. Drop boxes and inspection pipes are installed on each of the trenches. The drainfield area is about 40% smaller than a conventional system drain-field based on water savings from the low-flow toilets.
Wastewater Parameters/Effluent Characteristics
Grab samples were taken from the clarifier chamber of the graywater system each week for four weeks and then each month for four months. Table 1 presents the data for treated graywater Values for parameters related to the organic components (BOD, TOC, COD) and indicates that the effluent from the system is of typical domestic origin (Eckenfelder, 1970). Fecal coliform levels are in the 103 range compared to reported values of septic tank effluent of 104 to 108(EPA). Phosphorus levels in the treated graywater are higher than predicted for domestic graywater for households in the Great Lakes States where phosphorus in laundry detergents is banned. This discrepancy is attributed to the use of phosphate detergents purchased via mail (11.4% P or 6.4 grams/ laundry load as recommended).
Although dissolved oxygen (DO) levels range from 2 to 5 mg/l, the system is not expected to achieve nitrification due to relatively short retention times. Of two analyses for nitrate-N, none was detected in the first and 1.0 mg/l in the second. Since organic carbon is preferentially metabolized over organic-N, it is expected that much of the nitrogen would remain in the organic form, and as ammonia, with only a small amount converted to nitrate (both organic-N and ammonia-N are measured by the TKN analysis).
The variability of individual parameters on different sample days is shown in Figure 2. Total Solids, N & P, and BOD, COD, TOC (the measures of organic loading) generally tend to increase or decrease together. This suggests that the measured variability in concentration is more related to waste water volume than to variability in the characteristics of the waste itself. Relative Standard deviations of the means are less than 23% for all wastewater parameters except fecal coliform.(data for fecal coliform are not normally distributed).
Compost Leachate Characteristics
Because the compost leachate is diverted to the graywater aerobic tank, separate samples of the leachate were taken on two occasions to assess the relative contribution of the 'blackwater' to the bulk graywater. The compost leachate (Table 3) is characterized by a relatively high organic loading (COD and TOC), high nutrient loading (TKN and TP), and the presence of fecal coliform. The leachate is colored by organics but is not especially turbid. Direct measurement of the compost leachate volume was difficult since this stream is introduced intermittently into the graywater system. However, direct observation of this system and measurements at other Equaris installations indicate the volume is generally less than 10 liters per day. The compost leachate volume is estimated to be less than 3 percent of total graywater effluent volume based upon comparison of the leachate mass loadings for individual constituents to the total graywater effluent loadings for those same constituents. Fecal coliform from the leachate could account for up to half of the coliform bacteria count in the graywater effluent by similar calculation. As a comparison, leachate fecal coliform levels are in the 103 to 104 range while reported values for septic tank effluent range from 105to 108(EPA).
The BOD5 values of 'zero' in Table 3 are not thought to be representative due to the lack of appropriate sample seeding at the laboratory. BOD is primarily used as a process control variable -- test results are dependent upon the biological process from which the seed source is derived. TOC and to a lesser extent COD are less subject to this limitation. Ratios of BOD/COD and BOD/TOC in the leachate would be expected to be lower than in typical raw domestic wastewater since the compost process is aerobic, and much of the soluble organic matter available to the organisms involved in the BOD5 test has already been converted to residual organics or carbon dioxide.
Outdoor and Indoor Water Use
Table 4 summarizes the cumulative and per capita water use at the caretaker's residence since early 1995. Outdoor water use is much higher than expected. This is largely attributed to purging of the well and water lines as a means of reducing iron levels in the water supply. Iron levels were measured at 3.0 mg/l and aesthetic problems generally occur at levels greater than 0.3 mg/l.
Indoor water use is slightly less than the national average. It was expected that the very low flow toilets installed at the site would result in even lower per capita water use. These toilets use about seven liters per capita per day (lcd) compared to almost forty lcd for standard toilets. Apparently savings from toilet use were offset by increased water use elsewhere in the household.
The drainfield area at Chester Woods was designed for a flow of about 1,000 liters/day. Had standard toilets and septic system been installed at the site, the drainfield would have been sized at about 1,650 liters/day. Actual flow has averaged about 355 liters/day (or 177 lcd) or about one-third capacity.
Incremental energy use attributed to the Chester Woods installation includes direct electricity and indirect heat-loss with vented air. While the intent was to separately measure electrical consumption of the wastewater treatment system components (aerobic tank, composting tank, toilets, sump pump), other co-located appliances were connected to the same electrical circuit(water softener, dehumidifier). Data from other Equaris installations indicate the pump supplying air to the graywater tank to be the major electricity user At seven cents per kilowatt hour, the monitored electricity costs about $11.50/month.
There is more air exhausted from the Chester Woods house than would normally be the case. Incremental heat-loss attributed to this type of installation is estimated at 5 million btu/month through the heating season. This represents about $25/month for propane. Some of this heat-loss would occur in any case with bathroom venting in a conventional setting. Passive air-to-air heat exchange on the compost tank vent could reduce some of the loss. A benefit of the exhaust design is that bathroom air is exhausted through the Neponâ Toilet system reducing typical bathroom odors.
To compare the Chester Woods system to a conventional ISTS, the residuals must be considered. Septic systems generate about 4,000 liters of septage every three years (1,200 liters/year) of approximately the composition shown in Table 8 (EPA, 1993). The caretaker's residence is expected to produce about 2,000 liters of compost per five years (around 400 liters per year) with the approximate characteristics shown in Table 9. Actual measurements of the exact volume or characteristics of compost at the Chester Woods site have not been made to date. Compost characteristics will be the subject of future testing.
Although most of the bulking agent was added at the time of BMRC system installation, some additional bedding was added periodically. It is estimated that an average of about 55 kg/year of pine bedding (equal to about 22 kg/yr carbon) was added. The amount of pine bedding that would be theoretically needed to compost the human wastes could be calculated from published values of carbon/nitrogen (C/N) ratios. The actual amount of pine bedding needed will depend on the carbon contributed to the BMRC from toilet paper and food wastes. Carbon contributions at the site are estimated at 11 kg/cap/year each for human wastes, toilet paper, and pine bedding, and 5 kg/cap/year for food wastes. Based on visual observations, it appears that these carbon additions do produce a stable finished compost in the system.
Calculated Reduction in Pollutant Loadings
For purposes of comparing the Equaris system to other treatment systems, the data from Tables 1 and 3 have been converted to loadings in grams/capita/day (gcd). Results are shown in Table 6 along with published values for per-capita waste generation for graywater, blackwater, and combined raw domestic waste. Based on these generic loading rates, the Chester Woods system is calculated to achieve greater than 94% for TSS and BOD5, greater than 83% reduction in total organics, and over 90% reduction of total-N. Nitrogen and phosphorus levels in the graywater effluent were similar to the published values for raw graywater. The calculated removal rates were largely due to the separation and retention of the solids and the nitrogen in the compost mass, and through nitrogen loss to the atmosphere as ammonia, and possibly through denitrification. Additionally, there may be some retention of organics and nutrients in the graywater tank biomat.
Much of the phosphorus measured in the graywater is thought to originate from household products rather than human waste (as discussed earlier). Like other aerobic wastewater treatment systems, the Chester Woods system would not be expected to remove phosphorus above the small amounts, which accumulate in the aerobic tank biomat. This is consistent with the data.
In Table 6 the Chester Woods effluent was compared to raw domestic graywater and blackwater on a per capita loading basis. Table 7 illustrates the degree of reduction calculated for TSS, TKN, BOD5, and coliform compared to reported septic tank effluent (in mg/l). While these concentration comparisons may not exactly relate 'one-to-one' to per capita loadings, they do show the same general pattern of nitrogen and organics reduction as described above. Calculated reductions for the Chester Woods site are less when compared to septic tank effluent (Table 7) than when compared to raw domestic wastes (Table 6), as would be expected since some settling and removal of nitrogen and organics occurs in septic tanks.
Comparison of Costs
The cost of the Equaris system including the BMRC tank, aerobic graywater tank, and toilets, was $8,050. The drainfield (five - 9 meter trenches) and 4,000 liter dose tank cost $4,690. For representative comparisons with other individual sewage treatment systems, $500 in toilet costs must be subtracted from the Chester Woods system costs (toilets are generally not considered part of the wastewater treatment system). The total cost of the Chester Woods system is then $12,240 -- this included $800 in shipping (note that the manufacturer has since relocated from Alaska to Minnesota). These are strictly capital costs and do not include the routine operating costs, for example utilities and septage hauling. For comparison of capital costs, Table 11 summarizes costs for six individual sewage treatment systems (note that local costs are estimated to be slightly higher than those shown below). These systems do not significantly reduce nitrogen loadings and only the aerobic tank system allows a downsizing of drainfield area.
While the cost of the Chester Woods system is more than conventional systems, the subject installation was the first of its kind in Minnesota and the costs were higher than if a larger number of units were produced and installed. Our assessment indicates that much of the treatment advantage of the Equaris system is associated with the composting component, while a disproportionate cost per degree of treatment is associated with the graywater component (both capital and operating). This may be important in highly sensitive sites, such as shorelands or where direct surface discharge is proposed. At many other sites, it may be more cost effective to utilize a downsized soil treatment system for graywater, rather than the Equaris graywater treatment component.
User Acceptance and System Problems
After initial startup, minor odor episodes occurred with the graywater system. This was attributed to shock loading of accumulated compost leachate in a special sump provided for sampling. Removal of the sump after compost leachate sampling was completed eliminated this problem.
An episodic fruit fly hatch occurred shortly after a modification was made to the graywater system to increase surface area for biofloc activity. The floating wiffle balls provide a surface area above the water line that appears to provide a hatching surface for the flies. The problem has diminished as the balls accumulate biofloc and submerge. Venting the tank to the outside may further alleviate the problem.
When it was observed that there was unintended seepage between internal chambers in the graywater tank, the structure was reinforced. Pump motors on the SeaLand Toilet (lower level) have required replacement. Several sheer pins in the compost tank were broken during the initial startup period.
Many people and organizations contributed to making this project a success. Olmsted County, through the Research and Development Committee, provided financial support for installing the Equaris system. Lawler Environmental Services provided the engineering and design work for the dosing tank and soil treatment system. The Minnesota Pollution Control Agency, through the Clean Water Partnership Program, provided funding for monitoring. Rochester Water Reclamation Plant provided BOD5, TSS, TOC, and TP analyses. Rochester Public Utilities provided water meters. Peoples Cooperative Power Association provided electric meters. The South Zumbro Watershed Partnership conducted the sampling, and Equaris Corporation provided technical assistance. Special recognition must be given to Earl and Barb Bosshardt, caretakers at Chester Woods Park, whose participation made the project possible.