September 1974 Environmental Protection Technology Series DEMONSTRATION OF WASTE FLOW REDUCTION FROM HOUSEHOLDS £ 55 \ W \ UJ CD National Environmental Research Center Office of Research and Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 ------- EPA-670/2-74-071 September 1974 DEMONSTRATION OF WASTE FLOW REDUCTION FROM HOUSEHOLDS by Sheldon Cohen and Harold Wallman General Dynamics Electric Boat Division Groton, Connecticut 06340 Contract No. 68-01-0041 Project No. 11010 GXJ Program Element No. 1BB033 Project Officer Harry E. Bostian Advanced Waste Treatment Research Laboratory National Environmental Research Center Cincinnati, Ohio 45268 NATIONAL ENVIRONMENTAL RESEARCH CENTER OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268 ------- REVIEW NOTICE The National Environmental Research Center, Cincinnati, has reviewed this report and approved its publication. Approval does not signify that the contents necessarily reflect the views and policies of the U. S. Environ- mental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. 11 ------- FOREWORD Man and his environment must be protected from the adverse effects of pesticides, radiation, noise and other forms of pollution, and the unwise management of solid waste. Efforts to protect the environment require a focus that recognizes the interplay between the components of our physical environmentair, water, and land. The National Environmental Research Centers provide this multidiscipli- nary focus through programs engaged in: o studies on the effects of environmental contaminants on man and the biosphere, and o a search for ways to prevent contamination and to recycle valuable resources. This report covers work on conservation of water in the home. Through the use of reduced flow plumbing fixtures and wastewater recycle systems, it has been demonstrated that significant reductions in water use and sewage flow can be achieved. A. W. Breidenbach, Ph.D. Director National Environmental Research Center, Cincinnati iil ------- CONTENTS FOREWORD ill LIST OP FIGURES V LIST OF TABLES vi ACKNOWLEDGMENTS vll Sections I CONCLUSIONS 1 II RECOMMENDATIONS 5 III INTRODUCTION 7 IV NORMAL WATER USE PATTERNS 11 V BATHROOM FLOW REDUCTION DEVICES: 20 SELECTION, INSTALLATION AND PERFORMANCE VI RECYCLE SYSTEMS: DESIGN, DEVELOP- 42 MENT, INSTALLATION AND PERFORMANCE VII HOMEOWNER ACCEPTANCE 85 VIII COST ANALYSIS 89 DC REFERENCES 96 X APPENDIX 98 ABSTRACT AND REPORT DATA 103 lv ------- FIGURES No. Page 1 Schematic Diagram for Homes With Flow 8 Restricting Shower and Dual Cycle or Shallow Trap Toilet 2 Schematic Diagram for Homes With Flow 9 Restricting Shower, Dual Cycle or Shallow Trap Toilet, and Wash Water Recycle System 3 Typical Data Collection Form 13 4 Shallow Trap Toilet 21 5 Econo-Flush Toilet Device 24 6 Sink-Bob Toilet Device 26 7 Saveit Toilet Device 27 8 Econo-Flush Toilet Device: Installed 28 9 Saveit Toilet Device: Installed 30 10 Speakman Auto-flo Shower Head: 13.3,1pm 31 11 Speakman Auto-flo Shower Head: 9.5 1pm 32 12 Recycle System With Cartridge Filter 43 13 Recycle System With Diatomite Filter 44 14 Recycle System With Cartridge Filter: Installed 47 15 Recycle System With Diatomite Filter: Installed 48 16 Polyethylene Storage Tank 55 17 Low Level Control System 56 18 Low Level Control Float Rod Assembly 57 19 Typical Data Collection Form - Recycle System 60 20 Diaclear Diatomite Filter 61 21 Diatomite Filtration System: Installed 62 22 Filter Pressure Drop vs. Time 65 23 AMF/Cuno Cartridge Filter 67 24 Pram MCM Cartridge Filter 69 25 Air Lift Clorox Feeder 75 26 Air Lift Feeder Installation 76 27 Chlorine Tablet Feeder 77 28 Pressurization System ------- TABLES No. Page 1 Water Savings Summary ^ 2 Test Home Characteristics 12 3 Distribution of Household Plow Reduction Devices 12 4 Phase I - Water Consumption Data 1^ 5 Phase III - Water Consumption Data 15 6 Phases I & III - Water Consumption Data 15 7 Average Household Water Use. Test Data (Phases 16 ; I & III) '8 Statistical Summary of Normal Water Use Patterns 1? 9 Hot Water vs. Cold Water Usage for Bath and 19 Laundry, Ipcd 10 Comparison of Water Closet Plush Volumes 23 11 Test Period Water Consumption Data 33 12 Statistical Summary of Water Use Patterns During 34 Test Period (Phase II) 13 Statistical Significance of Individual Water 36 Savings 14 Water Savings Obtained With Shallow-Trap Toilets 37 15 Water Savings Obtained With Dual Flush Devices 39 16 Water Savings Obtained With Plow Limiting Shower 40 Heads 17 Shower Hot Water Savings 41 18 Effect of Lawn Sprinkling on Various Soil Char- 51 acteristics 19 Wash Water Recycle System: Flow Reduction Summary 52 20 Diatomite Filtration System Performance Data 64 21 Pram MCM Cartridge Filter Performance Data 70 22 AMF/CUNO CG4-DC1 Cartridge Filter Performance Data 72 23 Filter System Performance Summary 73 24 Clorox Air Lift Feeder Performance Data 79 25 Chlorine Tablet Feeder Performance Data 82 26 Summary of Questionnaire Results 86 27 Cost Summary - Bathroom Water Saving Devices 90 28 Cost Summary - Wash Water Recycle System 91 29 Cost Comparison 93 vi ------- ACKNOWLEDGMENTS This report was submitted in fulfillment of Project 11010 GXJ and Contract 68-01-0041 by General Dynamics, Electric Boat Division under the sponsorship of the Environmental Protection Agency. The experimental portion of the project ran from May 1971 to May 1973. The performance of this project was carried out under the guidance and supervision of Harold Wallman, Project Manager, and Sheldon Cohen, Project Engineer. The valuable assis- tance of Donald E. Leone, Senior Biologist/Microbiologist throughout the course of the program is gratefully acknow- ledged. The assistance of C. Douglas King, Life Sciences, Convair Division of the General Dynamics Corporation, in coordinating activities in the San Diego area is acknow- ledged with sincere thanks. Acknowledgment is made of the cooperation given by the following volunteer homeowners and their families during the two-year demonstration program: Messrs. Andrew J. Ciminera, George J. Erkan, William J. Fish, Roy F. Holmes, C. Douglas King, Donald Manley, Paul Murphy, and Gordon W. Thomson. The support given the project by Dr. Harry E. Bostian, Project Officer, Advanced Waste Treatment Laboratory, Environmental Protection Agency, is acknowledged with sincere appreciation. vii ------- SECTION I CONCLUSIONS WATER SAVING TOILET DEVICES The results of the program show that water requirements for toilet flushing can be substantially reduced by commerci- ally available devices, and in an acceptable manner in terms of functional, economic, and aesthetic considerations. All of the water saving devices tested were convenient to oper- ate, provided satisfactory operation and performance, and scored very well in terms of user acceptance. Water saving (shallow trap) toilets were found capable of providing reductions in water closet consumption by way of flush volume reductions afforded with the six units tested. The average reduction observed (25$) was somewhat less than anticipated because of the generally lower than average flush volumes of the original conventional toilets. Be- cause of the smaller flush volume utilized by the shallow trap toilet, proper adjustment of the tank level and fill rate was found to be important in achieving satisfactory performance. The unit costs no more than its standard size counterpart, and is definitely warranted for new homes or necessary replacements. However, because of its relatively high initial cost, the shallow trap toilet appears to be of limited economic value for replacement of workable toilets. Toilet insert devices which converted conventional toilets to dual cycle operation yielded average reductions in toi- let flushing requirements of from 18$ (ECono-Plush) to 26# (Sink-Bob) by achieving respective flush volume reductions of from 25$ to 50$ for liquid wastes. The design of the toilet bowl was found to be important in terms of the ade- quacy of the reduced flush for liquid and/or solid wastes. Hence, although the Sink-Bob device has greater water- saving potential, its effectiveness would be more depen- dent upon the design of the toilet in which it is installed. ------- Installation of the toilet inserts in most standard toilets can be readily accomplished by the average homeowner. How- ever, some of the newer toilet models have a rather elab- orate flush valve design which would preclude the insertion and use of the devices tested. Due to their low initial cost, as well as their ability to produce significant re- ductions in water usage, toilet insert devices possess a greater potential for cost savings than any of the other bathroom devices tested. Increased usage and production of the toilet devices should result in even lower initial costs additional innovative designs, such as the Saveit toilet insert which was given a preliminary evalua- tion (when modified for dual flush operation, its flow re- ducing potential appeared similar to the Econo-Flush), should promote greater acceptance of these devices by the general public. FLOW LIMITING SHOWER HEADS The adequacy of a reduced shower flow rate of 13.3 Ipra (3.5 gpm) was clearly established in terms of providing an ade- quate water supply and spray pattern. A further reduction in flow to 9.5 1pm (2.5 gpm) proved unsatisfactory to 33$ of the participants. Despite their low cost, the flow limiting shower heads proved to be of less economic value than anticipated largely because of the rather limited water savings obtain- ed with these devices. This can be traced primarily to the personal bathing habits (lower than average bathing fre- quency as compared to previous surveys1 and preference for tub baths as indicated by personal interviews) of the pro- gram participants and does not necessarily invalidate their potential. Significant savings in hot water heating costs, in addition to water savings, should also be achieved with flow reducing showers where more frequent shower usage is encountered than in this test program. WASH WATER REUSE FOR TOILET FLUSHING AND LAWN SPRINKLING In three of the test homes, the reuse of waste wash water (bath and laundry) for toilet flushing and lawn sprinkling was successfully demonstrated throughout the one-year test period. The recycle system proved to be manageable and simple to use, and capable of reliable and safe operation. No impairment of toilet operation was observed due to the recycling of filtered, disinfected wash water. No signif- icant effects, adverse or beneficial, were noted on lawn growth or appearance throughout the test period or during the next growing season. ------- In general, the supply of waste bath and laundry water collected sufficed to meet the demands for toilet flushing. The supplemental feedwater requirements averaged only 5# of the total water reused. A 380 liter (100 gallon) storage tank should provide an adequate reservoir for an average size family. The Incorporation of lawn sprinkling was found to be useful as a supplemental reuse mode in those homes with an excess of wash water in order to reduce the waste overflow to the waste treatment (septic) system. The performance attained by the diatomite filter system proved to be satisfactory in terms of aesthetic acceptance by the homeowner and was achieved with relatively low oper- ating costs. Maintenance requirements were found to be similar to those involved in the operation of a small swimming pool. The residual suspended solids in the re- cycled wash water did produce temporary stains in the toilet bowls which tended to increase cleaning requirements. No problems due to foaming were experienced during refill of the toilet tank. Continual chlorination of the stored wash water with rela- tively simple and inexpensive feeders effectively inhibited bacterial growth, and no unpleasant odors were detected during normal operation. Stable chlorine residuals were maintained at the point of reuse despite the presence of interfering agents (soaps, detergents), and any potential health hazard for toilet flushing reuse is considered ex- tremely remote. Reuse for lawn sprinkling may present a possible hazard because of the greater accessibility to the recycled wash water, should any pathogens be present. This potential hazard can be eliminated by suitable underground discharge, but with substantial increased cost. A comparison of the relative effectiveness of each of the four different kinds of units tested, in terms of water savings, is summarized in Table 1. As anticipated, the wash water recycle system produced the greatest savings in total water consumption. The overall average water savings for toilet flushing and/or lawn sprinkling was 305 Ipd, corresponding to a reduction in total water consumption of Although the wash water recycle system afforded approxi- mately three to four times as much water savings as was provided by either the shallow trap toilet or dual flush devices (26# vs. 6.9-8.6$), the projected annual cost for a mass produced recycle system of $43/yr is about one to two orders of magnitude greater than the annual cost for the bathroom flow reduction devices. A comparison of the ------- projected recycle system cost of $0.39/1000 liters with the water and sewer use rates in Connecticut indicates that the system can effect marginal cost savings for single homes in high water and sewer use rate areas, and becomes economi- cally attractive only when septic systems with poor drain- age are encountered. Cost comparisons with and without septic systems and for ranges of water and sewer charges are shown in Section VIII, Table 29. Table 1. WATER SAVINGS SUMMARY Unit tested No. of units tested Average no. of occupants Water savings <$> reduction in total water usase Wash water recycle system Shallow trap water closet Dual Sink-Bob flush device Econo- Plush Flow limiting shower head 11 6.3 4.5 2.8 4.6 44.0 14.8 20.5 12.4 2.7 26.0 6.9 8.6 1.0 Ipcd = liters per capita-day The reuse of waste wash water for toilet flushing and/or lawn sprinkling can provide tangible benefits above and beyond the savings in water consumption. The recycle system, by minimizing the surges in outflow to the septic system associated with laundry and bath discharges, as well as by reducing total waste flow, allowed the septic tank and soil absorption system to operate more effective- ly. In two of the three homes, septic system backup nor- mally experienced was not observed throughout the test period. ------- SECTION II RECOMMENDATIONS 1. In order to promote greater public awareness of both the need for and benefits of water conservation, as well as the availability of flow reducing devices for both toilets and showers, broad-based educational/demonstration programs should be considered in selected areas. Such large-scale programs will also generate data with a higher degree of statistical significance than was possible In this program. The water saving customer education and appliance test pro- gram recently completed by the Washington Suburban Sanitary Commission2*3 is an excellent example of the utilization of both software (water saving handbooks) and hardware (toilet inserts, flow limiting shower heads, and inlet pressure re- ducing valves) in a comprehensive effort aimed at promoting water conservation by the general public. More demonstra- tions can serve to stimulate further development and mass production of water conservation devices. In Implementing such broad-based programs, a particularly fruitful approach might be to induce participation by the major toilet manufacturers in adaptating one of their standard models for use as a dual cycle toilet. The devel- opment of such toilets would preclude any compatibility problems which may exist between certain toilet inserts and some of the current toilet models, provide much greater ex- posure for this approach to water conservation, and assure professional installation by an experienced plumber. 2. The extension of the wash water reuse concept to mul- tiple family dwellings should be investigated in order to determine its potential attractiveness for water conserva- tion and waste flow reduction. Reuse on a multiple-dwell- ing basis should be more economical than for a single dwelling. Scale-up of the system developed for single family dwellings should not only permit substantial reduc- tions in both initial and operating costs, on a per capita basis, but also allow significant improvements in recycle system design and performance with minimal impact on total system costs. The need for installation of a separate ------- distribution system for toilet flushing should not alter the economics in a decisive manner. All of the foregoing pre- sumptions need to be assessed in the light of further de- tailed design and cost analyses. In addition, public atti- tudes toward the reuse of waste wash water on a communal basis will have to be further explored. Public safety in regard to the possible transmission of pathogenic organisms must also be carefully considered. It is therefore recom- mended that a preliminary study phase be conducted, prior to the design and construction of a prototype unit, in order to ascertain economic feasibility, public acceptance, and any potential health hazards. 3. The results of this project show that reuse of waste wash water at the single household level for toilet flushing and lawn sprinkling can be applied successfully. Also, the projected economics look marginally favorable for high user charge areas and where septic systems are flow limited. Further work on fixing the cost of a mass produced reuse system for single family dwellings could make the economics more favorable. It is therefore recommended that a larger- scale demonstration program be considered, for single family dwellings, which would incorporate as an initial phase the development and detailed costing of a prototype unit suitable for mass production. Such a program would (a) result in a cost and performance optimlz-ed system, (b) clearly establish the acceptability and safety of this approach to the homeowner, and (c) could lead to the avail- ability of a marketable unit. 4. The previous recommendations regarding home water con- servation are Justifiable on the basis of cost savings to the individual homeowner. On the larger scale of net cost savings to the community, however, there may be more effective ways of conserving water and reducing waste flow. Technological concepts related to wide scale reuse, reduc- tion of infiltration and integration of utilities could be more important than home water saving. User attitudes and rate structures could also have a determining influence. It is therefore recommended that a broad study be made of the technological, economic, public health, aesthetic and sociological factors relative to water conservation and waste flow reduction, in order to determine where the most incentive exists for further research, development and demonstration. ------- SECTION III INTRODUCTION BACKGROUND The austere reality of a fixed natural fresh water supply combined with a continuously expanding demand has pro- moted an increasing awareness of the need for developing new long-term approaches to water management. In recent years, this need has become more acute as the country becomes increasingly involved with controlling water. pollution. According to the Office of Saline Water,^ our water needs are expected to double by 1980 (from 1.5 to 3 billion cu m/day or 400 to 800 billion gal/day), and by the year 2000, total water consumption for the United States may be as much as 4 to 8 billion cu m/day. In a previous engineering study performed by General Dynamics for the Federal Water Pollution Control Adminis- tration,1 it was concluded that reduction of water usage appeared to be the most practical and economically fea- sible approach to both water conservation and waste treat- ment at the household level, and would not become obsolete as new treatment technology is developed. The study cited many household functions in which water is being used wastefully. In particular, the study indicated that water for toilet flushing and bathing could be reduced by approx- imately 35$ by using presently available devices and technology. The practice of water reuse at the household level is another possible means of reducing both water usage and waste treatment requirements. In particular, the use of waste wash water for toilet flushing (or lawn sprinkling) appears to be the simplest and most practical method of conserving water through reuse. In a city of 100,000 these savings could amount to more than eight thousand cubic meters of water per day that would not have to be supplied to the user and eventually treated in the waste treatment plant. Besides savings in operating costs for water supply and waste treatment, the decreased usage would delay the need for construction of new waste treatment facilities, for construction of larger sewer lines and ------- water mains, and for the development of new water sources which are becoming increasingly scarce and costly. PROGRAM OBJECTIVES AND SCOPE OP WORK The main thrust of this program was to evaluate the rela- tive merits, water savings, costs, acceptability, and potential attractiveness of (a) various presently avail- able bathroom flow reduction devices, and (b) schemes for the reuse of household wash water for toilet flushing and/ or lawn sprinkling. The flow reduction devices and re- cycle systems were installed on a voluntary basis in a representative group of eight homes to reduce water con- sumption in three major water use areas: (l) toilet flushing, (2) bathing, and (3) lawn watering. In five of the homes, shallow trap water closets and/or dual-flush devices were used to reduce the water requirements for toilet flushing, and flow limiting shower heads were used to reduce bathing water consumption (see Figure l). In the three remaining homes, in addition to the flow reducing de- vices, an alternate approach was implemented in which waste laundry and bath water were filtered, disinfected, and pressurized for reuse in toilet flushing and/or lawn water- ing (see Figure 2). FLOW I RESTRICTING SHOWER (& BATH) , METER) DUAL CYCLE OR SHALLOW TRAP TOILET LAWN WATER Figure 1. Schematic diagram for homes with flow restrict- ing shower and dual cycle or shallow trap toilet 8 ------- I METER LAUNDRY MAQHINE FLOW RESTRICTING SHOWER I (BATH) 1 FV r~ OVERFLOW DISINFECTANT FEEDER STORAGE TANK T METER] METER] LAWN WATER DUAL CYCLE OR SHALLOW TRAP TOILET PS FILTER [ PRESSURE A I TANK I NOTE: CV = CHECK VALVE FV = FEED VALVE PS = PRESSURE SWITCH Figure 2. Schematic diagram for homes with flow restrict ing shower, dual cycle or shallow trap toilet, and wash water recycle system ------- The work performed in this program was accomplished in three major phases as outlined below: 1. Meters were installed in the eight test homes to determine normal water use patterns for a period of six months. During this period, the design, selection and procurement of all necessary devices and systems were accomplished. 2. The bathroom flow reduction devices and recycle systems were then installed in the eight test homes for a period of one year. Throughout this period, both water consumption and performance of all devices and systems were continually monitored. 3. The water saving devices and systems were then removed, and the monitoring of normal water usage was continued for an additional six months to minimize any bias in the results due to seasonal changes or long-term variations in family habits. In addition to the analysis and evaluation of all data collected during the program, all systems and devices tested were evaluated in terms of homeowner acceptability through the use of informal interviews and formal ques- tionnaries completed by all adult occupants. 10 ------- SECTION IV NORMAL WATER USE PATTERNS HOMEOWNER SELECTION The homes of eight volunteer families were chosen for the test program. To facilitate installation, inspection, and data collection, six of the homes were chosen in the vicin- ity of the Electric Boat Division facilities (four in southeastern Connecticut and two in adjacent Rhode Island). Two additional homes were chosen in a water-short region (San Diego, California). The homes and families were care- fully chosen to obtain representative conditions of water use and typical installation costs. Additional guidelines for selection of the demonstration homes included homeowner interest and cooperation, single family dwelling with 3 or more occupants, accessible plumbing for meter installation, and drain lines suitably located for installation of re- cycle systems. Various relevant characteristics of each of the test homes and families are shown in Table 2. The number of occupants (adults and children) listed in the table reflect condi- tions at the start of the two-year test program. Due to fluctuations in the number of occupants for some of the families throughout the program, the final average number may not be the same as that shown in the table. Toilet characteristics are also included to serve as a basis for correlation of the water use data obtained during both control and test periods. Table 3 indicates both the num- ber and specific distribution of each of the household flow reduction devices and systems installed in each test home. HOUSEHOLD WATER CONSUMPTION DURING CONTROL PERIODS (PHASES I AND III) Throughout most of the two-year test program (May 1971 - May 1973), water usage was recorded by each of the eight homeowners, on a weekly basis, on data forms similar to that shown in Figure 3. In all eight homes, toilet, bath/ 11 ------- shower and inlet (total) usage were monitored. In five of the homes, laundry usage was also monitored, and the two homes in San Diego (#1 and #2) recorded lawn water consumption. Table 2. TEST HOME CHARACTERISTICS Single family homes #1 #2 #3 #4 #5 #6 #7 #8 Service water Water pressure Supply cm He City (metered) City (metered) Individ- ual well City (metered) City (metered) City (metered) Regional (metered) Regional (non- metered ) 284 320 104- 208 232 475 300 388 284 Waste disposal Sewers Sewers Septic tank Sewers Sewers Septic tank Septic tank Septic tank Adults5 1-W 1-NW 1-W 1-NW 1-W 1-NW 2-W 1-W 1-NW 1-W 1-NW 1-W 1-NW 1-W 1-PT b Children 1-S 2 -PS 1-S 1-PS 2-S 1-S 7-S 1-PS 3-S- 3-S Number of Toilet baths dssiKn 1 Reverse trap ^ Siphon jet 1 Reverse trap 1 Siphon Jet 1 Siphon jet 1 Siphon Jet 2 Reverse trap Wash- l|r down Plush vol., liters 17.2 14. 3/ 16. 3/ 16.7 12.5 16.3- 20.8 14.4 17.9 13. 6/ 17.0 15.5/ 15.5 a W=W0rking NW=Non-w orking PT=»Part time S=School PS=Pre-school Table 3. DISTRIBUTION OF HOUSEHOLD FLOW REDUCTION DEVICES Single family homes #1 #2 #3 #4 #5 #6 #7 #8 Shallow trap toilets 1 1 1 1 1 1 Econo- Plush __ 2 2 Sink- Bob _._ 2 2 Shower 9-^ *w i i i heads 13.3 1pm 1 3 1 2 1 Toilet . flushing 1 1 1 Lawn snplnklinz __ 1 1 __ 12 ------- HOME NO. HASE DATE TIME MCCER READIftflaR GALLONS ' WATEF CLOSE BATH LAUNDRY 1 T COLD HOT COLD HOT INLET NUMBER OF OCCUPANTS UNUSUAL CONDITIONS * COMMENTS: Figure 3. Typical data collection form ------- The control period, during which normal water use patterns were monitored, was broken down into two six-month periods (Phases I and III) covering the beginning and end of the two-year test program. Water consumption data for the con- trol period are summarized in Tables 4, 5, and 6. Average values for each household as vjell as overall averages for the entire group are shown for each function monitored. The data are'best presented on a per capita basis due to changes in the number of occupants in all but two of the test homes. Atypical data resulting from leaks, unusual circumstances, etc., were not included in the data summar- ies. Table 4. PHASE I - WATER CONSUMPTION DATA (5/ri - Home- owner #1 #2 #3 #4 #5 #6 #7 IB Overall Average Ho. or occu- pants 5.0 3.0 4.0 3.5 2.7 10.5 5.0 4.5 Inlet locd I6la 356a 166 252 260 177 133 200 213 Flush locd 36.8 105 44.0 67.5 111 42.5 34.4 72.0 64.0 toilet gihlet 22.8 29.4 26.5 26,7 42.5 24.0 25,9 36.0 29.2 Bath/Shower Ipcd 11.8 29.9 8.71 20.4 46.3 26.9 is. 5 22.4 22.7 $inlet 7.29 8.40 5.30 8.10 17.8 15.2 11.7 11.2 10.6 Laundrv Ipcd 79.3 35.6 34.0 30.3 44.6 $inlet 49.2 10.0 25.7 15.2 25,0 Lawn water ID ^ SSlnlet 1340 1000 -- 1170 62 48 - - - - - - 55 .4 .4 - - - - - - .4 a Inlet values shown = Total inlet - lawn water The per capita water use data show considerable variation among the eight homes selected. Some of the more recent dafa publlshed5*o correlated water consumption with such parameters as property valuation, education, income, occu- pation, etc. In the current program, the relatively small sample size precludes a meaningful evaluation of such rela- tionships. The variability in the data, from one home to another, is largely a function of the difference in family habits and "life styles." Another significant factor appears to be the type of waste disposal used. The aver- age total water consumption for the sewered homes was ------- Table 5. PHASE III - WATER CONSUMPTION DATA (11/72 - 4/73) Home- owner #1 #2 #3 #4 #5 #6 #7 #8 No. of occu- pants 8.2 2.0 4.0 4.5 2.3 9.0 5.0 3.6 Inlet Ipcd 121a 432a 144 193 267 224 153 153 Plush Ipcd i 31.0 103 36.4 76.6 116 50.9 50.9 66.8 toilet frLnlet 25.6 23.8 25.3 39.6 43.4 22.7 33.1 43.4 Bath/Shower Ipcd 17.4 46.2 14.0 13.3 33.7 32.2 35.6 18.9 S&lnlet 14.4 10.7 9.70 6.86 12.6 14.4 23.2 12.3 Laundry Ipcd 54.4 33.3 37.8 28.8 22.0 JSlnlet 45 7 - - - 16 18 14 .0 .7 - - - .9 .8 .3 Overall Average 210 66.4 32.1 26.4 13.0 35.2 20 .5 Lawn water Ipd S&lnlet 53.0 227 140 5.0 20.8 12.9 a Inlet values shown = Total Inlet - lawn water Table 6. PHASES I & III - WATER CONSUMPTION DATA Home- owner #1 #2 #3 #4 #5 #6 #7 #8 Overall Average No. of occu- pants 6.3 2.6 4.0 4.0 2.5 10.0 5.0 4.1 Inlet Ipcd I45a 384a 157 220 263 190 143 179 210 Plush Ipcd 34.4 104 41.0 72.5 113 44.8 42.1 69.8 65.2 toilet fSinlet 23.8 27.1 26.0 32.8 43.0 23.6 29.5 38.9 30.6 Bath/Shower Ipcd ? 14.0 36.0 10.8 16.6 40.2 28.4 24.9 20.8 23.8 Sinlet 9.68 9.37 6.82 7 -.51 15.2 15.0 17.5 11.6 11.6 L^undrv Ipcd 69.0 34.8 37.8 31.4 26.5 39.8 S&inlet 47 9 20 22 11 19 .6 .10 .0 .1 .6 .1 Lawn Water Ipd" 811 714 763 5&lnlet 47.0 41.6 - 44.3 alnlet values shown = Total Inlet - lawn water 15 ------- 253 Ipcd, as compared with 167 Ipcd for those connected to septic tanks. Although two of the eight test homes had non-metered water supplies prior to the test program, this was not found to be a meaningful parameter in this study in terms of correlating total water consumption. A comparison of the test data with the results of a previ- ous survey by General Dynamics reported in Reference 1 is shown in Table 7. The total (inlet) water consumption of 210 Ipcd is only 13$ less than the published figure and within the range of values commonly cited in the literature, The average water closet and bath consumption values, how- ever, are considerably less than the results of the previ- ous survey. Whereas Reference 1 reports toilet and bath usage as representing 39.2$ and 31.4$, respectively, of total water usage, the test data for phases I & III show respective levels of 30.6$ and 11.6$. On the other hand, laundry, and kitchen and lavatory usage are significantly higher than anticipated. The discrepancies noted probably reflect the different basis on which the data were collec- ted. While the test data were based on suburban/rural single-family dwellings, the previous survey was probably more reflective of multiple-family dwellings within an urban setting. Table 7. AVERAGE HOUSEHOLD WATER USE, TEST DATA (PHASES I AND III) No. of occu- pants Test data 4.8 Inlet Plush Ipcd Ip'c'd a 210 65.1 toilet %lnlet 30.6 Kitchen & Bath/Shower Laundry lavatory ipca 5*»inJ.et Ipcd ^.Inlet Ipcd 5&lnlet 23.8 11.6 39.8 19.1 68.3 29.4 Pre- vious survey 4.0 242 95.0 39.2 75.9 31.4 33.1 13.8 37.8 15.7 aOutslde uses included. In order to facilitate subsequent statistical evaluation of the data in terms of each' of the flow reduction devices tested, an additional tabulation was prepared (Table 8). The actual time periods covered in each home for Phases I and III are shown along with the mean per capita usage, and the standard deviations for the entire control period. 16 ------- Table 8. STATISTICAL SUMMARY OF NORMAL WATER USE PATTERNS (For uses later modified by water saving devices) A. Flush toilet,shallow trap, homes .Phase !_ Phase III Phases I & III House- hold #1 #2 #3 #4 #5 #6 Time Period 5A4-12/22 5/12-12/22 4/30-11/T 4/27-10/3 4/26-10/22 5/5-1A3 No. of days Ipcd 180 214 191 138 165 223 36.8 108 44.0 67.5 111 42.5 B. Sink-Bob homes Phase I House- hold #2 #8 Time Period 5 A2 -12/22 4/23-11/12 No. of days Ipcd 214 202 74.6 72.0 C. Econo-Plush homes Phase I House- hold #2 #7 Time Period 5/12-12/22 4/28-10/29 D. Plow limiting House- hold #1 #2 #3 #4 #5 #6 #7 #8 Time Period 5A4-12/22 __ 4/30-11/7 4/27-10/3 4/26-10/22 5/5-1A3 4/28-10/29 4/23-11A2 No. of days Ipcd 214 184 shower No. of days 180 191 138 165 223 184 202 74.6 34.4 Time Period No. Of No. Of days Ipcd days locd 12/27-5/2 28 12/27-5/2 126 12/24-4/28 119 11/24-5/6 163 11/22 -5 A 156 2/5-5/2 86 37.1 208 95.5 340 36.4 310 76.5 301 116 321 50.9 309 Phase III time Period sa 36.8 5.3 103 20.8 41.4 6.9 72.0 12.0 113 15.9 45.5 7.2 Phases I & No. of wo. of days locd days Ipcd 11/16-4/29 160 214 66.8 362 Phase III Time Period 11/24-4/29 wo. or days Ipcd 163 - 50 - .8 74.6 69.5 Phases I No. of days 214 347 & Ipcd 74 42 .6 .1 III s 14.2 9.2 III S 14.2 13.1 head homes locd 11.8 8.75 20.4 46.3 26.8 15.5 22.4 Time Period 12/27-5/2 12/27-5/2 12/24-4/28 11/24-5/6 11/22-5/11 2/5-5/2 11/16-4/29 11 A6 -4/29 No. of days 28 126 119 163 156 86 163 160 locd 20 46 14 13 33 32 35 18 .0 .3 .0 .3 .7 .2 .6 .9 No. of days 208 126 310 301 321 309 347 362 Ipcd 14. 46. 11. 16. 40. 28. 25. 20. 6 3 0 2 2 6 4 9 S 5.7 5.1 3.6 4.6 8.0 4.4 6.5 7.6 S = Standard deviation calculated from monthly averages. 17 ------- The standard deviations for toilet consumption were com- puted on the basis of weekly average values. Bath water usage exhibited significantly greater irregularities, on a weekly basis. In order to obtain a more meaningful statis- tical analysis, monthly averages for bath usage were used for computation of the standard deviation in each household. The per capita usage data were obtained from Tables 4, 5, and 6, with some modifications required for homes #1 and #2 in order to obtain a meaningful comparison with Phase II data. In home #1, the number of occupants changed abruptly from five to nine early in Phase III. The change was regarded as invalidating the family "identity" in terms of water use habits. Hence, only the data collected up to the time the change occurred (180 days + 28 days) were utilized. In home #2, a similar modification was made due to a decrease in the number of occupants from three to two which affected separately monitored toilets #2 and #3 in which the Sink- Bob and Econo-Flush devices were installed. Toilet #1 (shallow-trap toilet), used primarily by the two remaining occupants, was essentially unaffected by the change. Also, a radical change in bathing habits reported by homeowner #2 at the start of Phase II necessitated deletion of Phase I bath data for comparison purposes. Significant long-term shifts in water use patterns, from Phase I through Phase III, are evident from a comparison of per capita bath water consumption. For toilet usage, relatively minor changes were the rule. In addition, bath water usage showed greater variability than water closet consumption based on relative deviations from the mean. Hot and cold water service to the bath and clothes washing machine (laundry) was separately monitored and recorded throughout the test program. The data are presented in Table 9 for all three phases of the program, and will serve as the basis for an analysis of the savings in hot water usage (and associated fuel costs) effected by the flow lim- iting shower heads. Bath and laundry hot/cold water ratios are also tabulated for each household. A summary of all the data indicates that hot water represents 63^ of bathing water consumption and only 43$ of laundry water usage. 18 ------- Table 9. HOT WATER VS. COLD WATER USAGE FOR BATH AND LAUNDRY, LPCD Phase I Phase II Phase III House- Bath hold Hot Cold #1 #2 #3 #4 8.3 14.6 4.5 10.2 3.4 19.3 3.4 10.2 Laundry Hot Cold 24.2 9.8 54.9 25.7 Bath Laundry Hot Cold Hot Cold 12.5 13.2 4.9 8.7 3.8 26.5 22.3 9.5 4.2 -- 4.18 48.1 26.5 Bath Hot Cold 15.5 18.2 7.6 8.7 4.5 27.3 6.4 4.6 Laundry Hot Cold 28.0 11.7 26.1 21.6 #5 23.1 23.1 -- 18.2 15.5 ~ 23.8 9.8 -- #6 18.6 13.6 16.7 6.8 27.6 25.4 23.5 8.4 17-4 21.6 #7 7.6 7.6 19.3 14.8 13.6 12.5 17.8 . 14.4 23.1 14.0 15.9 12.1 #8 15.2 7.2' 11.7 18.6 18.9 8.3. 14.0 22.8 13.3 5.7 8.3 13.2 Overall Average Ratios 1.35 0.69 1.77 0.77 2.03 0.87 19 ------- SECTION V BATHROOM FLOW REDUCTION DEVICES: SELECTION, INSTALLATION AND PERFORMANCE TOILET FLUSHING REQUIREMENTS According to toilet consumption data reported in Reference 1, flushing accounts for a major fraction of the total household water requirements, ranging from 30$ to 40$. How- ever, despite the recent emphasis on water and waste manage- ment in this country, cleaning action and appearance are considered at least as important as water consumption. The predominant type of flushing action when a tank is used is the siphon jet type. When flushing occurs, a jet of water is activated under the water level creating a siphon action which cleans out the bowl. The average consumption for each flushing action is approximately 20 liters (5»3 gal.), a substantial portion of which is involved in the bowl clean- ing function. Another common toilet design is the reverse trap type. The flushing action and general appearance of the reverse trap bowl is similar to the siphon jet. The water surface and size of trapway are smaller, however, and the depth of seal is less. Two distinct approaches have been implemented in this pro- gram in an attempt to reduce toilet flushing requirements in a manner consistent with homeowner acceptability. The approaches, described below, are: (a) shallow trap toilets and (b) dual cycle flush devices. SHALLOW TRAP TOILET One of the approaches involved the use of a water saving toilet designed to use approximately one-third less water than ordinary toilets. The specific model selected for testing was the American Standard Water Saving Elongated Cadet, shown in Figure 4. It is similar in appearance and cost to the standard model except for a noticeably smaller tank. Less water is required for flushing due to the special design of the bowl (shallower trap). The Crane Company manufacturers a toilet similar in design and flush volume called the Radcliffe Water Miser. This toilet was not tested due to availability problems rela- tive to our program schedule. 20 ------- Figure 4. Shallow trap toilet 21 ------- The original plan called for the incorporation, inside the toilet tank, of one of the dual flush devices described be- low. However, preliminary testing indicated that this was not possible because (a) the shallow trap toilet would not operate properly with a reduced flush (less than 11.5 liters), and (b) the unique design of the toilet flush valve pre- cluded insertion of either dual flush device. Shallow trap toilets were installed in six of the eight test homes, four in southeastern Connecticut and two in San Diego. The water saving toilets had the standard 30.4 cm (12 ) roughing-in dimension and no problems were encountered by the plumbing contractors in their installation. In each case, the float arm had to be bent downwards substantially in order to keep the water level at the prescribed mark. The maximum inflow rate to the tank was also adjusted in order to provide an adequate flush with minimum flush vol- umes. The entire installation, including adjustments, averaged about 30 minutes. Plush volumes actually used for each of the six shallow trap toilets are presented in Table 10 and compared with the flush volumes of the original, conventional toilets. Init- ial flush volumes for the water saving toilets ranged from 9.5 to 12.5 liters, representing an average flush volume reduction of 24$. Two of the toilets (homes #5 and #6) required follow-up adjustments due to double flushing prob- lems noted by the homeowners. Several field adjustments had to be made at home #5. Recurrence of the flushing problem was finally traced to a defective flush valve (metering type) which was closing prematurely. Repair of the flush valve corrected the flushing problem. The double flushing reported in home #6 was found to be related to the wash water recycle system installed in that home for toilet flushing. This was verified by temporarily switch- ing to the city water supply, during which time the shallow trap toilet performed satisfactorily. After switching from a cartridge to a diatomite filter system, and doubling the capacity of the pressure tank, the toilet was reconnected to the recycle system and satisfactory performance was ob- tained for the remainder of the test period. DUAL FLUSH TOILET DEVICES The second approach utilized devices which converted a con- ventional water closet to dual cycle operation, i.e., a short flush for liquids and a normal flush for solids. Three different devices were examined during the program, 22 ------- Table 10. COMPARISON OP WATER CLOSET PLUSH VOLUMES Home 1 2 3 4 5 6 7 Number of baths 1 3 1 1 1 1 2 Conventional toilet f lush vol., liters 17.2 14.8A6.3/ 16.7 12.5 16.3/20.8 14.4 17.9 13.6A7.0 Shallow Trap Sink-Bob toilet flush flush vol., vol., liters liters 11.4 12.1 9.5 12.1 12.5-14.4 11.4-12.9 Light 8.0/ 8.7 Normal 16.7/ 17.0 __ EC ono -Flush flush vol., liters Light 14. V 14.0 10. 6/ Normal 17. 1/ 17.8 14.0/ 14.0 17.1 8 3 -4 15.5A5.5 7.6/ 8.0 16.7/ 15.5 the Econo-Plush, the Sink-Bob, and the Saveit. Although similar in effect, all the devices differed substantially in design, and are described below. Water consumptions with the Sink-Bob and Econo-Plush were shown previously in Table 10. Econo-Plush This toilet device consists of two interconnected plastic tanks open at the bottom which are positioned inside the toilet tank, and a handleAever assembly incorporating a unique valve arrangement. A picture of the assembled de- vice is shown in Figure 5. With the exception of some of the newer toilet models (American Standard, Sears), which have special flush valves, most standard models with stan- dard flush valves will accommodate this particular device. The Econo-Plush operates in the following manner: (a) Light flush - This is activated by pushing the handle up. The handle assembly, through a unique linkage arrangement, simultaneously opens the toilet flush valve and closes a plastic valve which seals both plastic tanks from the atmosphere. The contents of both tanks (approximately one gallon) are thereby trapped by the vacuum created and a reduced flush results. 23 ------- Figure 5. Econo-Flush toilet device ------- (b) Normal flush - This is activated by pushing the handle down in the usual manner. The plastic valve now opens in conjunction with the toilet flush valve, breaks the vacuum seal and thereby allows a full flush to occur. A label is included for posting on or near the toilet in order to remind the household occupant of the new flushing procedure. Sink-Bob As shown in Figure 6, this dual flush device consists of a polystyrene float and lead sinker connected to the float stem by a split brass ring. As with the Econo-Plush de- vice, most standard toilet models will accommodate the Sink-Bob. The Sink-Bob attaches to both rod and flapper- type seals at a point just above the flush valve. The device operates in the following manner: (a) Light flush - The toilet handle is tripped in the normal manner, opening the flush valve and allow- ing the water in the closet tank to drain into the bowl. When the level inside the tank has de- creased by approximately 50$, the Sink-Bob attains sufficient negative buoyancy to prematurely seat the flush valve. (b) Normal flush - For full flush, the handle must be held down during the entire flushing operation to prevent premature closing of the flush valve. Saveit Water Saver The Saveit water saver, unlike the other devices described above, will not convert a toilet to dual-cycle operation without modification. It does provide a reduced flush of approximately 50$ in a manner similar to that of the Sink- Bob. The device consists of a pre-folded plastic sheet which is formed around the flush valve and secured with two anchor rods (see Figure 7). When flushing occurs, the flush valve closes prematurely as approximately one-half of the water in the tank is blocked from gaining access to the drain. 25 ------- Figure 6. Sink-Bob toilet device Home Installation Dual-flush toilet devices were installed in homes #2, #7, and #8. Pour Econo-Flush units were installed; two in home #7 for one year, and two in home #2 for a six-month period. In home #2, installation was performed by the homeowner in about 15 to 20 minutes. A slight increase in normal flush volumes was observed due to slight adjust- ments to the float rod necessitated by the presence of the plastic tank inserts. In home #7, due to a special toilet flush valve arrangement, the new trip lever assembly had to be modified accordingly (see Figure 8). In addition, some minor alterations were required in the float rod orientation. In both homes, the dual-flush modification provided a light flush volume nearly 3.8 liters less than the normal flush volume. 26 ------- Figure 7. Saveit toilet device 27 ------- ro oo Figure 8. Econo-Flush toilet device: installed ------- In order to correct occasional binding of the handle which occurred after light flush tripping of one of the toilets in home #7, the handle assembly was removed and lubricated with Teflon spray. Subsequent flushing problems in home #7 were related to intermittent clogging of the inlet valves with recycled water solids, and traced to the flow restric- tive design of the downcomer tube. Replacement of both in- let valves with valves of newer, more modern design cor- rected the problem. Pour Sink-Bob units were installed, two in home #8 for one year, and two in home #2 for a six-month period. Installa- tions of all units were accomplished by the homeowners in approximately 3 to 5 minutes, without any modifications required. In home #2 the Sink-Bobs were attached to rod- type flush valvesj in home #8 the devices were attached to flapper valves. In the light flush mode, flush volumes were reduced by about 50$ The Saveit toilet device was installed in two homes not connected with the test program for a preliminary qualita- tive evaluation. The installation was readily performed by the homeowner in about 5 minutes. As noted above, the Saveit device is not a dual-flush device, and provides a reduced flush only. In one of the toilets, the device did not provide an adequate flush for solids. The Saveit was subsequently modified for dual flush operation by removing a small semi-circular section at the base of the unit, and operated satisfactorily thereafter. The light flush was achieved by depressing and releasing the toilet handle in the normal manner. For a full flush it was necessary to keep the handle down until the toilet tank had been com- pletely emptied. For the last two months Of Phase II, the modified Saveit was installed in home #7 (see Figure 9) for a limited qualitative comparison with the Econo-Flush device. The modified device provided a flush reduction of about 4.8 liters (similar to the Econo-Flush). Because of the limited use of the Saveit, results with this device are not Included in the summary tables elsewhere in this report. FLOW LIMITING SHOWER HEADS Published surveys of water usage (references 1 and 5) indi- cate that approximately 30$ of the total household water consumption is used for bathing. It is interesting to note the disparity between this value and the overall average usage recorded during Phases I and III of approximately 12$. This points to the possibility that the group of eight families selected are atypical with respect to bath- ing patterns (lower than average bathing frequency). They 29 ------- u> o :. - Figure 9. Saveit toilet device: installed ------- also seemed to have a greater preference for tub baths than showers. Shower heads with built-in flow limiting orifices are avail- able which can reduce water consumption rates from the typical 19 to 38 1pm (5 to 10 gpm) to 9.5 or 13.3 1pm (2.5 to 3.5 gpm). The actual amount of water saved will depend primarily on the system water pressure and the personal habits of the bather. Two different Speakman flow limiting shower heads were selected for testing. The first of these, shown in Figure 10, is equipped with a 13.3 1pm integral "Auto-flo" flow limiting orifice. This shower head has a fully^adjustable spray, integral ball Joint and a 5 cm face. The second shower head, shown in Figure 11, is equipped with a 9.5 1pm integral "Auto-flo" limiting orifice. It is also of the adjustable spray, ball Joint type but has a much narrower shape. Both shower heads have standard 1.27 cm (l/2M) I.P.S. female inlets which are compatible with stan- dard shower arms. A total of eleven flow limiting shower heads were installed in the eight test homes, with the specific distribution shown previously in Table 2. Install- ation can be performed by a homeowner in about 5 minutes. Figure 10. Speakman Auto-flo shower head: 13.3 1pm 31 ------- i':"; m* Figure 11. Speakman Auto-flow shower head: 9.5 1pm WATER METERS Pre-calibrated, domestic-type water meters were used for monitoring total water consumption as well as water used for toilet flushing, bathing, laundering and lawn sprink- ling as indicated before in Figures 1 and 2. Neptune Tri- Seal split-case meters, 1.59 cm (5/8") size, were selected for all cold water lines, and Neptune Trident Type S meters were used for all hot water lines (bath and laundry). Both- meters provide very close to 100$ accuracy from 0.9^ 1pm to 75 Ipro and beyond, and produce a negligible pres- sure drop over the range of Interest for all fixtures mon- itored. The meter registers were calibrated in gallons, and could be read to the nearest tenth of a gallon. TEST PERIOD WATER CONSUMPTION DATA Daily average water use values are presented in Table 11 on a per capita basis for all eight test homes. The data ------- cover roughly a one-year period froip November 1971 to November 1972. In Table 12, the data have been recast in a more convenient form for evaluation of water savings and statistical analyses. The actual time periods covered are shown along with the mean per capita usage and correspond- ing standard deviations. As noted in Section IV, the standard deviation for water closet consumption and bath usage were computed on the basis of weekly averages and monthly averages, respectively. Table 11. TEST PERIOD WATER CONSUMPTION DATA (PHASE II) Hdtne- owner 1 2 3 4 5 6 7 8 No. of occu- pants 5.2 2.5 4.1 3.4 2.4 9.9 5.0 4.1 Overall Average Inlet Ipcd 136a 300a 149 172 260 195 99.8 117 178 Flush toilet Ipcd 20.4 82.3 35.6 50.0 103 27.6 39.4 47.5 52.0 S&inlet 15.0 27.4 23.9 29.1 43.2 14.2 39.5 40.5 29.1 Bath/Shower Ipcd J&lnlet 16.1 35.9 9.1 14.4 32.2 23.8 25.8 27.6 23.1 11.8 12.0 6.1 8.4 12.4 12.3 25.9 23.7 14.1 Laundry Ipcd 74.8 36.0 53.0 32.2 36.8 46.6 JSlnlet 54.7 12.0 27.2 32.3 31.5 31.5 Lawn t{ Ipcd 1520 497 1010 later J&lnlet 68.0 39.8 53.9 a Inlet values shown » Total inlet - lawn water. In home #2, Econo-Flush dual-flush devices were tested for a period of eight months in water closet #2 and #3. The Econo-Flush devices were then replaced by two Sink-Bobs for a comparative evaluation. However, as indicated in Table 12, only 31 days of testing were completed. This was because the primary user of water closet #2 and #3 vacated the home for the remainder of Phase II. ------- B. C. D. Table 12. STATISTICAL SUMMARY OP WATER USE PATTERNS DURING TEST PERIOD (PHASE II) Household Time period No. of days Shallow Trap Toilet 1 2 3 4 5 6 Sink -Bob 2 8 1/5 - 12/20 1/6 - 12/27 11/13 - 11/25 10/29 - 10/27 11/7 - 11/19 1A3 - 12/4 9/27 - 11/1 10/4 - 11/12 328 309 378 356 342 290 31 363 Ipcd 20.6 76.3 35.6 50.0 113 27.6 55.8 47.4 Q sa 3.1^ 10.1 4.55 7.97 17-4 6.07 12.0 8.8 EC ono -Plush 2 7 1/6 - 9A3 10/29 - 11/22 215 387 49.8 39.4 13.2 6.1 Plow Limiting Shower Heads 1 2 3 4 5 6 7 8 1/5 - 12/20 1/6 - 12/27 11/13 - 11/25 10/29 - 10/27 11/7 - 11/19 1/13 - 12/4 10/29 - 11/22 10/4 - 11/12 328 309 378 356 342 321 387 363 16.1 35.8 10.1 14.6 32.2 23.8 25.8 27.6 3.0 6.1 2.0 3.1 6.9 3.7 5.9 10.3 aS = Standard deviation ------- Statistical Analysis Prior to a comparison of test and control period data, a statistical analysis was performed to determine whether or not the differences between the mean values being compared (water savings) were statistically significant. For this analysis, the student's_t-distribution was used to test the null hypothesis: X., - X2 = 0. The t-distribution is a measure of the deviation between the means of two random groups of data. The specific values of t were calculated from the following equation': t = Xl - X2 (n, S nl +n2 1 where: X, = mean usage, control period &2 = mean usage, test period S, = standard deviation, control period S2 = standard deviation, test period n- = number of data points, control period n2 = number of data points, test period The calculated values of t were compared with the theoreti- cal values of t at P = 0.05 (95$ confidence level) in Table 13 at the corresponding number of degrees of freedom (n, + ru - l). The decision to reject the null hypothesis (equivalent to a confirmation of statistical significance at the 95$ confidence level) was made when the calculated value of t was found to be greater than the theoretical value. For the shallow trap toilets, the null hypothesis was re- jected in five of the six homes. In home #5* the means were statistically indistinguishable as the average values were, in fact, identical. As for the dual-flush devices, both sets of data for the Sink-Bobs received statistical confirmation, while only one set of Econo-Flush data was found to be statistically significant. In the other set (home #7) the control period standard deviation is seen to be excessively large relative to the rather small differ- ence in means observed. 35 ------- Table 13. STATISTICAL SIGNIFICANCE OF INDIVIDUAL WATER SAVINGS » P .p H £ J ^ CO 1 3 "o to ,0 §§ o c O J- W * g « Ital W H 'C P (G a x H h 8 i l-l X Control period data (Phases I & III) House- hold 1 2 3 4 5 6 2 8 2 7 1 2 3 4 5 6 7 8 Xl» Ipcd 36.8 L03 41.4 72.0 L13 45.5 74.6 69.5 74.6 42.1 14.6 46.3 11.0 16.2 40.2 28.6 25.4 20.9 Standard deviation 5.3 20.8 6.9 12.0 15.9 7.2 14.2 9.2 14.2 13.1 5.7 5.1 3.6 4.6 8.0 4.4 6.5 7.6 Nl 31 45 41 30 39 35 27 37 27 37 9 6 12 11 12 12 12 12 Test period data (Phase H) V Ipcd 20.6 76.3 35.6 50.0 113 27.6 55.8 47.4 49.8 39.4 16.1 35.8 10.1 14.6 32.2 23.8 25.8 27.6 Standard deviation 3.1 10.1 4.5 7.9 17.4 6.0 12.0 8.8 13.2 6.1 3.0 6.1 2.0 3.1 6.9 3.7 5.9 10.3 N2 47 44 25 40 48 36 4 28 13 26 12 12 12 12 12 11 12 12 Student's t - test t _ calc. 16.7 7.5 3.7 9-1 0.0 11.2 2.43 9.6 5.1 0.96 0.74 3.4 0.73 0.94 2.51 2.69 0.15 1.74 t at P - 0.05 1.99 1.98 2.00 1.99 1.99 1.99 2.04 2.00 2.02 2.00 2.12 2.46 2.39 2.40 2.39 2.40 2.39 2.39 w tf *tir f\ H * 9S X mm X C3 1} V 12 Reject Reject Reject Reject Accept Reject Reject Reject Reject Accept Accept Reject Accept Accept Reject Reject Accept Accept GO ON NOTE: The decision to reject is equivalent to a confirmation of statistical significance. ------- In only three of the eight homes using flow limiting shower heads (#2, #5, and #6) were the data found to be statistically significant. In those cases for which the null hypothesis was accepted, any observed differences in average water use (positive or negative) will be set equal to zero. WATER SAVINGS Bathroom Plow Reduction Devices Conventional toilets were replaced by shallow trap toilets in six of the eight test homes. The water saving toilets were advertised as consuming approximately one-third less water for flushing then conventional toilets. Actual measured flush volume reductions (as shown previously in Table 10) averaged 25$, ranging from 1% to 35$. Table 14 shows the actual water savings in each of the six homes. In general, the percentage reduction in water usage correlated fairly well with the measured flush volume re- ductions, ranging from 0$ to 44$, with an average reduction of 25.6$. This indicates that the toilets operated depend- ably. Water savings ranged from 0 to 26.7 Ipcd, with an average savings of 14.8 Ipcd. The percentage reduction in total water usage averaged 6.9$, with the maximum value of 11.2$ attained in home #1. Table 14. WATER SAVINGS OBTAINED WITH SHALLOW TRAP FLUSH TOILET Conventional toilet House- hold 1 2 3 4 5 6 No. of occupants 5.1 2.0 4.0 4.0 2.5 10.1 locd 36.8 103 41.4 72.0 113 45.5 Shallow trap toilet No. of occupants 5.2 2.0 4.1 3.4 2.4 9.9 Average savings Ipcd 20.6 76.3 35.6 50.0 113 27.6 Water savings Ipcd 16.2 26.7 5.8 22.0 0 17.9 14.8 $ reduction 44.0 26.1 14.0 30.6 0 38.8 25.6 % reduction in total water usage 11.2 7.0 3.7 10.0 0 9.4 6.9 37 ------- In three of the test homes, six toilets were modified for dual cycle operation by either the Sink-Bob or Econo-Plush toilet inserts. The Sink-Bob and Econo-Plush devices pro- vide respective reductions of approximately 50$ and 25$ respectively when operated in the light flush mode. The overall reduction in« water consumption will depend on how often the light flush is used as well as the adequacy of bowl cleaning by the light flush. Table 15 presents the water savings actually obtained over the one-year test period. The Sink-Bob devices performed well in both homes #2 and #8, resulting in an average water savings of 20.5 Ipcd, equivalent to an average reduction in toilet flushing of 28.6$. The Econo-Plush device gave better 'than anticipated results in home #2, but proved in- effective in home #7. The insignificant water savings in the latter home was partially attributable to a relaxation of toilet flushing habits (failure to use the reduced flush) due to the recycle system installation and the resulting abundance of wash water for toilet flushing. Another sig- nificant factor was the relative infrequency of light flush usage by the three children. Average savings for the Econo- Plush was 12.4 Ipcd (l6.6$). The overall average water savings for both dual flush devices was 16.5 Iped (22.6$). The percentage reduction in total water usage averaged 6.0$. Eleven flow limiting shower heads were installed in the eight test homes. As noted earlier, two different types were employed. In five of the homes, 13.3 1pm flow limit- ing shower heads were used, and 9.5 1pm shower heads were used in the remaining three. The water savings produced by these devices are shown in Table 16. As indicated in Table 13, in only three of the homes (#2, #5, and #6) were the data found to be statistically significant. In the re- maining five homes, the mean usage values (test vs. control) were found to be statistically indistinguishable. In general, the observed lack of effectiveness appears to be primarily a function of personal bathing habits. In four of the homes (#1, #3, #4, and #7), significant shifts in bathing frequency and/or habits from Phase I to Phase III were ob- served (see Table 8). In homes #3 and #4, showering accounted for only a minor fraction of all bathing activ- ity. An overall average water savings of only 2.7 Ipcd, corresponding to a percentage reduction of 7.1$ was ob- served. The percentage reduction in total water usage was only l.< 38 ------- Table 15. WATER SAVINGS OBTAINED WITH DUAL FLUSH DEVICES House- hold 2 7a 8 Convent: Toll No. of occu- pants 1.5 5.0 U.I ional ,et Ipcd 7k.6 1*2.1 69.5 Dual flush devices Sink bob No. of occu- pants 1.5 . k.l Ipcd 55.8 kj.k Econo-flush No. of occu- pants Ipcd 1.5 ^9.8 5.0 39-U Average savings Water Sav: Sink-bob Ipcd 18.8 ... 22.1 20.5 % Reduc. ofFlusl 25.2 ... 32.0 28.6 t Reduc. i of total U.9 ... 12.3 8.6 .ngs Econo-flush Ipcd 2k. 8 2.7* ... 12.U * Reduc. ofFlus! 33.2 6.Ua ... 16.6 Reduc L Of tot 6.5 1.9a 3.3 U) Overall average savings Iped 16.5 % Reduction % Reduction of Flush of total 22.6 6.0 Water savings set = 0 based on statistical analyses ------- Table 16. WATER SAVINGS OBTAINED WITH FLOW LIMITING SHOWER HEADS House- hold Ia 2 3a k& 5 6 7* 8 Conventional shower heads No. of occu- pants 5.1 2.7 ^.0 k.O 2.5 10.1 5.0 k*l Ipcd lk.6 1*6.3 11.0 16.2 kO.2 28.6 25A 20.9 Flow limiting shower heads 13.3 1pm No. of occu- pants 5.2 2.5 Z.k . 5.0 *.l Ipcd 16.1 35.8 32.2 25.8 27.6 9.5 1 No. of occu- pants *.l 3.^ 9.9 ... tan Ipcd mum^m 10.1 1^.6 23.8 _ Average savings Overall avg. savings Water savings 13.3 1pm Ipcd -1.5 10.5 8.0 -o.U -6.8 3.7 % Redu. of bathing -10.3 22.7 19.9 -1.6 -32.5 8.5 Ipcd 2.7 % Red. of total -1.0 2.8 3.1 -0.3 -3.8 1.2 9.5 Ipa Ipcd ... 0.9 1.6 l* .8 1.6 $ Redu. of bathing 7.1 % Redu. of bathing 8.2 9.9 »*»« 16.8 5.6 % Red. of total 0.6 0.7 2.5 0.8 % Redu. of total 1.0 s water savings considered statistically insignificant (set = 0) for calc. of average savings, ------- Shower hot water savings, presented in Table 17, ranged from 0'to 5.2 Ipcd (0 to 21.8$ reduction), with an average savings of 1.5 Ipcd (7.2$ reduction in shower hot water usage). Evaluation of the shower hot water savings was based on the same statistical considerations as were applied to the data in Table 16. Table 17. SHOWER HOT WATER SAVINGS House- Phases I & III hold Ipcd la 2 3a 4a 5 6 7a 8a 12.2 16.1 5.7 9.4 23.4 20.2 14.9 14.4 Phase II Ipcd 12.5 13.2 4.9 8.7 18.2 16.7 13.6 18.9 Hot water savings Ipcd -0.3 2.9 0.8 0.7 5.2 3.5 1.3 -4.5 % reduction -2.5 18.0 14.0 7.5 21.8 17.3 8.7 -31.0 Average savines 7.2 aHot water savings in these homes set equal to zero based on results of statistical analysis summarized in Table 12. ------- SECTION VI RECYCLE SYSTEMS: DESIGN, DEVELOPMENT INSTALLATION AND PERFORMANCE OVERALL SYSTEM DESIGN, INSTALLATION AND PERFORMANCE Treatment, Storage and Toilet Flushing Reuse Since toilet flushing normally accounts for the major fraction (30-40$) of household water usage, the reuse of waste wash waters (bath and laundry) for this function is a potentially attractive scheme. The actual amount of water saved will depend upon the nature of the balance between waste wash water generated and toilet usage in a particular household. If necessary, a more favorable bal- ance can be effected by (a) a reduction in toilet flushing requirements through installation of a suitable dual-flush device, or (b) an increase in the"available supply by in- clusion of lavoratory waste water. Apart from the necessity of providing an adequate water supply at all times, other desirable system criteria in- clude minimum odor, minimum staining properties, accept- able clarity and color, and prevention of health hazards. It is, of course, also necessary to achieve these criteria under the practical constraints of acceptable costs, main- tenance, and space requirements. One of the aims of this program was to develop sufficient information with which to establish an acceptable compromise in terms of perfor- mance, cost, and safety. All of the functions and components involved in the house- hold wash water recycle systems are illustrated in Figures 12 and 13. These configurations are the ones chosen for extended testing after optimization of component choice and arrangement. The prior development work and individual component performance are described in more detail in later subsections of this chapter. Both systems finally chosen are fundamentally similar with the exception of the spec- ific design of the filtration and disinfection systems. The recycle system incorporating the diatomite filter 42 ------- LAUNDRY DOLE SOLENOID VALVE TO FLUSH TOILET A 1/2" COPPER! POLYETHYLENE STORAGE TANK 1/3 HP SHALLOW WELL JET PUMP CLOROX FEEDER ASSY CARTRIDGE FILTER Figure 12. Recycle system with cartridge filter ------- BATH 1" FLEXIBLE HOSE LAUNDRY 1-1/2" COPPER OR PVDC 1/2" TO TABLET FEEDER POLYETHYLENE STORAGE TANK 1/3 HP SHALLOW WELL JET PUMP 1-1/4" BYPASS 1"CHECK VALVE ' DIATOMITE FILTER 1"CHECK 115V Zir \ VALVE 1/30HPRECIRC. PUMP PVC // BACKWASH ' VALVE DF = DISINFECTANT FEEDER Figure 13. Recycle system with diatomite filter ------- (shown in Figure 13) was somewhat more complicated because of the need for recirculation as explained below. Laundry and bath water were collected in a suitably sized vented storage tank, provided with an overflow pipe, side bottom outlet, and a low-level control system for supplemental feed water. The stored wash water was either continuously (see Figure 12) or intermittently (when the pressurization pump operated, see Figure 13) disinfected prior to filtra- tion. The treated water was pressurized by a 1/3 HP shallow well jet pump mounted on either a % or 115 liter pressure tank, controlled by a pressure switch over the range of 105 to 210 cm Hg. When the pump was activated, wash water was pulled through a cartridge or dlatomite filter and pressurized. This arrangement, with the filter upstream of the pump, had several advantages: i (l) The filtration system operated under a vacuum (0-65 cm Hg) while the pump was running, and only 4.5 to 11.5 cm Hg positive pressure at all other times. Hence, the system, up to the pump check valve, was under minimal pressure at all times, thereby minimizing leaks as well as filtration system design requirements. (2) The pump was protected by the filter from any harmful debris. A filter by-pass with an in-line strainer was also pro- vided in case of filter breakdown. In order to preclude contamination of an existing water supply, no cross-connections were allowed, and a suitable air gap was provided for the feed water inlet to the storage tank. For additional protection of the potable water system, an A.S.S.E. (American Society of Sanitary Engineers) approved double-check valve assembly (Watts No. 9 backflow preventer) was installed in the house service line. This device provides positive protection against back-syphonage and backflow if the supply pressure should ever fall below recycle system pressure at the same time that a cross-connection (pressure or gravity type) existed. Some localities may require the installation of a more sophisticated backflow prevention device such as a reduced pressure device (similar to the double-check valve assembly with an added differential pressure controlled relief valve). An additional requirement of the national plumbing code, not imposed on any of the test units in- stalled, states that all non-potable water supplies must be color coded yellow. In general, plastic piping, fittings and valves were used for those lines not under a ------- positive pressure greater than 35 cm Hg. All other lines consisted of copper tubing and bronze fittings. The acceptance of plastic piping for uses other than cold water drain lines, varies from one locality to another, and appears to be gaining in favor in many areas. The design and installation of all recycle systems were subject to prior approval by local and state authorities in both Connecticut and Rhode Island. Two local plumbing contractors were used to install the wash water recycle systems in the three test homes. The systems in homes #7 and #8 were installed by the same con- tractor (see Figures l4 and 15). The installations in homes #7 and #8 were completed in approximately 8 hours each. In home #7, about one-half of the time was spent plumbing the recycle system (cartridge filter), per se, with the remaining time (4 hours) spent on connection of the system to the house plumbing. In home #8, 5 hours was spent on recycle system (diatomite filter) interconnec- tions, with only 3 hours required for house plumbing con- nections. In home #6, approximately 10 hours were required for system interconnection. This excessive installation time reflects the plumber's lack of experience with FVC piping and fittings. Pour hours were spent on house plumb- ing connections. In general, no impairment of flush toilet operation was observed in the three test homes due to the recycling of filtered, disinfected wash water. In two of the three homes, reduced flush operation was successfully employed. Continual chlorination of the stored wash water effectively inhibited bacterial growth, and no unpleasant odors were detected in any of the bathrooms during normal operation. Stable chlorine residuals were maintained at the point of reuse, despite the presence of interfering agents (soaps, detergents). In general, the susceptibility of the more resistant enteric bacterial pathogens to free chlorineois essentially the same as that of the coliform bacteria," which were readily controlled. Viruses are more resistant to chlorine than coliform bacteria, and a negative coliform test does not necessarily exclude viruses. However, with- in a home, the possibility of viruses (or other pathogenic organisms) being transmitted by toilet flushing reuse, although possible, is less likely than by normal means (physical contact, inhalation, or ingestion). The clarity and color of the recycled water attained by the diatomite (swimming pool) filter proved to be satisfactory in terms of aesthetic acceptance by the homeowner and was 46 ------- Figure 14. Recycle system with cartridge filter: installed 47 ------- Figure 15. Recycle system with diatomite filter! installed 48 ------- achieved with relatively low operating costs and maintenance requirements. The residual suspended solids in the recycled wash water did produce temporary stains in the toilet bowls which resulted in increased cleaning requirements. Since dissolved soaps and detergents were not removed by the dia- tomaceous earth, the foaming tendency of the wash water was not significantly reduced. However, no problems due to foaming were observed at the toilet inlet valve during re- fill of the water closet tank. Lawn Watering Reuse The original program plan called for the installation of two wash water recycle systems in San Diego, California, similar to those described above, to explore the feasibil- ity of reducing lawn water consumption in a water short region. The normal water use patterns summarized previously show that lavin water consumption averaged 1340 and 1000 Ipd respectively in homes #1 and #2 during the growing season. Based on full utilization of laundry and bath wastes, lawn usage could have been reduced by 455 Ipd (34$) and 197 Ipd (20$), respectively. Maximum utilization would probably have required a significant change in lawn watering habits by both homeowners in order to use the treated water as it became available. Because of possible harmful effects of waste wash waters on the lawns, the reuse of laundry water was pre-tested during the preliminary monitoring period (Phase I). The principal concerns are with chemical effects on vegetation and soil chemical makeup. In general, soils that are acid, sandy and well-drained should provide the most favorable results. Homeowners #1 and #2^ watered a small test plot with laun- dry water on a weekly basis and no short-term (3-4 months) adverse effects were noted. Just prior to the scheduled installation, however, the San Diego Public Health Department failed to approve the ex- perimental reuse of waste water on lawns. They were opposed, in principle, to any decentralized control of waste water management, and did not wish to encourage such a trend by setting a precedent in this case. In order to accomplish the original program objectives, the reuse of wash water for lawn irrigation was subse- quently incorporated into two of the toilet flushing recycle systems in the Southeastern Connecticut area (homes #6 and #7) primarily to supplement existing toilet flushing reuse and further reduce waste flow. This was done with the full cooperation of local public health officials even after being appraised of San Diego's objections. 49 ------- In home #6, a 33 sq m test plot received average waterings of 100 Ipd for three months and 400 Ipd for another three- month period. In home #7, a 14 sq m test plot was watered an average of only 15 Ipd. Revolving sprinklers were operated by shut-off valves located near the pressure tank outlet, for convenience as well as safety. No oper- ational problems, such as clogging of the orifices, were encountered during the test period. The soil at both locations was tested and characterized as generally sandy and acidic. At home #6, where much larger amounts of wash water were reused for lawn sprinkling, the soil drainage was characterized as fair. The drainage at home #7 was generally poor. No significant effects, adverse or beneficial, were noted on lawn growth or appearance throughout the test period or during the next growing season. The sprinkled areas were not sheltered from the normal rainfall and were, in fact, exposed to record rainfalls throughout portions of the test period. Prob- lems may have developed If drought conditions prevailed. It is unfortunate that the lawn sprinkling tests could not have been conducted in San Diego to obtain more conclusive information. The effect of lawn sprinkling with filtered, disinfected wash water on various soil characteristics is shown in Table 18. The soil analyses were performed on samples taken at the end of the test period. Soil samples taken from the non-sprinkled area were assumed to be representa- tive of soil conditions prior to the period of reuse. Although detergents are formulated of sodium salts of weak acids, the pH of the sprinkled wash water was found to be essentially neutral. Excessive amounts of phosphate may tie-up iron, magnesium, calcium and other essential metals needed for plant growth. In both homes, low-phosphate detergents were used. The soil analyses indicate no significant decrease in the avail- ability of the essential metals listed in Table 18. Spray- ing the lawns with wash water apparently effected small re- ductions in soil moisture content of from 10 to 20$. The moisture reductions may reflect an increased water repell- ancy, or may result from the normal differences in soil characteristics at the locations sampled. The effect of sodium buildup on soil permeability may become significant on a long-term basis. However, the lawn sprinkling tests were conducted on such a limited basis that sodium levels were not evaluated. Additional long-term testing of the lawn watering scheme would be required for a definitive evaluation of its effect on soil moisture. 50 ------- Table 18. EFFECT OF LAWN SPRINKLING ON VARIOUS SOIL CHARACTERISTICS Location pH Calcium Potassium Phosphate % Moisture Particle size distribution sprinkled 6.2 #6 non- sprinkled 6.6 #7 sprinkled 5.3 non- sprinkled 5«^ NOTES: MH - Medium high M - Medium L - Low VL - Very low MH M MH M VL VL L VL VL VL M M 8.7 9.6 11.4 13.9 Pine - very fine sand Fine - coarse » « o M.-3 sand These level designations are relative to normal values found in the soils in this area and were assigned by the University of Connecticut Agronomy Testing Laboratory. Reuse for lawn sprinkling may present a possible hazard because of the greater accessibility to the recycled wash water, should any pathogens be present. This hazard could be eliminated by a suitable underground discharge system. Water Savings Waste wash water (bath and laundry) was reused for toilet flushing and/or lawn sprinkling in homes #6, w7, and wo. Data for each specific type of, reuse are shown in Table 19. for each of the three home's. In home #8, wash water was recycled exclusively for toilet flushing. In home #6, it became necessary to reconnect the water closet to the city water supply during the months of August and September in order to make changes in the filtration and pressuriza- tion system. During that period, wash water was reused exclusively for lawn watering through use of the filter by-pass line which incorporated a heavy duty Cuno car- tridge filter. The maximum amount of water savings attainable is equal to the water normally used for toilet flushing and/or lawn sprinkling. In actual operation the water savings will be diminished by an amount equal to the supplemental feed- 51 ------- Table 19. WASH WATER RECYCLE SYSTEM: FLOW REDUCTION -SUMMARY ui Household and Phase I & III average inlet #6 190O Ipd #7 715 Ipd #8 735 Ipd Type of re -use toilet flush. toilet & lawn lawn sprink. toilet flush. toilet & lawn toilet flush. Net storage volume liters 455 455 455 305 305 322 Toilet Bath & laundry Ipd Ipd Phases I & III 433 448 448 21O 210 286 Phase II 266 337 -__ 193 204 194 Phases I & III 662 662 662 282 282 195 Phase II 750 794 732 265 334 262 Lawn sprin- kling Ipd 106 402 15.5 Feed water Ipd 15.5 44.5 22.4 10.6 4.9 18.2 Tank over- flow Ipd 5OO 396 352 83 119 86 Overall average0 Average flow reduction1* Ipd 432 510 380a 189 220 268 (176) 305 Ipcd 43.7 51.5 38. 4a 37.8 44.0 65.4 (42.9) 44.0 % red. 23 27 20a 26 31 36° (24) 26 Not included in overall average, atypical reuse system. Note on calculation- procedure - Hones #6, 7 and 8 had reduced flow toilets and showers during Phases II but not during Phases I and III. To put the data on the same bases with no reduced flow fixtures during all phases, the following was done: - Effects- of reduced flow showers were neglected. - The flow reduction possible was assumed to be the average water used for toilet flushing during Phases I and III, plus any lawn watering in Phase II, minus the average feed water required in Phase II. The. feed water requirement could be higher but the effect is probably small unless there is insufficient bath and laundry water to provide for conventional toilet flushing. See note c below regarding home #8. °For hone #8, the bath and laundry water is not sufficient to provide for flushing conventional toilets. In addition to the procedure outlined under note b, flow reduction was calculated by another procedure. The flow reduction possible was taken as the bath and laundry water available during Phase II minus -the tank overflow. The flow reduction values resulting from this assumption are shown in parentheses. These values are thought more appropriate for hoae #8 and were used in calculating the overall averages. ------- water required due to both short-term and long-term im- balances in supply vs. demand. In homes#6 and #7, a comparison of wash water supply and water closet demand shows a substantial surplus of wash water available. In home #8, the supply deficit observed during the control period was converted into a surplus during the test period by the incorporation of dual-flush toilet devices, thereby substantially reducing water closet demand. In all three homes, the supplemental feedwater requirements averaged only 5$ of the total water reuse requirements. Average flow reductions (water savings) are also presented in Table 19. The overall average water savings for toilet flushing and/or lawn sprinkling was 305 Ipd. The average savings for toilet flushing reuse only was 266 Ipd. The incorporation of lawn sprinkling as a supplemental reuse mode in homes #6 and #7 increased the water savings by 16$ to 18$. Greater flow reductions could be achieved by con- verting lawn sprinkling from a manual to an automatic operation through the inclusion of a high-level control system similar to the low-level control system previously described. The overall average flow reduction, on a per capita basis, was 44.0 ipcd. The average percentage re- duction in total water consumption for all three recycle systems was 26$. Additional Benefits The reuse of waste wash water for toilet flushing and/or lawn sprinkling provided tangible benefits to two of the three homeowners above and beyond the savings in water consumption. The recycle system, by minimizing the surges in outflow to the septic system associated with laundry and bath discharges, as well as by reducing total waste flow, allowed the septic tank and soil absorption system to operate more effectively. In both homes #7 and #8, the septic systems performed poorly prior to recycle system installation. In home #7, the normal annual septic backup and associated odors were noticeably absent despite record rainfalls during the year. Inspec- tion of the septic system at the end of the test period indicated all lines were clear with no sign of clogging or backup. In home #8, before and after installation of the recycle system, the household experienced difficulty laundering multiple washes (in succession) without causing the septic system to back-up and overflow the standpipes. As a result, frequent trips to a commercial laundromat were necessary. This difficulty was not experienced during the test program. 53 ------- SUB-SYSTEM DEVELOPMENT AND PERFORMANCE Storage System On the basis of criteria such as corrosion resistance, adaptability and ease of handling, and cost per unit volume, Nalgene heavy-duty conventional polyethylene tanks were selected. These plastic storage tanks are also trans- lucent, thereby enabling water levels inside the tank to be monitored by the homeowner. Selection of tank capacity was limited by considerations of space and cost. Based on daily monitoring of the magnitude of both flush toilet and bath/laundry usage, as well as their relative balance or imbalance, 380 liter (100 gallon) tanks were deemed ade- quate for homes #7 and #8, and a 570 liter (150 gallon) tank was selected for home #6. Standard 455 liter galvanized pneumatic tanks were too difficult and costly to adapt, and custom-made galvanized tanks were more costly and difficult to handle than those selected. Two-hundred and eight liter (55 gallon) steel drums with polyethylene liners were eco- nomically attractive but were not considered the most con- venient for the first prototype test units. A sketch of a 380 liter polyethylene storage tank, with required modifications, is shown in Figure 16. The vented tank was provided with a 3.2 cm (1-1/4 ) side bottom out- let bulkhead fitting of PVC. A side bottom outlet was selected, as opposed to a bottom drain, to explore the benefits of using the storage system as a settling tank. 2.5 cm (l") and 3.8 cm (1-1/2") PVC bulkhead fittings were installed in the 0.95 cm LPE (linear polyethylene) cover to accommodate the laundry and bath wastes, respectively. A 3.8 cm overflow fitting was located approximately 12.5 cm below the cover. Design calculations indicated that a 9.2 cm allowance would have been sufficient to provide for a continuous inflow of 115 1pm. The tank rested on a sheet of 2.5 cm plywood and was elevated with concrete blocks to allow the overflow to gravity drain into the building main drain to the main septic tank. A check valve was incorpor- ated into the overflow line- to preclude backup of the septic system into the storage tank. The storage tank was also provided with a low-level control system (LLCS) in order to ensure an adequate water supply at all times. The LLCS consisted of a solenoid controlled feed line valve activated by a sump pump type float switch installed as shown in Figure 17. A specially designed float rod assembly was fabricated for this application (see Fig- ure 18). The LLCS was designed so that the feedwater line would be activated with the water level approximately 7.6 54 ------- 3/4" SOLENOID VALVE CONCRETE BLOCKS SUMP PUMP FLOAT SWITCH (INSTALLED UPSIDE DOWN) FLOAT ROD ASSEMBLY PVC BULKHEAD FITTINGS BASEMENT FLOOR Figure 16. Polyethylene storage tank 55 ------- Figure 17 Low-level control system cm above the top of the outlet pipe. This Is a point where sufficient water is still available to handle 3 to 4 sim- ultaneous toilet flushes. De-activation of feedwater supply occurred with the water level at 12.7 cm above the top of the outlet pipe. The effective wash water storage capacity of the tank was thus decreased by 28 to 47 liters. The size of the storage tanks proved to be adequate in terms of collection capacity required prior to reuse as evidenced by the relatively small quantities of fresh feedwater re- quired. The low-level control system performed reliably and maintained an ample supply of water in the tank at all times for toilet flushing and/or lawn sprinkling reuse. In homes #6 and #7, the tank overflow was capable of hand- ling simultaneous bath and laundry inputs with the water level remaining well below the top of the tank. In home #8, the overflow initially did not function properly due to a partially clogged septic system and improper installa- tion (insufficient slope from tank to soil pipe location). Subsequent cleaning of the septic system and increasing the pitch of the overflow line allowed for satisfactory overflow operation for the remainder of the test period. 56 ------- 1/2" BULKHEAD FITTING 1/2"PVCPIPE 'T-'20-1/2" LONG FOR 100 GAL TANK 2T LONG FOR ISO GAL TANK 3/4" X 1/2" ADAPTER. JKS 3/4" PVC PLUG. NPT W/1/4" TAPPED HOLE 1/2"PVCTEE.NPT 12" L BRASS FLOAT ROD. 1/4" -30 STD.4"X6"FLOAT Figure 18. Low-level control float rod assembly 57 ------- The side-bottom outlet permitted a substantial fraction of the suspended solids to settle out and accumulate at the bottom of the tank over a period of time. The obvious ad- vantage of using the storage container as a settling tank is the decreased load on the filter and associated reduc- tion in filter operating costs, particularly for the car- tridge filters. The accumulated sludge has to be drained or pumped out periodically in order to maintain this ad- vantage. An alternate design, more suitable for use with a diatomite filtration system, is to provide for bottom discharge, allowing virtually complete drainage and pre- cluding any significant sludge build-up. A tank with a conically-shaped bottom (commercially available in heavy- duty polyethylene) would be ideal for this design modifica- tion. A polyurethane support could be molded at nominal cost. Without sludge separation by settling, the load on the filter would be increased and backwashing and pre- coating would probably be required at least twice as often. However, the increased filtration costs should be offset by reduced disinfection requirements. Leaks developed at the tank outlet location in both homes #7 and #8. In both cases, threaded plastic (FVC) fittings were cracked by the application of excessive torque to a mating metal reducing bushing or adapter during installa- tion. The connections were subsequently made up using all FVC fittings and no further leaks were observed. Filtration System The two primary reasons for the incorporation of a filtra- tion system into the wash water recycle system are: (l) To remove debris such as large dirt particles and lint that might adversely affect operation of the pressurlzation system or toilet control valves. (2) To provide an effluent of sufficient clarity (as indicated by turbidity and suspended solids levels) to be generally acceptable to the user. The first function can be readily accomplished by a simple basket-type strainer. The second function obviously re- quires a more sophisticated and costly filter. The cri- teria of aesthetic acceptability is a highly subjective and variable one, and may be amenable to public condition- ing and education. The toilet flushing water standards suggested in Reference 1 were used as tentative guidelines ------- and included a recommended maximum turbidity level of 20 turbidity units. In the search for an optimum filtration system, the desired degree of filtration must be tempered by considerations of economic feasibility and convenience to the homeowner. In general, the finer the filtration, the more costly the op- eration. After a survey of available filter types, and discussions with various manufacturers, two alternate approaches were selected for evaluation. Cartridge filters were incorporated into the recycle systems of home #6 (8 months) and home #7 (12 months). Diatomite filters were installed in home #8 (12 months) and home #6 (3 months). The performance of each filter was continually monitored and evaluated in terms of effluent clarity (turbidity), suspended solids levels and filter capacity. In addition, the homeowner recorded, as shown in Figure 19, such data as filter pressure drop, storage tank level, and quantity of disinfectant used. Information relating to both routine maintenance and operational problems was also noted. Tur- bidity and suspended solids analyses were performed in accordance with the standard methods recommended in Refer- ence 9. Diatomite Filter The specific unit selected was a Diaclear LP-18 diatomite filter, and is shown in Figure 20 with the pagoda-shaped filter septum removed. The septum is fabricated from woven polypropylene and has a filtering surface area of 1.6? sq m. In order to facilitate backwashing, the filter was modified by incorporation of a PVC reciprocating slide valve as illustrated in Figure 21. Prior to filter opera- tion, the filter is pre-coated by circulating a pre-mixed slurry containing approximately 0.7 kg of diatomite through the filter, the slurry being educted through a separate valve controlled feed port (see Figure 21). The average time required for pre-coating the filter was approx- imately 5 minutes. The pre-coat procedure was performed in accordance with the manufacturer's recommendations and follow-up consultations, and was considered optimum for this application. Changing the position of the slide valve reverses the flow through the filter and allows the dirty filter cake to be backwashed out to the drain with the use of 75 to 95 liters of filtered wash water. Fresh water for backwashing can be fed in through the same connection as used for pre-coating. This particular filter model was selected because of the relative ease with which it can be backwashed. Sand filters were not considered because of the large flow rates (75-115 Ipm) normally required for back- 59 ------- HOME NO. /PHASE JH DATE TIME METER READINGS, GALLONS WATER CLOSET BATH LAUNDRY COLD HOT COLD HOT FEED WATER INLET FILTER AP, INCHES HG FILTER RENEWAL BACKWASH VOL. GAL TIME REQ'D, MIN. STORAGE LEVEL SAMPLE NUMBER STORAGE TANK TOILET TANK NUMBER OF OCCUPANTS O COMMENTS: Figure 19. Typical data collection form - recycle system ------- washing of systems that would be suitable In size for home use. Figure 20. Diaclear dlatomite filter 61 ------- Figure 21. Diatomite filtration system: installed 62 ------- Preliminary testing of this filter unit without the normal recirculation gave poor results. In the absence of flow, the filter cake rests at the bottom of the filter. Each time the jet pump is activated, the filter cake is re- applied to the system, but sufficient time apparently elap- ses to permit a significant portion of the wash water to pass through unfiltered. Normal recirculation rates are in the order of 40-80 Ipm/sq m filter area. In this applica- tion, recirculation levels of this magnitude would be im- practical in terms of operating costs and noise levels. Hence, a compromise solution was sought, and a 1/30 HP re- circulation pump was incorporated which did provide accep- table effluent clarity after several hours of recirculation. The recirculation pump required a by-pass line and check valve in order to preclude cavitation during pressurization of the; filtered wash water. A diatomite filter was first evaluated in home #8, with the filtered wash water reused exclusively for toilet flushing. Sears standard swimming pool grade diatomaceous earth was utilized in all but one instance. For the third filtration cycle, a much finer grade (Johns-Mansville supercel) was used for a comparative evaluation. Table 20 summarizes the filter performance data for homes #8 and #6. The volume of wash water processed and number of days is tabulated for each filtration cycle. Backwashing was performed when the filter pressure drop, as indicated by the pump vacuum gage, exceeded 64 cm Hg. The typical variation of filter pressure drop with time observed for this type of filter is shown in Figure 22. After the first few days of operation, a pla- teau of /^ 40 cm is reached and maintained until the last few days of the cycle. The dropping and re-application of the filter cake by temporary shutdown of the recirculation pump can result in a temporary lowering of the pressure drop with a concomitant extension of filter life. In home #8, the filter was backwashed prematurely during the 2nd, 3rd, 5th, and 6th cycles. The 2nd and 5th cycles were prematurely terminated to accommodate replacement of the recir.culating pump. The first recirculating pump de- veloped excessive noise which was later traced to a defec- tive s tat or winding. The second pump developed a leak at the shaft seal due to the corrosive effect of excessive chlorine residuals discussed in a later paragraph. A seal- less magnetic drive 1/30 HP pump was used, for the remainder of Phase II. No further problems were encountered. The 3rd cycle was terminated as a result of the poor perform- ance being obtained with the supercel grade of diatomaceous earth and the 6th cycle due to a penetration of the filter septum. The wire form on which the filter septum is positioned snapped at a welded Joint (which was probably 63 ------- weakened by overchlorination during initial testing) and subsequently penetrated the polypropylene bag. Some corro- sion of the interior surfaces of the filter housing was also noted in home #8. A change in materials of construc- tion for both the wire form and filter housing from galvan- ized to stainless steel or a suitable plastic is apparently indicated for this application Table 20. DIATOMITE FILTRATION - PERFORMANCE DATA HOME #8 Length of Time Period Period (days) 1 2 3 4 5 6 7 124 50a I6a 56 I6a 7a 83 Diato uradel std std Super - eel std std std std mit oat Wt. 0. 0. 0. 0. 0. 0. 0. e Volume of Turbidity levels, ppm 68 68 68 68 82 82 82 23,700 11,700 3,350 14,500 2,690 1,100 15,500 14-30 19-29 40-60 18-40 40 110 15-39 Toilet 13-30 25 20-60 25 90 15-29 Suspended solids, mg/1 Storage Toilet 12-21 15-25 23-32 29 23-45 66 36 10-38 HOME #6 Time Period 1 2 Length of Period (days) 49 57 Diatomate pre-coat aradeiwt. .K-K std std 0. 0. 82 82 Volume of wash water processed, Tit ere /eye 12,8oob 14,000C Turbidity levels. pom Suspended noli fin. mer A . Storage we/lawn Storage we/lawn 23 22 19 20 12 15 20 19 Baclcwashed prematurely for reasons other than excessive filter pressure drop (see text). bReused primarily for lawn sprinkling. GReused primarily for toilet flushing. ------- cr\ FRAM CARTRIDGE FILTER AMF-CUNO CARTRIDGE FILTER Figure 22. Filter pressure drop vs time ------- Three complete filtration cycles were obtained during Phase II. The volume of wash water processed per cycle ranged from 14,500 liters to 23,700 liters, with a corresponding cycle life of from 56 to 124 days. The observed spread in the data probably reflects the effect of sludge build-up at the bottom of the storage tank. The sludge was pumped out of the tank at the end of period #5 by simultaneous stirring and dilution with supplemental feed water. The implications for storage tank redesign were discussed in a previous paragraph. In home #6, two full filtration cycles were obtained. The filter capacity ranged from 12,800 to 14,000 liters of wash water, with corresponding cycle lives of from 49 to 57 days. The relatively short cycle life, as compared with home #8, is a consequence of the much greater daily water consumption in home #6. Turbidity and suspended solids levels were determined at both the storage tank and water closet locations. In gen- eral, the filtered wash water had a slight greyish cast as compared with tap water. At the lower turbidity levels the color was scarcely noticeable. During normal operation, the turbidity of the water recycled to the water closet ranged from 13 to 30 ppm in home #8, and 19-20 ppm in home #6. Suspended solids levels ranged from 15-25 mgA in home #8 and 19-20 mgA in home #6. The 5th and 6th filtra- tion cycles in home #8 were not included with the data summarized above due to the developing loss of Integrity of the filter septum. The variation in the turbidity and suspended solids data reflects, to some extent, the time interval between sampling and the most recent bath and/or laundry input, as well as influent turbidity and suspended solids levels. Although the effluent turbidity levels ranged somewhat higher than the proposed design criteria, they were found to be aesthetically acceptable. COD (chemical oxygen demand) levels for the filtered, dis- infected wash water were also determined and ranged from 53 to 85 mgA. Cartridge Filters As an alternate approach, two different cartridge-type filters were evaluated during the test program. Since no circulation was required, the systems incorporating these filters were inherently less complicated. One of the cartridge filters selected was an AMP/Cuno CG 4DC-1 stain- less steel filter shown disassembled in Figure 23. Four cartridges surround the internal centerpost. The model was selected primarily on the basis of ease of maintenance, 66 ------- Figure 23. AMF/CUNO cartridge filter ------- availability, corrosion resistance, and the relatively large number of cartridge types and porosities available. Micro- klean depth-type, disposable cartridges (wool fiber, phenolic resin) with nominal solids removal down to 5 to 25 microns, and Cuno-Cel surface type disposable cartridges (pleated cellulose) with a nominal solids removal down to 10 microns were chosen for testing. A Pram MCM epoxy-coated steel filter shown in Figure 24 was the other unit tested. This model was selected primarily on the basis of its low initial cost. The filter is de- signed to accept a single large cartridge of the surface type (pleated cellulose), with nominal solids removal down to 5 to 15 microns or the depth type (glass fibers), with solids removal down to 10 microns. Both the Pram and AMP/ Cuno filters were provided with vent and drain plugs. Pram Filter - A Pram MCM filter was tested in home #6, with an AMP/Cuno 1M filter installed in the by-pass line. Cartridges were replaced when the filter pressure drop exceeded 64 cm Hg. The typical variation of filter pressure drop with time is illustrated in Figure 22. The pressure drop is seen to increase very gradually and remains relatively low through- out most of the filtration cycle. Toward the end of the cycle, the rate of increase is substantially accelerated. Cartridge replacement was readily accomplished by the home- owner in approximately 10 minutes. Table 21 summarizes the filter performance data for home #6. The volume of wash water processed by the surface-type cartridges progressively diminished from 15,650 liters to 5,200 liters, with a correspondingly shorter cartridge life. This apparent diminution in filter capacity was presumably due to the gradual build-up in and partial carryover of sludge at the bottom of the storage tank. The sludge carry- over was apparently enhanced in periods 7, 8, and 9 by the supplemental reuse of wash water for lawn sprinkling. Hence, periods 1 and 5 would be indicative of filter car- tridge capacity if the sludge were periodically removed every 3 to 4 months by draining or pump-out. Periods 7, 8, and 9 reflect conditions of filtration without the aid of a settling tank. A single 10-micron depth-type cartridge was tested and became clogged after having processed only 3305 liters (10 days) of wash water. Due to a delay in the procure- ment of additional extended surface cartridges, the by- pass filter was operated for a total period of 39 days. 68 ------- 'Mi ------- Table 21. PRAM MCM CARTRIDGE FILTER/CUNO IN By.-PASS PILTER PERFORMANCE DATA Turbidity Cartridge Volume levels, ppm Time Cartridge life, processed period description days Suspended , me: A ._ efflu- llters/cyc.influent ent efflu influent ent- 1 C-744, 5 52 micron, ex- tended sur- face 2 C-729, 10 10 micron, depth type 3 Cuno by -pass, 3 10 micron depth 4 Cuno by -pass, 11 25 micron depth 5 Cuno by-pass, 25 heavy-duty 6 C-744-15, 15 46 micron, exten- ded surface 7 C-744-15, 15 22 micron exten- ded surface 8 c-744-15, 15 12 micron, exten- ded surface 9 c-744-15, 15 8 micron, exten- ded surface 15,650 3,305 855 3,100 6,200 12,000 9,450 5,150 70-90 40-55 49-67 26-36 90 50 90 68 85-90 78-85 85 82 75 55 90 60 85 60 64 68 66 52 29 60 64 29 65 32 5,220 110 65 66 33 Depth-type cartridges with nominal ratings of 10 to 100 (heavy-duty) microns were used successively, and evidenced respective filter capacities of 855 to 6200 liters (3 to 25 days). Two Cuno 1M filters equipped with heavy-duty cartridges should be equivalent to the Pram MCM filter in terms of filter cartridge life. Effluent turbidity levels for C-744 (extended surface) cartridges ranged from 40 to 65 ppm, representing turbidity reductions of from 27 to 50$. Filtrate suspended solids levels ranged from 26 to 36 mg/1, corresponding to per- centage reductions of from 44 to 50$. The heavy-duty by-pass filter achieved negligible reductions in turbidity and suspended solids levels. As was the case with diatomite filtration, the filtrate had a slightly greyish cast. 70 ------- AMF/Cuno filter - A C^no stainless steel cartridge filter received testing in home #7. Replacement of the filter cartridges was easily accomplished by the homeowner in approximately 10 minutes. Variation of filter pressure drop with time is shown in Figure 22. Filter performance data for home #7 is summarized in Table 22. The capacity of the 10-micron depth-type (F2278-C1) cartridges progressively decreased from 11,000 liters (52 days) for period #3 to 4370 liters (18 days) for period #5 due to the sludge build-up and carry-over effect noted earlier. Based on a comparison of the average weight pick- up of the filter cartridges with the observed reduction in suspended solids levels, it is estimated that approximately 2/3 of the wash water suspended solids settled to the bottom of the storage tank during normal operation. The sludge was pumped out of the tank at the end of period #7. The per- formance of the last two sets of cartridges, in terms of filter capacity, was markedly improved. In period #9, the improved G78 cartridges attained a filtration capacity of 22,450 liters (108 days), higher than any other cartridge tested. As anticipated, the 5-micron depth-type cartridges were able to process only one -half the amount filtered by comparable 10-micron cartridges. The 10-micron extended surface cartridges exhibited capacities intermediate be- tween the original and improved depth-type cartridges. Effluent turbidity levels for the 10-micron depth-type cartridges ranged from 30 to 95 units, equivalent to tur- bidity reductions of from 62 to 80$. Effluent suspended solids levels ranged from 3^ to 68 mgA, corresponding to reductions of from 63 to The extended surface cartridges produced similar average effluent turbidity and suspended solids levels. Filter Performance Comparison Each filter system tested allowed for satisfactory toilet flushing and lawn sprinkling operation. The major perfor- mance differences between the filters were in the areas of filter capacity or life, and quality of the filtrate in terms of turbidity and suspended solids. Filtrate quality is essentially an aesthetic criteria involved in the promo- tion of homeowner acceptance. Filter capacity is important primarily because of its impact upon operating costs as well as homeowner maintenance requirements. 71 ------- Table 22. AMP/CUNO CG4-DC1 CARTRIDGE FILTER PERFORMANCE DATA Cartridge Volume Time Cartridge life, processed period description davs liters /eye 1 2 3 4 5 6 7 8 9 F-2278-cl, 10 micron, depth-type F-2278-B3, 5 micron, depthr- type F-2278-cl, 10 micron, depth- type F-2278-cl ; 10 micron, depth- type F-22878-F2, 25 micron, depth- type 38285-32, 10 micron, surface type 078 -c-1, 10 micron, depth- type 38285-32, 10 micron, surface type G78 -c-1, 10 micron, depth- type 39a 26 52 35 18 30 34 49 108 4,950 7 4,750 11,000 7,600 4,370 6,450 7,400 12,600 22,450 Turbidity , levels, com i influent 220-280 150 175-280 205-300 190 165-200 170 70-130 90-160 effluent 135-140 60 40-70 40-80 95 46-60 65-95 50-60 30-65 Suspended solids, mg/1 prrxu- influent ent 92-97 38-44 110-193 36-68 168-195 49-66 135 92 95-135 33-38 60-90 39 93-158 34-51 cartridges replaced prematurely for inspection. 72 ------- Table 23 summarizes filter system performance for each of the three units tested. The diatomite filter achieved the best overall performance. The filter provided an average capacity of approximately 17,000 liters per cycle and a corresponding average backwashing interval of about three months. The average filtrate turbidity and suspended ?° l2S ifv!ls,?f ?? ppm and 21 ""SA* respectively, achieved by the diatomite filter, were judged to be quite satisfac- tory in terms of aesthetic acceptability. The two car- tridge filters, roughly equivalent in performance, pro- cessed 74 to ob> as much wash water per cycle as did the diatomite filter, while permitting effluent suspended solids and turbidity levels approximately two to three times higher. Table 23. FIHTER SYSTEM PERFORMANCE SUMMARY Average volume Filter processed, system liters /cvc . Dlaclear LP-18 Diatomite Pram MCM 15 micron surface- type AMP/Cuno CQ4-DC1 10 micron 17,000 12,600 15,000 Equivalent filtration period, days 86 48 71 Average effluent turbidity levels. Dom 23 60 62 Average effluent suspended solids. me A 21 31 43 Annual operating costs, I/year 16 43 40 depth-type Disinfection It was anticipated at the start of the program that some provision for disinfection of the stored wash water would have to be made in order to control odors as well as to prevent health hazards. In order to establish disinfec- tion requirements, several samples of non-disinfected wash water were collected one to two weeks after installation of two of the recycle systems, and were analyzed for colifortn levels. The most probable'number of total collforms (MEM) ranged from <3 to 100/100 ml for home #8 and from 160 to 1000/LOO ml for home #7. After four to five weeks holding, bacterial growth had increased by two to three orders of magnitude, and septic odors were detected. The data thus 73 ------- indicated that normally used amounts of both laundry deter- gents and bleach (Clorox) were clearly inadequate in terms of bacterial growth inhibition. Based primarily on considerations of cost and disinfecting power, chlorination was selected as the best means of des- troying pathogenic and odor-causing bacteria. The use of alkyl ammonium chlorides such as Roccal are not suitable for this application since they are neutralized by soaps and anionic detergents. A preliminary evaluation of the chlorine demand of untreated wash water samples was made for homes #7 and #8. Freshly collected wash water required chlorine concentration levels of /*- 5 mg/1 to provide 1 hour chlorine residuals 2 0.5 ppm. Untreated wash water samples collected after one to two months of storage showed higher chlorine demand levels, ranging from 10 to 15 mg/1. i Two different techniques were utilized for disinfection of the stored wash water at the required dosage levels. Both represent a relatively simple and inexpensive means of chlorination. The first of these provides for the con- tinuous introduction of diluted laundry bleach (NaOCl) with the use of an air lift feeder as illustrated in Figures 25 and 26. An alternate approach utilized an inexpensive chlorine tablet feeder designed for above ground pools (see Figure 27). Both calcium hypochlorite tablets and the more slowly dissolving chlorinated isocyanurates were eval- uated in conjunction with this feeder. Both approaches are capable of providing pre-determined chlorine dosage levels based on average daily throughput rates. Specific dosage (concentration; levels at any given time may vary considerably, and will depend on such factors as (a) daily and weekly wash water supply and water closet demand pro- files, (bj effective tank capacity, and (c) amount of dirt and sludge present in storage tank and filter at any time. The performance of each feeder was periodically monitored, and samples were taken for analyses of chlorine residual, odor, and coliform count. Corresponding chlorine dosage levels were also determined. Chlorine residuals were de- termined by the Ortho-Tolidine method.9 Testing for the coliform group was carried out by the multiple-tube fer- mentation technique, the results of which were expressed in terms of the Most Probable Number (MPN) per 100 ml.9 Operation and performance details are given below for each of the methods of disinfection. ' ------- Figure 25. Air lift Clorox feeder ------- Figure 26. Air lift feeder installation 76 ------- Figure 27. Chlorine tablet feeder ------- Air Lift Clorox Feeder An air lift feeder was installed and tested in home #7. A standard aquarium air pump forced air through flexible plastic tubing partially filled with the bleach solution, via a tee arrangement, and lifted droplets of .bleach up into the storage tank through 0.64 cm OD (0.32 cm ID) poly- ethylene tubing. The chlorine dosage rate was controlled by adjusting either the feed rate or the strength of the bleach solution (0.85 - 1.7$ NaOCl). The feed rate was governed by the level of bleach inside the main feeder bottle (in which the bubbler was immersed). The average feed rate was approximately 25 cc/hr, corresponding to chlorine dosage levels of from 20-25 mg/1. The bleach solution was main- tained at the desired level with the aid of a feeder set-up as shown in Figure 26. The Clorox reservoir consisted of an air tight (screw cap closure) plastic aspirator bottle, filled with bleach solu- tion with flexible plastic tubing attached. The tubing was inserted into the bleach solution primary feed container (Figure 25) to the desired depth. As long as the end of the tubing remained immersed in the bleach solution, no transfer of bleach took place due to the vacuum which existed in the Clorox reservoir bottle. When the level of bleach in the primary container dropped so as to expose the end of the tubing, the vacuum was broken and gravity flow took place with air bubbles displacing the fluid transferred. The flow continued until the end of the tubing was no longer exposed and the vacuum re-established. Replenishment of the diluted laundry bleach was required about every 30 days. The yearly requirement for household laundry bleach (5$ hypochloritej is estimated at 42 liters (ll gallons). The data in Table 24 show that the air lift feeder, once properly adjusted, provided satisfactory performance In terms of odor control and germicidal action. In general, chlorine residuals of from 0.15 to 0.5 mg/1 were main- tained in the water closet tank throughout the test period. As expected, chlorine residuals in the storage tank showed a wider variation, ranging from 0.2 to 5 ppm. As long as measurable chlorine residuals were maintained, no unpleas- ant odors were detected at the water closet or storage tank location, and the coliform counts were essentially negative ( £ 11/100 ml). Fortunately, the wash water was disinfected at pH values near neutral, which is favorable to a high level of disinfecting efficiency. ------- Table 24. CLOROX AIR LIFT FEEDER - PERFORMANCE DATA Average Total chlorine Chlorine Coliform dosage residual, . count, Date level, mcc/l me A Odor MPN/100 ml Remarks 11/22 11/30 12/21 3/1 3/6 3/14 3/21 3/27 3/28 4/13 4/17 4/19 4/24 4/28 5/3 5Al 5/16 5/22 6/5 6/9 6/12 6/18 6/30 0 0 0 15 10 50 20 27 27 20 10 10 28 40 20 29 10 15 10 20 25 20 35 0 0 0 0 0 0.5 0.3 0.2 0.15 0.25 0.1 0 0.25 0.5 0.15 0.15 0 0.15 0 0.25 0.2 0.15 0.1 TW₯ 160 Without disin- fection SS 1000 TWW 140,000 TW₯ Feeder in- stalled ND MC ND Incorporated "Clorox Reservoir" ND ND <11 ND ND ND ND MC < 3 ND ND < 3 ND ND MS 24,000 Feeder In- operative ND ND ND < 3 ND 79 ------- Table 24. CLOROX AIR LIFT FEEDER - PERFORMANCE DATA-(CONT'D) Average chlorine Chlorine dosage residual, Date level, rag/1 rag A 7/6 1/12. 7/17 8/1 8A6 9/5 9/8 9/15 9/22 9/28 10/3 10A6 50 25 25 15 55 60 26 20 35 25 25 50 0.15 0.15 0.2 0 0.5 0.5 0.15 0.1 0.3 0.2 0.15 0.25 Total Collform count, Odor MPNAOO ml Remarks ND ND ND MS MC < 3 MC ND ND ND ND < 3 ND ND Increased Clorox con- centration Readjusted concent. downward aAnalysis performed on samples taken at water re-use location. Odor glossary: TWW - Typical wash water MC - Mild chlorine MS - Mild septic SS - Strong septic ND - None detected (equivalent to chlorinated tap water) 80 ------- The excessively high coliform count of 24,OOOAOO ml noted in Table 24 was due to the inadvertent disconnection of electric power to the air pump. This rendered the air lift feeder inoperative for several days. Such an incident would not occur in a permanently installed system. HTH Chlorine Tablet Feeder A commercially available chlorine tablet feeder was in- stalled and tested in homes #6 and #8. The feeder was first incorporated into the diatomite filtration system of home #8, with filtered wash water continuously recirculated through the feeder. Chlorine dosage levels were controlled by (a) adjusting the immersion depth of the chlorine tablets, and (b) adjusting the rate of flow to the feeder. HTH (calcium hypochlorite) tablets were first used, but problems in con- trolling the dosage level were encountered due to its rapid dissolution rate and incomplete solubility At normal recir- culation rates, excessive chlorination resulted. At low flow rates, the tablets exhibited a tendency to coalesce in the feeder, blocking further dissolution. The HTH tablefes were subsequently replaced by completely soluble chlorinated iso-cyanurate tablets which dissolved at a much slower rate. However, the chlorinated iso-cyanurates were found to be significantly less effective than hypochlorltes, per unit weight of chlorine, in maintaining odor control, and higher residuals had to be maintained (1-2 ppm). In addition, with normal recirculation rates and present feeder design, the ex- tent of tablet immersion was critical in order to avoid under of over chlorination, making frequent adjustments necessary. A second chlorine tablet feeder was installed in home #6. The set-up incorporated a solenoid-controlled return line to the feeder. The solenoid valve was, in turn, activated by the Jet pump pressure switch, and admitted filtered, pressurized wash water to the feeder concurrently with pump operation. Hence, in this case, the chlorine feed profile coincided with the household toilet flushing pattern. Both the return flow rate and tablet immersion depth were adjusted to provide an average dosage level of 20 to 30 mgA based on the average water closet consumption. Due to the intermittent type of operation, calcium hypochlorite tablets were able to be used successfully. Chlorine odors emanating from the dis- penser were minimal. The feeder held approximately 200 HTH tablets, providing a replenishment interval of about 80 days. Approximately 3.6 Kg (8 Ibs) of HTH tablets would be required on an annual basis. Performance data for the tablet feeder installed in home #6 are shown in Table 25. During normal operation, chlorine residuals of from 0.1 to 1.0 mg/1 were maintained. On one 81 ------- Table 25. CHLORINE TABLET FEEDER - PERFORMANCE DATA Date 1/17 1/26 2/2 2/9 2/11 3/2 3/3 3/15 3/29 4Ao 4/19 4/26 5/26 5/31 6/5 6/7 6/9 6/13 7/13 Average chlorine Chlorine dosage residual, level, rag/I mg/1 0 0 0 20 10 28 200 0 25 20 10 5 25 50 25 25 200 25 25 0 0 0 0.1 0 0.5 5 0 0.15 0.1 0 0 0.15 1.0 0.15 0.15 5.0 0.3 0.1 Odor TWW TWW TWW ND ND ND MC TWW ND ND ND TWW ND ND ND ND SC ND ND Total Coliform count, MPN/ioo ml Remarks Without disin- fection « Installed floats- ing basket dispenser Incorp. manual recirc. Feeder not in service Re -installed dispenser -._ -- < 3 Installed HTH feeder -- * mm -- Pump running excessively < 3 __ 82 ------- Table 25. CHLORINE TABLET FEEDER - PERFORMANCE DATA (CONT'D) Lverage chlorine Chlorine dosage residual. Date level, ms A mgA Odor Total Coliform count, MPN/100 ml 8/1 10/5 20 10 10A2 15 10A8 20 11/19 25 0.3 0.2 0.1 0.1 0.15 ND ND ND ND ND Lawn sprink- ling only Diatomite filter oper- ational occasion, over-chlorination was produced when the jet pump remained running for an excessively long period of time due to clogging of the Fram filter. Such an occurrence can be prevented by the incorporation of a pump vacuum switch set for 65 cm Hg. The intermittent chlorination provided by this tablet feeder proved to be as successful as the continuous (Clorox) feeder in controlling odors and bacterial growth. Fluid Transfer and Pressurization A pump is required to pressurize the filtered wash water and make it available for toilet flushing and/or lawn irrigation. A pressure range of 100 to 200 cm Hg is generally considered adequate for these purposes. An additional pressure differential is required (0-50 cm Hg) to draw the water through the filtration system. In order to meet these pressure requirements at flow-rates ranging from 19 to 38 1pm, a 1/3 HP shallow well Jet pump was selected (Flint and Walling Model C833). The pump was mounted on either a 45 or 115 liter pressure tank, and was provided with an air volume control to maintain sufficient air space inside the tank, as shown in Figure 83 ------- 28. Operation of the pump was controlled by a pressure switch set to maintain the pressure between 100 and 200 cm Hg. The pump motor was rated at 115V/5.6 amps with a starting surge of 8.6 amps. The 1/3 HP jet pump provided reliable and satisfactory per- formance in terms of transferring wash water through the filter and subsequent pressurization. The pump was capable of handling filter pressure drops (vacuum) up to 65 cm Hg while pressurizing the wash water to 210 to 260 cm Hg during water closet reuse. During normal operation, flow rates of from 20 to 40 1pm were provided. The pump running time was very short, averaging about one-half minute per water closet flush. The air volume control worked satisfactorily in two of the three pumps installed. Recurrent flooding of the pump installed in home #7 was traced to defective air volume control fittings. Subsequent installation of a Schrader air valve allowed for satisfactory pump operation. In home #6, a leak developed at the pump inlet connection due to improper installation of a threaded PVC adapter. Subsequent replacement with cast iron fittings back to the check valve eliminated the leak problem. Figure 28. Pressurization system 84 ------- SECTION VII HOMEOWNER ACCEPTANCE GENERAL The continued use and success of the various water saving devices and systems installed in the eight test homes ul- timately depends upon acceptance by the homeowner. The homeowner's decision may be based primarily on socially conditioned attitudes, or on functional criteria such as performance, cost, and maintenance. In order to determine and evaluate the reactions of the people who participated in the demonstration program, a formal questionnaire was prepared and distributed to all adult occupants. Despite the relatively small sample size (a total of 16 respon- dents), the survey did provide a clear indication of the relative acceptability of each of the water Saving devices tested. A sample of the questionnaire with the tabulated results of the survey is included in the Appendix. A summary of the survey results in terms of user acceptance is presented in Table 26. The responses were generally favorable toward the use of all of the devices and systems tested. It should be noted that all of the volunteer homeowners were General Dynamics employees. The selection of General Dynamics employees was made in order to (a) help assure continuance in the program throughout its entirety, (b) maximize exchange of information, (c) facilitate monitor- ing and maintenance of experimental devices and systems, and (d) achieve better homeowner participation. ------- Table 26. SUMMARY OP QUESTIONNAIRE RESULTS Device or Number Percent of responses system of which indicated tested respondents user acceptance Plow limiting 16 88 shower heads Shallow trap 12 83 water closet Dual-flush 6 100 devices Toilet flushing 6 67 reuse lawn sprinkling 4 50 reuse Although the possibility of biased attitudes affecting the results of the survey cannot be absolutely refuted, it is believed that the respondents answered the questionnaire honestly and objectively, with any bias due primarily to pre-conditioned attitudes unrelated to their association with the contractor. The fact that the participants were selected on a voluntary basis does suggest a certain level of identification with the objectives of the program and perhaps a greater predisposition towards acceptance of a particular device than if chosen at random. FLOW LIMITING SHOWER HEADS A total of eleven flow limiting shower heads were installed in the eight test homes. Eight shower heads (5 homes) were of the type which limited the flow to 13.3 1pm (3.5 gpm). Unanimous acceptance of this particular type was indicated in terms of providing an adequate water supply, spray patt- ern, and overall performance* The remaining three shower heads (3 homes) incorporated a 9.5 1pm (2.5 gpm) flow limiting orifice. Two-thirds of the respondents using this type shower head found the device acceptable. 86 ------- SHALLOW TRAP TOILET Shallow trap water closets were installed in six of the eight test homes. Eighty-three percent of the respondents indicated that they found the toilet performance acceptable and would recommend it to others. The remaining Ytf> found that they had to double flush the water closet approxi- mately 10^ of the time, and apparently Judged this to be unsatisfactory performance. A small increase in the water closet flush volume, by either raising the water level or increasing the fill rate, may have eliminated the problem in one of the homes. All of the respondents stated that they were satisfied with the appearance of the toilet and indicated maintenance requirements were minimum. DUAL-PLUSH DEVICES A total of eight dual-flush devices were tested in three homes. All six respondents indicated that each device was easily installed and operated. All of the people surveyed indicated that the devices provided an adequate flush for solids and liquids. Two-thirds of the respondents found the reduced flush adequate for solids as well as liquids. Use of the reduced or "light" flush averaged approximately 50$ of all flushes. The necessity of having to hold the flush handle while the water closet tank was being emptied in order to effect a full flush with the Sink-bob or Saveit device apparently did not discourage any of the users. All of the adult participants considered the retail price of each device to be within reasonable limits. Costs are given in Table 27 in the Cost Analysis Section. WASH WATER RECYCLE SYSTEM Toilet Flushing Wash water recycle systems were installed in three homes in Southeastern Connecticut. All six adult respondents indi- cated that the system was manageable, simple to use, pro- vided an adequate water supply for toilet flushing, and performed satisfactorily. Fifty percent considered the' maintenance requirements minimum. All of the occupants indicated that there were additional cleaning requirements for the toilet bowl with this system due to temporary dis- coloration or staining. One of the six adult participants found the reuse of treated wash water for toilet flushing objectionable and disapproved of the appearance of the water. One-third of the participants expressed concern about odor control and for a possible health hazard. None of the respondents indicated any problems due to excessive 8? ------- foaming. Two-thirds of the adult users indicated they would recommend the use of this system to others. By way of comparison, a recent public opinion survey by Bruvold and Ward10 in the State of California showed only 5$ of the sampled population to be opposed to the reuse of reclaimed water for toilet flushing. Lawn Sprinkling All four adult participants involved in the reuse of wash water for lawn watering did not notice any obvious adverse nor beneficial effects as compared with conventional lawn watering. No effect on sprinkler system operation was noted, Both households expressed concern for a possible health hazard. One-half of the participants were willing to con- tinue using this system, as well as recommend it to others. GENERAL Fifty percent of all respondents indicated that they real- ized a cost savings during the test program as a result of a decreased water or electric bill. Seventy-five percent of the households served by a septic system experienced a noticeable decrease in service requirements for these facilities during the test period. Based on the commercial availability of recycle systems, 44$ of all respondents ex- pressed an interest in toilet flushing reuse, and 31$ in lawn watering reuse. The prices they would be willing to spend on a recycle system ranged from $200 to $500. 88 ------- SECTION VIII COST ANALYSIS INTRODUCTION Most people are interested in water pollution abatement and are willing to encourage the spending of government money for various pollution abatement programs. However, they are understandably less willing to spend their own money for private pollution control measures when their neighbors are not also compelled to do so, except in those circumstances where private water supply and/or waste treat- ment facilities are inadequate. There is also a general lack of interest in devices to reduce water usage in this country as the result of generally plentiful water supplies, rate schedules which tend to penalize water conservation, and Indirect methods for the recovery of capital costs asso- ciated with the construction of pollution control facilities, Household flow reduction is not likely to become widespread unless water users are thoroughly convinced, not only of the desirability of using water saving devices, but also convinced that use of these devices would not result in a cost penalty, and could even result in a cost savings. COST SUMMARY Bathroom Water Saving Devices The installed cost for each of the water saving devices tested is shown in Table 27. The material costs are based on the retail prices at which these devices are currently available to the homeowner. It is assumed that the toilet inserts (reduced flush devices) and flow limiting shower heads will be installed by the homeowner. The total annual cost was based on amortization* of the installed cost over *No interest was charged for determining costs in this study, the assumption being made that interest would not be a fa.ctor with the relatively small costs involved. 89 ------- the expected life of the device. The cost of toilet inserts, such as the Econo-Flush, Sink-Bob, and Saveit, are expected to decrease as the market for these kinds of devices expands. Table 27. COST SUMMARY - BATHROOM WATER SAVING DEVICES Water saving device Shallow -trap flush toilet Dual flush devices Sink-Bob Econo-Plush Saveit Material Cost-$ 60 4 14 6 Labor Cost-$ 15 0 0 0 Installed Cost-$ 75 4 14 6 Operating Cost-$ 0 0 0 0 Expected Life.yrs. 20 10 10 10 Total Annual Cost-$/yr. 3.75 0.40 1.40 0.60 Plow limiting shower heads 13.3 1pm 9.5 1pm 6 8 0 0 6 8 0 0 15 15 0.40 0.53 Wash Water Recycle Systems The cost of the various components associated with the in- stallation and operation of the two different types of re- cycle systems are shown in Table 28. For the prototype recycle systems, the installed cost of the system with the diatomite filter is seen to be $100 higher than the car- tridge filter system ($640 vs. ?54o), the difference attributable to higher costs for filtration, piping, and installation. Storage system costs were based on the use of a 380 liter polyethylene tank, and included provisions for a low-level control system and tank elevation, if re- quired. Approximately one-half of the labor costs shown are due to connection to the house plumbing and will vary somewhat from home to home. Tabulations of the operating costs associated with filtra- tion, pressurization, and disinfection are shown for the prototype systems. The rather modest filter media costs for the diatomite filter are based on the use of 3.7 kg of diatomite per year (at $0.40Ag), and replacement of the filter system every five years. The relatively high oper- ating costs for the cartridge (Fram) filter reflect an average cartridge replacement interval of approximately two months, and a cartridge replacement cost of $5.85. Filter media costs for the Cuno filter (10 micron depth cartridges) averaged $40/yr. Approximately 90$ of the electric power 90 ------- Table 28. COST SUMMARY - WASH WATER RECYCLE SYSTEMS Prototype Recycle Systems Projection for Diatomite filter A. Initial cost Storage sya» $175 Filter sys.- 135 Pressuriza- Disinfectant feeder ----- 20 Valves, pipe, fl 4- +-! VM-KCI _ _ OC Total Mat'l Cost- R40 Labor Cost 100 Total Installed Cost - $640 B. Annual Opera - t4ng cost Filter media $3.50 Electric power 12.00 Disinfectant 5.50 $21.00 C. Total annual cost Expected life Total cost per ,rr, _ $63.50 Cartridge filter $175 60 115 20 80 450 $540 $38.80 1.20 5.50 $45.50 15 $81.50 recycle system (Diatomite filter) $70 100 85 20 -Z5L 350 $400 $3.50 7.00 5.50 $16.00 15 $43.00 aFram filter selected for cost analysis 91 ------- requirements for the diatomite filtration system arise from the need for recirculation, with the remaining portion due to pressurization for toilet flushing. Calculation of the power requirements was based on a 16-hour-on/8-hour-off cycle for, the recirculation pump, and a unit electricity cost of $0.02 per Kwh. Chlorination costs for both the dia- tomite and cartridge filter systems were based on maintain- ing an average dosage level of 25 mg/1 using chlorine tab- lets (CaOCl) at a rate of 3.6 kg per year. Other methods of disinfection were about equivalent in cost. The total operating cost for the recycle system, incorporating a cartridge filter, is seen to be more than twice as high as that for the diatomite filter system due to the difference in filter operating costs of $24.50 per year. The total annual costs were determined by amortizing the initial costs over an expected life of 15 years and adding the respective operating costs. The total cost for the .cartridge type recycle system is almost 30$ higher than the diatomite system, as the relatively high filter media cost for the former system overshadows its lower initial cost. Projected costs for a mass produced, cost optimized version of the diatomite type recycle system are also included in Table 28. A substantial cost reduction is projected for the storage system based on the use of two modified 208 liter (55 gallon) drums with polyethylene liners. The use of pre- assembled subsystems should decrease installation costs by about one-half. Due to the short pump running time required for toilet pressurization, a less expensive jet pump should be satisfactory. The 42$ reduction in electric power re- quirements shown in Table 28 reflects an anticipated cor- responding reduction in filter recirculation requirements which should be attainable without compromising effluent clarity at the point of reuse. This can be implemented by the inclusion of an additional, limited range pressure switch mounted on the pressure side of the jet pump which would activate the recirculation pump. With this set-up the filter cake would be applied to the filter system, and filtered water would be circulating through the lines prior to jet pump activation. The total annual cost for the mass produced recycle system is estimated at $43 per year, cor- responding to a projected cost reduction of 32$. ECONOMIC FEASIBILITY An assessment of the economic feasibility of the various water savings devices tested during this program was made in terms of their potential for cost savings through flow reduction. In Table 29, the cost per unit volume of flow reduction effected by each device or system is presented 92 ------- Table 29. COST COMPARISON Flow reduction device Shallow trap water closet Dual flush devices Flow limiting shower heads Wash water recycle system Sinkbob Econoflush Saveit 13.3 1pm 9.5 1pm Prototype Mass-produced Cost per unit vol. of flow reduction $/1000 liters 0.15 0.02 0.07 0.04 0.08 0.22 0.57 0.39 Typical water rates $/1000 liters 0.16 - 0.42 0.16 - o.42 0.16 - 0.42 0.16 - 0.42 0.16 - 0.42 0.16 - 0.42 0.16 - o.42 0.16 - 0.42 Typical sewer rates $/1000 liters 0 - 0.13 0 - 0.13 0 - 0.13 0 - 0.13 0 - 0.13 0 - 0.13 0 - 0.13 0 - 0.13 Septic tank system - poor soil $/1000 liters ... ... <* MM o.4o o.4o Net Savings $/year $.25 to 9.80 $4.10 to 15.60 $1.72 to 9-20 $2. 40 to 10.20 $1.10 to 5.32 $.52 to 3.53 $-45.70 J to -2.301 $-130 J to 27.6O $-25.20 I to 18.20H $19.20 J to MJ.lol vo tA) savings per year based on water and sewer rates %et savings per year based on water rate and septic system cost ------- and compared with typical water and sewerage rates. For the recycle systems, the basis of comparison was expanded to include septic tanks with poor soil absorption systems. In most instances, water supply facilities are operated as a utility, and water use rates are adjusted to recover the total cost incurred (operating cost plus level debt service) plus some nominal profit. Domestic water rates throughout the State of Connecticut range from $0.16 to $0.42 per 1000 liters. Water rates in the primary test area (Southeastern Connecticut) average $0.32 per 1000 liters. Waste treatment facilities are normally operated by the municipal government, and sewerage use charges are much less than the actual cost of wastewater collection and treatment. Usually, homeowners are billed directly only for operational costs of sewerage services. Other costs, such as planning, engineering, and construction of waste treatment and collection facilities, are covered by initial assessment fees, bond issues, and property taxes. Because of encouragement by recent federal guidelines for supplemental funding of new waste facilities, there is a trend towards consolidation of water and waste treatment management and towards the distribution of oper- ating costs on the basis of water usage rather than fixed service charges. Typical sewerage use rates in the Connec- ticut area range from zero for increments in use (fixed service charges)to $0.13/1000 liters. Treatment costs for septic systems with poor soils are based on data from References 11 and 12, (not updated, therefore a conserva- tive basis). When compared with typical water and sewer use rates, all of the bathroom flow reduction devices listed were shown to be economically acceptable in terms of cost savings. In general, the dual flush devices show the greatest potential for cost savings, even when considered for those communi- ties in which water rates are low to average, and sewer rates are not based on water consumption. The flow limiting shower heads, despite their low cost, proved to be of less economic value than anticipated largely because of the rather limited water savings obtained with these devices. At least one-half of the cost savings shown for shower heads are attributable to a reduction in hot water heating requirements. The shallow trap toilet is definitely war- ranted in terms of cost savings for new installations or necessary replacements, since the material and installation costs for the standard and water saving toilets are approxi- mately the same. Because of its relatively high cost, however, the shallow trap toilet appears to be of limited economic value for replacement of workable toilets. ------- The cost analysis of the system to reuse waste wash water for toilet flushing and/or lawn sprinkling indicates that household recycle units become economically attractive when septic tank systems with poor soils or other inad- equacies are encountered. It should be borne in mind that the water savings indicated ($19 to $Wyr.) apply only to situations where existing septic systems are inadequate, or where new septic tank installations are being planned and incorporation of the recycle system will permit a less ex- pensive septic system. Comparison of the projected cost of the mass produced version with typical water and sewer rates indicates that the system can also effect marginal cost savings in high water rate areas. However, from a strictly economic point of view, the potential cost savings attainable in such circumstances ($18.20/yr.) is less than one would obtain by investing an amount equal to the system initial cost at 5$ interest. Steadily dwindling reserves of fresh water for domestic con- sumption in many areas of the country and the continuing emphasis on advanced waste treatment will undoubtedly en- courage future reuse of waste water at both the household and community levels through higher water and waste treat- ment rates. Another method of encouraging water conserva- tion and reuse would be to make the true cost of waste water collection and treatment more visible to the indivi- dual user. This can be accomplished if the municipality is able to recover some or all of the associated capital costs by direct charges to the consumer on the basis of water usage. Federal guidelines on user charges show a trend in this direction. 95 ------- SECTION DC REFERENCES 1. Bailey, J.R., et_ al.. A Study of Flow Reduction and Treatment of Waste Water From Households. Federal Water Quality Administration, Cincinnati, Ohio. Pub- lication Number 11050 FKE. NTIS Order No. PB 197-599. December 1969. 154 P. Published papers based on this report: a) Bailey, J. and H. Wallman. Plow Reduction of Waste From Households. Water and Sewage Works, 118(3): 68-70, March 1971. b) Bailey, J. and H. Wallman. A Survey of Household Waste Treatment Systems, Journal Water Pollution Control Federation, 4^. (12): 2349-2360, Decem- ber 1971. c) Wallman, H. Should We Recycle/Conserve Household Water? Proceedings Sixth International Water Quality Symposium, Water Quality Research Council, April 18-19, 1972. pp 75-76. 2. Cabin John Drainage Basin Water-Saving Customer Educa- tion and Appliance Test Program. Washington Suburban Sanitary Commission, February 14, 1973. 3. Bostian, H.E., S. Cohen, and H. Wallman. Water Con- servation by the User. Presented at the International Public Works Congress and Equipment Show sponsored by American Public Works Association, September 19, 1973 in Denver. Scheduled for publication in the American Public Works Association Reporter, June 1974. 4. Saline Water Conversion Summary Report, 1972-73, Office of Saline Water, U. S. Department of the Interior. 5. Lineaweaver, P.P., Jr., Geyer, and Wolff, Residential Water Use, Report V, Phase Two. John Hopkins Univer- sity, 1966. 6. Dunn, D.F., and T.E. Larson. Relationship of Domestic Water Use to Assessed Valuation, With Selected Demo- graphic and Socio-economic Variables. Journal of Amer- ican Water Works Association. 55, (4): 441, April 1963. 96 ------- 7. Snedecor, G.W., Statistical Methods. Fifth Edition. Ames, Iowa, Iowa State College Press, 1957. p 91. 8. Chambers, W.C., Chlorination for Control of Bacteria and Viruses in Treatment Plant Effluents. Journal Water Pollution Control Federation. 43, (2): 228-241, February 1971. 9. Standard Methods for the Examination of Water and Waste- water, 13th Edition. New York, N.Y. American Public Health Association, 1971. 874 p. 10. Bruvold, W.N., P.O. Ward.. Using Reclaimed Wastewater - Public Opinion. Journal Water Pollution Control Feder- ation. 44. (9): 1690-1696, September 1972. 11. Thomas, H.A., J.B. Coulter, T.W. Bendixen, and A.B. Edwards. Technology and Economics of Household Sewage Disposal Systems. Journal Water Pollution Con- trol Federation,. 3JL, (2): 113-145, February I960. 12. U.S. Department of Health, Education, and Welfare, "Manual of Septic-Tank Practice," Public Health Service Publication No. 526, 1967. 97 ------- SECTION X APPENDIX HOMEOWNER'S QUESTIONNAIRE SUMMARY NUMBER OF RESPONSES Yes No A. FLOW LIMITING SHOWER HEAD DEVICES 1. Was the device attractive in appearance? 16 0 2. Did the fixture provide an adequate supply of shower water for your purposes? 14 2* 3. Did the fixture provide an excessive supply of water for your purposes? 1 3U5. 4. Did the device perform satisfactorily? 16 0 5. Was the maintenance considered minimum? 16 0 6. Did you require more time to shower with this head than with your previous shower head? 2* 14 7. Would you recommend the use of this de- vice to others? 14 2* 8. Did the device provide an acceptable spray pattern? 14 2 9. Please indicate roughly what percentage of your total bathing was accomplished by showering? 30-100$ *9.6 1pm shower head B. SHALLOW-TRAP TOILET 1. Was the toilet attractive and acceptable from a decorative vi-ewpoint? 12 0 2. Did the toilet provide an adequate flush for everyday toilet use? 10 2 98 ------- B. SHALLOW-TRAP TOILET (Cont'd) C. 3. 4. 5. (a) Was a second flush ever required? (b) If yes, approximately how often? Did the device perform satisfactorily? Was the maintenance required minimum? 6. Would you recommend the shallow-trap toilet to others? DUAL-PLUSH DEVICE 1. Was this device easily operated? 2. Did this device provide for an adequate flush for the waste solids? 3. Was the reduced flush adequate for liquid wastes? 4. Did the device provide an excessive supply of water for flushing purposes? 5. Approximately how much of the time was the reduced flush used? 6. Was use of the full flush inconvenient? 7. With the toilet device installed, have you found it necessary to clean the toilet bowl more frequently? 8. Was the device easily installed? 9. Did the device function satisfactorily? Was the maintenance required minimum? 10 11 Would you recommend this toilet device to others? 12. Do you feel the retail price is too high? _0 NUMBER OP RESPONSES Yes Aver 10 12. 10 A JL d _6 _p_ 52 _o on \J \sL _0_ _6_ A JL _6 i? 0 No 3 . 10$ 2 _0 __2_ _0 _0 0 JL i _6 _6 jO 0 _0 -2 6 Econo-Plush: Sink-Bob: Saveit: !>13.95 A !> 3.00 - $4.00 6.99 99 ------- NUMBER OF RESPONSES Yes No D. RECYCLED WATER SYSTEM FOR TOILET FLUSHING 1. Did you find the recycled water system complex to use? JO 6 2. Did you find the system unmanageable? 0 6 3. Did the system provide an adequate water supply for flushing purposes? 6 0 4. Did the system perform satisfactorily? 6 0 5. Was the maintenance required minimum? 3 3 6. Did the system require a minimum of surveillance? 4 2 7. Did you find the reuse of treated wash water for toilet flushing objectionable? 1 5 8. Were there additional cleaning require- ments for the toilet bowl with this system? 6 0 9. Were there any problems because of: a) odors generated? 2 4 b) appearance or color of the water? 1 ^ c) excessive discoloration or staining of the toilet trap? 6 0 d) excessive foaming? 0 6 e) the requirement and use of disin~ fectants? 3 3 f) a concern for a possible health hazard? 2 4 g) the design of the system? _1_ 10. Would you recommend this system to others ?__4 E. RECYCLED WATER USED FOR LAWN OR GARDEN WATERING 1. Did the use of treated laundry/bath water for lawn or garden use have any obvious adverse effect on your lawn or soil? 0 2. Did it have any beneficial effect over conventional lawn watering? 0 100 ------- RECYC1 NUMBER OP RESPONSES Yes No WATER USED FOR LAWN OR GARDEN WATERING (Cont'd) 3. Would you recommend use of this system to others ? 2 4. Did you feel a concern for any possible health hazard? 2 5. Did the use of the system have an adverse effect on the operation of the sprinkler system? 0 *Homeowner #7 felt that manually operated lawn watering reuse was impractical for his situation, He did indicate approval of an automatically operated sprinkler set-up. 2* 0 F. GENERAL 1. 3. Did you realize a cost savings during the test program through the use of water saving devices or a recycling system by: Yes No way to evaluate a b c decreased water bill? decreased electric bill? decreased fuel bill? If the home was served by a sep- tic tank or leach field, were there any noticeable changes in service requirements for these facilities during the study? 6 £ If yes, explain: See Section VIII for explanations. Do you plan to include in your home any of the following devices when this test is completed? a) restrictive flow shower-head device? b) toilet flush water saving device? c) shallow-trap toilet? Yes No 101 ------- NUMBER OP P. GENERAL (Cont'd) RESPONSES Yes No 4. If recycle systems were commercially available, would you: a) reuse laundry /bath water for toilet flushing? _4 _2 b) reuse laundry/bath water for lawn or garden watering? 2 4 5. If answers to 4(a) and/or 4(b) were yes, how much would you be willing to spend? No more than: $200 1 . $300 2 . $400 2 , $500 2 . $600 0__ 6. Would you recommend homeowner's partici- pation in other test programs such as this? 7. Other comments and suggestions concern- ing this test program would be appreciated. G. IDENTIFICATION Name Age Sex Relation to homeowner 102 ------- * TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) EPA-670/2-74-071 4. TITLE AND SUBTITLE DEMONSTRATION OF WASTE FLOW REDUCTION FROM HOUSEHOLDS 3. RECIPIENT'S ACCESSION-NO. 5. REPORT DATE Sept. 1974; Date of Issue 6. PERFORMING ORGANIZATION CODE 7. AU Sheldon Cohen and Harold Wallman 8. PERFORMING ORGANIZATION REPORT NO U440-74-057 9, PERF /IING ORGANIZATION NAME AND ADDRESS Materials and Environmental Engineering, Electric Boat Division General Dynamics Corporation Groton, Connecticut 06340 10. PROGRAM ELEMENT NO. 1BB033 * ROAP 21 ASW; Task 010A 11. CONTRACT/GRANT NO. 68-01-0041 12. SPONSORING AGENCY NAME AND ADDRESS National Environmental Research Center Office of Research & Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 13. TYPE OF REPORT AND PERIOD COVERED Final 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES 16. ABSTRACT A two-year demonstration program was conducted to evaluate water savings, costs, performance and acceptability of various water-saving devices. Reduced flow toilets and flow limiting shower heads were installed in eight single- family dwellings. In three of the homes bath and laundry water was filtered, disinfected, and reused for toilet flushing and/or lawn sprinkling. The ex- perimental portion of the program ran from May 1971 to May 1973. ; Water requirements for toilet flushing were substantially reduced in an economically attractive and aesthetically acceptable manner. Shallow trap and dual-flush toilets resulted in average decreases in toilet water usage of 25% and 23%, respectively. Flow restricting shower heads proved to be relatively ineffective, however this result may have been due to use patterns unique to this study. Wash water recycle systems provided satisfactory operation throughout the test period. The average savings for toilet flushing reuse ranged between 23% and 26% of total water usage. The incorporation of lawn sprinkling as a supplemental reuse mode further reduced waste flow from homes by 16% to 18%. For single-family dwellings, recycle systems could effect marginal cost savings in high water and sewer use rate areas. They are definitely warranted when septic systems with poor drainage (due to soil or topography) are encountered. KEY WORDS AND DOCUMENT ANALYSIS DESCRIPTORS b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group *Water conservation, *Houses, *Water reclamation, Sewage treatment, Filtration, Disinfection, Cost Analysis *Sewage flow reduction *Home water reuse, Plumbing fixtures, Showers, Toilets, Lawn watering 13 B 8. DISTRIBUTION STATEMENT Release to public 19. SECURITY CLASS (ThisReport)' UNCLASSIFIED 21. NO. OF PAGES 111 20. SECURITY CLASS (This page) UNCLASSIFIED 22. PRICE EPA Form 2220-1 (9-73) 103 U.S. GOVERNMENT PRINTING OFFICE: 197A-657-585/5306 Region No. 5-M ------- |