139107

Transcript

1 Addressing Nitrate in California’s Drinking Water TECHNICAL REPORT 6: Drinking Water Treatment for Nitrate With a Focus on Tulare Lake Basin and Salinas Valley Groundwater Report for the State Water Resources Control Board Report to the Legislature California Nitrate Project, Implementation of Senate Bill X2 1 Center for Watershed Sciences University of California, Davis http://groundwaternitrate.ucdavis.edu Prepared for the California State Water Resources Control Board

2 Drinking Water Treatment for Nitrate T 6 echnical Report ddr A essing Nitrate in California’s Drinking Water Groundwater With a Focus on Tulare Lake Basin and Salinas Valley Report for the State Water Resources Control Board Report to the Legislature Pr b y : epared 1 Vivian B. Jensen Jeannie L. Darby of UC Davis and 2 and Craig Gorman of Jacobs Engineering Group, Inc. Chad Seidel Center for Watershed Sciences University of California, Davis California Nitrate Project, Implementation of Senate Bill X2 1 ared for: Prep California State Water Resources Control Board July 2012 1 Corresponding author: jdarby @ucdavis.edu 2 The first p ortion of this document (through the end of Section 3) was developed concurrently for the American Water Works Association (AWWA) as the report titled An Assessment of the State of Nitrate Treatment Alternatives (2011) through collaboration with Chad Seide l, Ph.D., P.E. and Craig Gorman, M.S., P.E. of Jacobs Engineering Group, Inc .

3 Suggested Citation: Jensen, V.B., Darby, J.L., Seidel, C. & Gorman, C. (2012) Drinking Water Treatment for Nitrate. Technical Report 6 in Addressing Nitrate in California’s Drinking Water with a Focus on Tulare Lake Basin and Salinas Valley : Center for Watershed Groundwater. Report for the State Water Resources Control Board Report to the Legislature. Sciences, University of California, Davis. An electronic copy of this Final Report is ava ilable from the following website: http://groundwaternitrate.ucdavis.edu Copyright ©2012 The Regents of the University of California All rights Reserved The University of California prohibits discrimination or harassment of any person on the basis of rac e, color, national origin, religion, sex, gender identity, pregnancy (including childbirth, and medical conditions related to pregnancy or childbirth), physical or mental disability, medical condition (cancer - related or genetic characteristics), ancestry, marital status, age, sexual orientation, citizenship, or service in the uniformed services (as defined by the Uniformed Services Employment and Reemployment Rights Act of 1994: service in the uniformed services includes membership, application for membersh ip, performance of service, application for service, or obligation for service in the uniformed services) in any of its programs or activities. University policy also prohibits reprisal or retaliation against any person in any of its programs or activities for making a complaint of discrimination or sexual harassment or for using or participating in the investigation or resolution process of any such complaint. University l laws. policy is intended to be consistent with the provisions of applicable State and Federa Disclaimer : This document is intended as a guide of nitrate treatment technologies and should be used as an informational tool only. The contents of this document are solely the responsibility of the authors and do not necessarily represent the of ficial views of supporting agencies. The selection and design of the most appropriate nitrate treatment alternative for a particular water system depend on a variety of factors and require the expertise f proprietary technologies is intended to provide information of experienced professional engineers. Discussion o about treatment alternatives and does not imply endorsement. For further inquiries, please contact Thomas Harter, Ph.D. [email protected] 125 Veihmeyer Hall University of California Davis, C A 95616 - 8628 Phone: 530 - 752 - 2709

4 Acknowledgments report titled Assessment of the State of Nitrate Treatment This document is largely based on the An (2011) prepared for the American Water Works Association (AWWA) Alternatives hrough collaboration t 3 with Ch ad Seidel, Ph.D., P.E. and Craig Gorman, M.S., P.E. of Jacobs Engineering Group, Inc. Special thanks to the members of the AWWA Inorganic Contaminants Research and Inorganic Water Quality Project Subcommittee: Name Affiliation Water Works Engineers Michelle De Haan, Chair Jennifer Baldwin, Chair CH2M HILL Jess Brown Carollo Engineers Susan Brownstein California Department of Public Health Dennis Clifford University of Houston AWWA Elise Harrington Tarrah Henrie e Company California Water Servic France Lemieux Health Canada Jerry Lowry Lowry Systems, Inc. Support for this research was provided from the following sources: American Water Works Association Technical & Education Council   California Department of Public Health Safe D rinking Water Revolving Fund Contract No. 06 - 55254  California State Water Resources Control Board, Contract No. 09 - 122 - 250 3 The first portion of this document (through the end of Section 3) was developed concurrently for the American Water Works (2011) through report titled An Assessment of the State of Nitrate Treatment Alternatives Association (AWWA) as the . collaboration with Chad Seidel, Ph.D., P.E. and Craig Gorman, M.S., P.E. of Jacobs Engineering Group, Inc Technical Report 6: Drinking Water Treatment for Nitrate

5 The authors would like to thank the following for their contributions to the project: Balazs University of California, Ber keley Carolina Self - Help Enterprises Paul Boyer Canada UCD – Center for Watershed Sciences Holly Paul CDPH Division of Drinking Water and Environmental Management Collins UCD – Rob Information Center for the Environment Coman Commandatore CDPH Division of Drinking Water and Environmental Management Marc Jesse Dhaliwal CDPH – Southern California Drinking Water Field Operations Branch Laurel Community Water Center Firestone Fresno County Environmental Health Wayne Fox Fryjoff - Anna UCD – Center for Waters hed Sciences Hung Charles Hemans Tulare County Department of Environmental Health – Water Program Center for Watershed Sciences Kristin Honeycutt UCD – Chris Johnson Aegis Groundwater Consulting, LLC Sally Keldgord Kern County Environmental Health Aaron UCD – Center for Watershed Sciences K ing Lichti CDPH – Fresno District Engineer Betsy Longley CSU Fresno Karl Elena UCD – Center for Watershed Sciences Lopez Moore Pacific Institute Eli Sam Perry Washington State Department of Health Engineer Joe Pra do Fresno County Jim Quinn UCD – Information Center for the Environment Cheryl Monterey County Health Department – Drinking Water Protection Sandoval Services Tran Andrew – Graduate Student Researcher UCD Leah Walker CDPH – Division of Drinking Water and Environmental Management Emily Wallace UCD – Undergraduate Student Assistant Additionally, the authors would like to thank the drinking water utilities that participated in the survey he project. and the treatment technology vendors for their contributions to t Technical Report 6: Drinking Water Treatment for Nitrate

6 Contents ... ... ... ... v Tables ... ... ... ... ... vii Figures ... ... ... ... Acronyms and Abbreviations ix ... Unit Conversions ... ... ... ... ... xi ... ... ... ... ... 1 Summary Objective ... ... ... ... ... 1 Background ... ... ... ... 1 ... Approach ... ... ... ... ... 2 Findings ... ... ... ... ... 5 Non Treatment Options ... ... ... ... 5 - Treatment Options ... ... ... ... 5 9 ... Conclusions ... ... ... ... 1 Introduction ... ... ... ... 12 ... ... ... ... 12 1.1 Management Options for Nitrate in Potable Water - Treatment Options for Nitrate Contaminated Potable Wa ter ... ... 2 Non 14 2.1 Well Abandonment, Inactivation, and Destruction ... ... ... 14 ... ... ... 14 2.2 Wellhead Protection and Land Use Management 2.3 Development of Alternative Sources and Source Modification ... ... 14 2.4 Blending ... ... ... ... ... 16 3 Treatment Options for Nitrate Contaminated Potable Water ... ... 18 3.1 Ion Exchange (IX) ... ... ... ... 21 3.1.1 Conventional Ion Exchange ... ... ... ... 21 ... 3.1.2 Ion Exchange Design Considerations - ... ... 23 28 ... ... 3.1.3 Ion Exchange - Cost Considerations ... i Technical Report 6: Drinking Water Treatment for Nitrate

7 3.1.4 Ion Exchange Selected Research ... ... ... 29 - - ... ... 29 3.1.5 Ion Exchange Summary of Advantages and Disadvantages ... ... ... 29 3.1.6 Modif ications to Conventional Ion Exchange - Case Studies ... ... 3.1.7 Ion Exchange ... 35 ... 3.2 Reverse Osmosis (RO) ... ... ... ... 48 ... 3.2.1 Reverse Osmosis Design Considerations ... ... 48 - 3.2.2 Reverse Osmosis - Cost Considerations ... ... ... 53 3.2.3 Reverse Osmosis - Selected Research ... ... ... 54 - Summary of Advantages and Disadvantages ... ... 54 3.2.4 Reverse Osmosis 3.2.5 Reverse Osmosis - Improvements and Modifications ... ... 55 57 3.2.6 Reverse Osmosis Case Studies ... ... ... . - 3.3 Electrodialysis (ED/EDR/SED) ... ... ... ... 73 75 ... 3.3.1 Electrodialysis - ... Design Considerations ... 3.3.2 Electrodialysis - Cost Considerations ... ... ... 77 ... 3.3.3 Electrodialysis Selected Research ... ... 78 - 3.3.4 Electrodialysis - Summary of Advantages and Disadvantages ... ... 78 3.3.5 Modifications to Electrodialysis ... ... ... . 79 3.3.6 Electrodialysis - Case Studies ... ... ... ... 80 ... 3.4 Biological Denitrification (BD) ... ... 88 ... 3.4.1 Biological Denitrification Design Considerations ... ... ... 90 - ... - Cost Considerations ... ... 95 3.4.2 Biological Denitrification 3.4.3 Biological Denitrification - Selected Research ... ... ... 96 3.4.4 Biological Denitrification Summary of Advantages and Disadvantages ... .. 9 6 - 3.4.5 Biological Denitrification - Case Studies ... ... ... 97 106 ... ... 3.5 Chemical Denitrification (CD) ... ... ii Technical Report 6: Drinking Water Treatment for Nitrate

8 3.5.1 Zero Valent Iron (ZVI) ... ... ... ... 106 ... ... ... 108 3.5.2 Catalytic Denitrification ... - Design Considerations ... ... ... 108 3.5.3 Chemical Denitrification - Emerging Technologies ... ... .. 110 3.5.4 Chemical Denitrification ... 3.5.5 Chemical Denitrification Cost Considerations ... ... 113 - 3.5.6 Chemical Denitrification - Selected Research ... ... ... 114 3.5.7 Chemical Denitrification - Summary of Advantages and Disadvantages ... . 114 ... 3.6 Brine Treatment Alternatives and Hybrid Treatment Systems 114 ... 3.6.1 Electrochemical Destruction of Nitrate in Waste Brine ... ... 115 3.6.2 Catalytic Treatment of Waste Brine ... ... ... 116 Entry) 3.7 Residential Treatment (Point of - Use, Point - of - - ... ... ... 117 ... 120 ... 4 Tulare Lake Basin and Salinas Valley - Water Quality Analysis 4.1 Water Quality - Treatment Interference ... ... ... 121 ... - Co - contaminants ... 4.2 Water Quality ... ... 121 4.3 Water Quality and Treatment Selection ... ... ... 128 5 Addressing Nitrate Impact ... ... 130 ed Potable Water Sources in California 5.1 Well Abandonment, Destruction and Inactivation ... ... ... 130 ... 5.2 Survey of Blending and Treat ing Systems ... ... 133 6 Treatment Cost Analysis ... ... ... 139 ... 6.1 Costs by Treatment Type ... ... ... ... 141 6.2 Costs by System Size ... ... ... ... 144 6.3 Costs by Water Quality Parameters ... ... ... 147 6.4 Disposal Costs ... ... ... ... . 149 7 Guidance for Addressing Nitrate Impacted Drinking Water ... ... .. 152 ... 152 ... 7.1 Checklist for the Selection of Mitigation Strategy ... Technical Report 6: Drinking Water Treatment for Nitrate iii

9 ... ... ... ... . 153 7.2 Decision Trees 8 Summary and Conclusions ... ... ... ... 155 ... ... 157 9 Literature Cited ... ... ... ... 175 10 Appendix ... ... ... ... ... 175 10.1 Tables of Selected Research ... ... ... iv Technical Report 6: Drinking Water Treatment for Nitrate

10 Tables ... ... 4 Table S.1. Utilities included in the case studies. ... ... ... 6 Table S.2. Potable water treatment options for nitrate management .3. Comparison of major treatment types. ... ... ... Table S 11 Table 1. Utilities included in the case studies. ... ... ... 19 . Potable water treatment options for nitrate management. ... ... 20 Table 2 Table 3. Selection of full - scale ion exchange installations for nitrate removal. ... ... 21 Table 4. Summary of design considerations for conventional IX. ... 24 ... Table 5. Selected published costs of ion exchange systems for nitrate removal. ... ... 29 Table 6. Summary of design considerations for reverse osmosis. ... ... 49 Table 7. Selected costs of reverse osmosis systems for nitrate removal. ... ... 54 Table 8. Summary of design considerations for electrodialysis/electrodialysis reversal. ... 75 78 ... Table 9. Sample EDR O&M costs. ... ... ... Table 10. Full - ... 89 scale biological denitrification systems for potable water treatment. l denitrification for nitrate removal. ... 90 Table 11. Summary of design considerations for biologica Table 12. Membrane biological reactor configurations. ... ... ... 94 Table 13. Cost information fo ... ... 96 r biological denitrification of potable water. Table 14. Summary of design considerations for chemical denitrification. ... ... 109 ble 15. Selected research on brine treatment alternatives and hybrid systems Ta ... . 115 119 Table 16. Costs of POU treatment for nitrate removal. ... ... ... Table 17. Summary of water quality data of high nitrate wells in Tulare Lake Basin and Salinas Valley. 120 Table 18. Comparison of major treatment types ... ... ... 129 . ... ... 129 Table 19. Influence of nitrate concentration on treatment selection. Table 20. Incidence of abandonment, destruction, and inactivation of nitrate impacted wells. ... 131 Table 21. Population and nitrate levels of systems in the study area treating or blending for nitrate. .. 137 v Technical Report 6: Drinking Water Treatment for Nitrate

11 Ta ... 138 ble 22. Nitrate level, well depth and well capacity for a Tulare County blending system. ... cost curves of IX for arsenic removal. Table 23. Cost estimation using ... 140 U.S. EPA Table 24. Summary of anion exchange and reverse osmosis cost information by system size. ... 146 Table 25. An exercise in the estimation of treatm ... 148 ent costs based on nitrate levels. Table 26. Brine disposal costs. ... ... ... ... 150 arch 175 ... Table A.1. Selected rese ... on the use of ion exchange for nitrate removal. 176 the use of reverse osmosis for nitrate removal. ... .. Table A.2. Selected research on 177 Table A.3. Selected research on the use of electrodialysis for nitrate removal. ... ... Table A.4. Selected research on the use of biological denitrification for nitrate removal. 178 ... Table A.5. Selected research on the use o f chemical denitrification for nitrate removal. ... 181 ... 182 Table A.6. Advantages and disadvant ages of the five major treatment options for nitrate removal. vi Technical Report 6: Drinking Water Treatment for Nitrate

12 F igures ... ... 2 Figure S.1. Summary of nitrate management options. ... ... ... 16 Figure 1. Selective well screening using a packer/plug. ... ... ... Figure 2. Conventional ion exchange schematic. 22 ... Figure 3. Process flow diagram for counter current MIEX® process. ... ... 30 ... ... 32 Figure 4. Vessel rotation in Calgon Carbon countercurrent ISEP® system. Figure 5. Example of flow through the ISEP® system. ... ... ... 32 Figure 6. Example of an Envirogen multiple bed proprietary anion exchange system. 33 ... Figure 7. Process schematic of weak base ion exchange for nitrate ("WIN" process). ... 35 Figure 8. Reverse osmosis schematic. ... ... ... ... 48 Figure 9. Flow chart of the HERO™ process. ... ... ... 56 ... ... ... 73 Figure 10. Electrodialysis reversal schematic. 74 ... ... Figure 11. Illustration of electrodialysis membrane stack. ... Figure 12. Biological denitrification schematic. ... ... 88 ... Figure 13. FBR configuration. ... ... ... ... 98 Figure 14. FBR treatment system schematic. ... ... ... 100 Figure 15. Surface chemistry o f ZVI particles ... ... ... 107 Figure 16. Process schematic for denitrification using SMI - III®. ... ... 111 Figure 17. Schematic of the brine treat ment system developed by Ionex SG Limited. ... 116 Figure 18. Schematic of an ion exchange system with brine regeneration coupled with catalytic treatment of brine for reuse. ... ... ... ... 116 Figure 19. Raw water nitrate levels exceeding the MCL ... ... ... 122 Figure 20. Raw water high nitrate wells with high arsenic levels. ... ... 123 ... ... 124 Figure 21. Raw water high nitrate wells with high perchlorate levels. ... ... 125 Figure 22. Raw water high nitrate wells with pesticides detected. vii Technical Report 6: Drinking Water Treatment for Nitrate

13 - Figure 23. Salinas Valley [As] versus total well depth (deepest water) and [NO ] versus depth to top of 3 screen (shallowest water). ... ... ... ... 127 - Figure 24. Tulare Lake Basin [As] versus total well depth (deepest water) and [NO ] versus depth to top 3 ... ... of screen (shallowest water). ... 127 ... Figure 25. Tulare Lake Basin: Incidence of nit rate and arsenic MCL exceedance with well depth. ... 128 Figure 26. Location of nitrate impacted abandoned, destroyed, and inactivated wells. ... 132 Figure 27. Digital survey distributed to drinking water systems treating and/or blending to address high nitrate levels. ... ... ... ... ... 134 Figure 28. California drinking water systems treating and/or blending for nitrate. ... 135 Figure 29. Drinking water systems treating or blending for nitrate in the Tulare Lake Basin and Salinas ... ... 136 Valley. ... ... ... Figure 30. Wells of a blending system in Tulare County. ... ... ... 138 Figure 31. Average cost comparison of nitrate treatment technologies. ... ... 142 Figure 32. Costs of anion exchange for nitrate treatment. ... ... ... 143 ... Figure 33. Costs of reverse osmosis for nitrate treatment. ... ... 143 Figure 34. Costs of biological denitrification in drinking water treatment. ... ... 144 Figure 35. Cost curve of IX (blue) and RO (red) for nitrate removal. ... ... 145 Figure 36. Decision Tree 1 - Options to address nitrate impacted drinking water sources ... 153 154 ... ... Figure 37. Decision Tree 2 - Anion exchange. ... viii Technical Report 6: Drinking Water Treatment for Nitrate

14 Acronyms and Abbreviations Acre Feet AF AFY Acre Feet per Year American Water Works Association AWWA BMP Best Management Practice BD Biological Denitrification Bed Volume BV CD Chemical Denitrification CIP Clean - In - Place Disinfection Byproduct DBP Detection Limit DL DO Dissolved Oxygen Empty Bed Contact Time EBCT Electrodialysis ED EDR Electrodialysis Reversal FBR Fluidized Bed Reactor Fixed Bed Reactor FXB Granular Activated Carb on GAC GHG Greenhouse Gas GPM Gallons per Minute Groundwater Under Direct Influence (of Surface Water) GWUDI ™ HERO High Efficiency Reverse Osmosis Hydraulic Loading Rate HLR ® ISEP Ion Exchange Separation System IX Ion Exchange LSI Saturation Index Langelier MBfR Membrane Biofilm Reactor MBR Membrane Bioreactor MCL Maximum Contaminant Level MGD Million Gallons per Day ® MIEX Magnetic Ion Exchange NDMA nitrosodimethylamine N - Operations and Maintenance O&M PICME Permits I nspection Compliance Monitoring and Enforcement POE Point - of - Entry POU Use Point - of - PRB Permeable Reactive Barrier RO Reverse Osmosis SBA IX Strong Base Anion Exchange Sequencing Batch Reactor SBR ix Technical Report 6: Drinking Water Treatment for Nitrate

15 Silt Density Index SDI SDWA Safe Drink ing Water Act SED Selective Electrodialysis SMI Sulfur Modified Iron TDS Total Dissolved Solids Total Suspended Solids TSS Ultra - Low Pressure Reverse Osmosis ULPRO flow Sludge Blanket Reactor - Up USBR United States Environmental . S . E PA U Protection Agency VSEP Vibratory Shear Enhanced Process WAC Weak Acid Cation Exchange Weak Base Anion Exchange WBA IX WQM Water Quality Monitoring Zero Valent Iron ZVI x Technical Report 6: Drinking Water Treatment for Nitrate

16 Unit Conversions Metric to US US to Metric Mass Mass 0.04 ounces (oz) 1 ounce 28.35 grams 1 gram (g) 2.2 pounds (lb) 1 pound 0.45 kilograms 1 kilogram (kg) 1.1 short tons 1 short ton (2000 lb) 1 megagram (Mg) (1 tonne) 0.91 megagrams 1 gigagram (Gg) (1000 tonnes) 1102 short tons 1000 short tons 0.91 gigagrams Distance Distance 0.39 inches (in) 1 inch 2.54 centimeters 1 centimeter (cm) 1 meter (m) 3.3 feet (ft) 1 foot 0.30 meters 1 meter (m) 1.09 yards (yd) 1 yard 0.91 meters 1 kilometer (km) (mi) 1 mile 1.61 kilometers 0.62 miles Area Area 2 2 10.8 square feet (ft ) ) 1 square foot 0.093 square meters 1 square meter (m 2 2 ) 1 square kilometer (km (mi 0.39 square miles ) 1 square mile 2.59 square kilometers 1 hectare (ha) 1 acre 0.40 hectares 2.8 acres (ac) Volume Volume 1 liter (L) 0.26 gallons (gal) 1 gallon 3.79 li ters 3 3 1 cubic foot 35 cubic feet (ft 1 cubic meter (m 0.03 cubic meters ) (1000 L) ) 0.81 million acre - feet 3 1 cubic kilometer (km ) feet - 1 million acre 1.23 cubic kilometers - ac MAF, million ( ft) Farm Products Farm Products 0.89 pounds per acre 1.12 kilograms per 1 pound per acre 1 kilogram per hectare (kg/ha) hectare (lb/ac) 1 tonne per hectare 0.45 short tons per acre 1 short ton per acre 2.24 tonnes per hectare Flow Rate Flow Rate 1 cubic meter per day 3.38 cubic meters per 0.296 acre - feet per year 1 acre foot per year - 3 - ft/yr) (ac /day) (m day 1 million cubic meters per day 0.0038 million cubic 1 mega gallon per day 264 mega gallons per day 3 m million (694 gal/min ) (mgd) ( meters/ day /day) Units Nitrate - as milligrams/liter as nitrate ( mg/L rwise noted, nitrate concentration is reported as NO * Unless othe ). 3 To convert from: - Nitrate - N (NO - N)  Nitrate (NO multiply by 4.43 ) 3 3 - N) multiply by 0.226 Nitrate (NO N (NO - )  Nitrate - 3 3 xi Technical Report 6: Drinking Water Treatment for Nitrate

17 Summary Objective The purpose of the first three sections of this document is to provide a detailed guide to the current state of nitrate treatment alternatives that can be used as a reference tool for the drinking water community. The remainder of this document focuses on nitrate tre atment of drinking water in California and specifically in the Tulare Lake Basin and the Salinas Valley. Background Nitrate contamination of potable water sources is becoming one of the most important water quality ted States. The maximum contaminant level (MCL) for nitrate is concerns in California and across the Uni - 45 mg/L as nitrate (NO ), which is approximately equivalent to 10 mg/L as nitrogen (N). The major 3 health concern of nitrate exposure through drinking water is the risk of methemoglobinemia, or “blue baby syndrome,” especially in infants and pregnant women. Due to the nature of the infant digestive system, nitrate is reduced to nitrite which can render hemoglobi 2010). n unable to carry oxygen (SWRCB Nitrate is naturally occurring at low leve ls in most waters, but it is particularly prevalent in groundwater or industrial activities. Of specific concern that has been impacted by certain agricultural, commercial , r treatment facilities, are crop fertilization activities and discharges from animal operations, wastewate d by nitrate (Pacific Institute 2011). and septic systems. Small rural communities are particularly impacte Nitrate presents unique water treatment challenges. The United States Environmental Protection . S . EPA) lists onl y anion exchange (IX), reverse osmosis (RO), and electrodialysis reversal (EDR) Agency (U e removal ( as accepted potable water treatment methods for nitrat 2010). Due to the U . S . EPA technologies is often production of high - strength brine residuals, sustainable application of these three limited by a lack of local residual disposal options and the challenge of increasing salt loads. The lack of affordable and feasible nitrate treatment alternatives can force impacted utilities to remove nitrate - contaminated sources from their available water supply. In many instances, this action can severely compromise a water utility’s ability to provide an adequate supply of safe and affordable potable water. The need for additional nitrate treatment technologies has driven the drinking water community to begin developing alternative options to effectively remove nitrate while limiting cost and brine production challenges. Promising treatment options include weak base anion (WBA) exchange and improvements in strong base anion (S BA) exchange such as low brine residual technologies; biological treatment using fluidized bed, fixed bed, and membrane biofilm (MBfR) reactors; and chemical reduction using media such as zero valent iron (ZVI) and sulfur modified iron (SMI). A summary of the g water is presented in Figure S.1. In this diagram options to address nitrate contamination of drinkin treatment options are classified in terms of their ability to either remove nitrate to a residual waste en species through reduction. stream or transform nitrate to other nitrog 1 Technical Report 6: Drinking Water Treatment for Nitrate

18 4 S. 1 . Summary of n itrate m anagement o ptions . Figure Approach This report includes a comprehensive literature review and case studies of specific systems across the range of nitrate treatment alt ernatives. The literature review is intended to provide background information about current and emerging potable water treatment alternatives to address nitrate contamination. In addition to peer - reviewed literature, information found in the “grey paper s” of conference proceedings has been included to assure capture of the most recent technology developments. For each of the major treatment technologies, subsections of the literature review detail the following: , lity, system layout esign considerations including water qua and site considerations; residuals  D , and operational complexity, management and disposal; and maintenance, monitoring c  ost considerations,  s elected research, and summary of advantages and disadvantages.  a 4 treatme For the purposes of this discussion, blending has been categorized as a “non - nt” option; however, in practice, blending is sometimes referred to as “treatment.” Treatment options throughout this report refer to treatment technologies - effectively address available for the removal or reduction of nitrate in drinking water. Blending can sometimes be us ed to cost the nitrate problem through dilution, but has been categorized separately from treatment options for simplicity. 2 Technical Report 6: Drinking Water Treatment for Nitrate

19 Information is summarized in t ables whenever appropriate, including a summary table of selected research studies for each of the major treatment technologies (Appendix). A survey was conducted to collect detailed information about the application of nitrate treatment. A ities, currently treating for nitrate and/or in design for future treatment, was included in subset of util the survey. The survey was developed to gather information with respect to the benefits and limitations of the various nitrate treatment technologies and was cond ucted via phone and in - person when applicable. The list of utilities included in the survey was developed with the intention of covering a ual range of utilities with respect to geographic location, treatment type, population size and resid ques (Table S.1). Detailed case studies have been compiled for each of the treatment handling techni technologies where full - scale facilities have been in operation or are moving ahead with design. This survey was conducted through collaboration with Jacobs Engineering i n the completion of the the American Water Works Association associated assessment of nitrate treatment alternatives for AWWA and is complemented by a parallel survey of nitrate treatment systems in California. Details ) ( from the initial survey are includ ed as examples following a discussion of each of the treatment technologies. Details of the complementary survey of California systems are included in the second half of this report. 3 Technical Report 6: Drinking Water Treatment for Nitrate

20 Table S. 1 . Utilities included in the c ase studies. Avg. Influent Nitrate Capacity (gpm) Treatment Type Location Case # - (mg/L as N) mg/L as NO 3 Ion Exchange Conventional ion exchange with blending California 400 1 31 – 53 (7 – 12) 2 Conventional ion exchange with blending California ~4 5 (~10) 400 ® Counter Current Ion Exchange (MIEX – ) Indian Hills, CO 50 53 3 71 (12 – 16) 35 4 California 500 – 900 Multiple vessel ion exchange – 89 (8 – 20) 5 Multiple vessel ion exchange Chino, CA 5000 40 – 200 (9 – 45) Reverse Osmosis 6 Reverse osmo sis and blending Bakersfield, CA 120 75 – 84 (17 – 19) Reverse osmosis, exploring biological 20) 7 – 89 (11 Brighton, CO 4600 49 – reduction Reverse osmosis and blending Arlington Desalter, Riverside, CA 4583 44 – – 20) 8 89 (10 Combined Reverse Osmosi s and Ion Exchange Reverse osmosis, ion exchange and blending Chino Desalter I, Chino, CA 4940 (RO), 3400 (IX) 148 – 303 (33 – 68) 9 224 (16 Chino Desalter II, Mira Loma, CA 4167 (RO), 2778 (IX) 70 – Reverse osmosis, ion exchange and blending – 51) 10 Electrodialysis /Electrodialysis Reversal/Selective Electrodialysis 3,260 (each, 2 systems) 11 ~80 (~18) Electrodialysis Reversal (EDR) Spain 89 (19 Israel 310 84 – Selective Electrodialysis (SED) – 20) 12 Biological Denitrification idized bed biological Implementing flu 13 ) 5 Rialto, CA 2000 – 4000 17 – 19 ( ~4 – reduction – 20) 14 Implementing fixed bed biological reduction Riverside, CA 1670 44 – 89 (10 4 Technical Report 6: Drinking Water Treatment for Nitrate

21 Findings Non - Treatment Options The focus of this assessment is the current state of nitrate trea tment alternatives. However, in practice, non treatment options are generally considered first as they can often be more sustainable and less - costly. Non - treatment options include wellhead protection, land use management, well inactivation, cation, development of alternative sources (including consolidation/connection to a source modifi nearby system), and blending. Blending was found to be the most common method to address nitrate contamination. When a low nitrate water supply source is available, dilut ion of high nitrate sources to effective than installing - produce water with nitrate levels below the MCL is typically more cost treatment. Treatment Options Nitrate treatment technologies were categorized into five major types. Ion exchange (IX), reverse osmosis (RO), and electrodialysis/electrodialysis reversal (ED/EDR) remove nitrate to a concentrated waste stream, while biological denitrification (BD) and chemical denitrification (CD) transform nitrate to other nitrogen species through reduction. Comm on concerns in the application of the removal technologies include waste management costs and treatment interference from other water quality parameters (e.g., hardness and sulfate). Pretreatment is often required to avoid fouling or scaling of the resin for IX and the membranes for RO and ED/EDR. Due to the destruction of nitrate, both biological and chemical denitrification have the potential for more sustainable treatment without brine residuals, cation of these nitrate treatment options is - scale appli but also have limitations to consider. Full currently limited. The selection of the most appropriate treatment option depends on various key factors specific to the needs and priorities of individual water systems. A brief comparison of fundamental desi gn and considerations, S.2. It is ent options is listed in Table advantages and disadvantages of these treatm ote that the contents of Table S.2 are not intended to provide a comprehensive set of important to n criteria for treatment options. Other importa nt criteria in determining the best treatment option, which are site specific and cannot be broadly generalized, include capital and operations and maintenance t option (O&M) costs, system size (capacity), and system footprint. Overall, there is no single treatmen that can be considered the best method for nitrate removal across all water quality characteristics and for all systems. 5 Technical Report 6: Drinking Water Treatment for Nitrate

22 Table S. Potable water treatment options for nitrate management (adapted from WA DOH 2005). 2 . Electrodialysis Biological Denitrification Chemical Denitrification Ion Exchange Reverse Osmosis scale Systems Full Yes Yes Yes Yes No - Removal to waste stream Removal to waste stream Removal to waste stream Biological reduction Chemical re duction Treatment Type Sulfate, iron, manganese, total Turbidity, iron, manganese, Turbidity, iron, manganese, Common Water suspended solids (TSS), metals Temperature and pH, anoxic SDI, particle size, TSS, TSS, hydrogen sulfide, Temperature and pH Quality Design hardness, organic matter, hardness, metals (e.g., (e.g., arsenic), hardness, conditions Considerations arsenic) c) organic matter metals (e.g., arseni pH ad justment, nutrient and Pretreatment - filter, address hardness Pre - filter, address hardness Pre - filter, address hardness Pre substrate addition, need for pH adjustment Needs anoxic conditions n pH adjustment, iro treatment Post - Filtration, disinfection, possible pH adjustment pH adjustment pH adjustment removal, potential ammonia substrate adsorption Remineralization Needs Remineralization control Waste/Residuals Sludge/biosolids Concentrate Waste media, Iron sludge Waste brine Concentrate Management Initial plant startup: Days to weeks Minutes Minutes Start up Time Minutes Minutes - After reaching steady state: Minutes Conventional (97%) scale - Water Recovery Not demonstrated full Nearly 100% Up to 85% Up to 95% Low brine (Up to 99.9%) No waste brine or concentrate, No waste brine or Multiple contaminant nitrate reduction rather than concentrate, nitrate Nitrate selective resins, removal, higher water Multiple contaminant transfer to a waste stream, high reduction rather than Advantages removal, desalination common application, (TDS recovery water recovery, and potential transfer to a waste stream, multiple contaminant removal removal) (less waste), desalination, and potential for multiple for multiple con taminant unaffected by silica removal contaminant removal Substrate addition, potentially Inconsistency of nit rate complex, high monitoring more reduction, risk of nitrite Potential for nitrate peaking, Membrane fouling and formation (potential needs, possible sensitivity to Energy demands, scaling, lower water recovery, high chemical use (salt), brine incomplete denitrification), environmental conditions, risk of waste disposal, potential operational complexity, operational complexity, for Disadvantages reduction to ammonia, lack nitrite formation (potential disinfection byproduct (DBP) energy demands, waste of full scale systems, pH and - waste disposal incomplete denitrification), post treatment to address temperature dependence, formation (e.g., NDMA) - disposal log - possible need for iron turbidity standards and 4 virus removal (state dependent) removal Technical Report 6: Drinking Water Treatment for Nitrate 6

23 Ion Exchange (IX) The most commonly used nitrate treatment m ethod is IX. Anion exchange for nitrate removal is similar to a water softener, with nitrate ions removed rather than hardness ions. Nitrate is removed from the treatment stream by displacing chloride on an anion exchange resin. Subsequently, regenerati on of the resin is necessary to remove the nitrate from the resin. Regeneration is accomplished by using a highly concentrated salt solution resulting in the displacement of nitrate by chloride. The result is a concentrated waste brine solution high in n itrate that requires disposal. The most significant drawback of this treatment option is the cost for disposal of waste brine, especially for inland communities. The brine volume is largely dependent on the raw water quality and the configuration of the system. Key factors in the consideration of IX include the pretreatment requirements to avoid resin fouling, the potential need for nitrate selective resin, the frequency of resin replacement, the possible post - r other product water quality concerns (e.g., the treatment requirements to address corrosion o potential for NDMA formation), and the management of waste brine. If waste brine disposal options are not limiting, IX can be the best option for low to moderate nitrate contamination and removal of and chromium). Application of IX may not be multip le contaminants (including arsenic, perchlorate , feasible for extremely high nitrate levels due to salt use and waste volume. Current research on brine nologies capable of effectively addressing treatment alternatives may lead to the development of tech scale implementation of this are unknown at this point. - the disposal concern; however, the costs for full Modifications to conventional IX have emerged in recent years offering low brine alternatives with efficiency. The efficiency of IX systems is dependent on the raw water characteristics. It is improved important to note that there can be cases where conventional IX systems yield greater water efficiency with lesser water quality. than a modified system that is implemented at a location Another promising alternative to consider for the future is weak base ion exchange (WBA IX). This emerging technology is more operationally complex than conventional IX, but may offer the advantage of recycling waste as fertil izer. Reverse Osmosis (RO) As the second most common nitrate treatment alternative, RO can be feasible for both municipal and Use applications and can be used simultaneously for desalination and removal of nitrate and - - Point of many co - contaminants. Follow ing pretreatment to prevent membrane fouling and scaling, water is - forced through a semi permeable membrane under pressure such that the water passes through, while contaminants are impeded by the membrane. - off between water etreatment requirements, the trade Key factors in the consideration of RO are the pr recovery and power consumption, the management of waste concentrate, and the typically higher costs the relative to IX. One deciding factor favoring the selection of RO over IX for nitrate removal would be need to address salinity. 7 Technical Report 6: Drinking Water Treatment for Nitrate

24 Recent advancements in membrane technology and optimization of pre - and post - treatment have led to - increases in the efficiency of RO treatment systems. For example, the use of Ultra Low Pressure Reverse Osmosis (ULPRO) membra nes enables lower power consumption. Electrodialysis (ED, EDR, SED) The use of ED in potable water treatment has increased in recent years, offering the potential for lower residual volumes through improved water recovery, the ability to selectively remo ve nitrate ions, and the minimization of chemical and energy requirements. ED works by passing an electric current through a series of anion and cation exchange membranes that trap nitrate and other ions in a concentrated g and thus the need for chemical addition, the polarity of the system waste stream. To minimize foulin can be reversed with electrodialysis reversal (EDR). By reversing the polarity (and the solution flow direction) several times per hour, ions move in the opposite direction through the membranes, minimizing buildup. Key factors in the consideration of EDR are the pretreatment requirements, the operational complexity of the system, the limited number of system manufacturers, the management of waste concentrate , and the lack of full - scale installations for nitrate removal from potable water in the United States. Like RO, EDR is commonly used for desalination and can be an alternative for nitrate treatment of high TDS R costs are similar to RO and waters. In contrast to conventional RO, EDR is unaffected by silica. ED evidence suggests that EDR can be the preferable option as the Silt Density Index (SDI) increases. For very small particle sizes, robust pretreatment can be necessary for RO. It is important to note that the EDR process does not directly filter the treatment stream through the membranes; contaminants are transferred out of the treatment stream and trapped by the membranes. This generally minimizes membrane fouling, decreasing pretreatment requirements in comparison to RO. Biological Denitrification (BD) Biological denitrification in potable water treatment is more common in Europe with recent full - scale systems in France, Germany, Austria, Poland, Italy , and Great Britain. To date, full - scale drinking water applications in the United States are limited to a single plant in Coyle, OK (no longer online). However, - two full scale systems are anticipated in California in the next couple of years. Biological denitrification through reduction. Substrate and nutrient s relies on bacteria to transform nitrate to nitrogen ga - treatment can be more intensive than for the removal processes. addition is necessary and post Biological denitrification offers the ability to address multiple contaminants and the avoidance of costly waste br ine disposal. Key factors in the consideration of biological denitrification are the chemical requirements, the need for anoxic conditions, the level of operator training, the robustness of the system, and the post - treatment requirements. State regulation s are expected to vary and, until more experience with the application of biological denitrification for potable water treatment is obtained in the United States, pilot and ought to have a biological treatment is th , demonstration requirements may be intensive. Typically larger footprint; however, with the latest design configurations, the system footprint may be comparable to that of RO or EDR systems. 8 Technical Report 6: Drinking Water Treatment for Nitrate

25 With reduction of nitrate to nitrogen gas, the lack of a problematic brine waste stream is a clear advant age of biological treatment over the removal processes. Biological treatment has the potential to provide a sustainable nitrate treatment option for the long term. More will be known with the completion of the anticipated full - scale systems in California ; cost estimation suggests that biological treatment can be economically competitive with IX. Chemical Denitrification (CD) Chemical denitrification uses metals to transform nitrate to other nitrogen species. As an emerging technology, no full - scale chem ical denitrification systems have been installed in the United States for and application for nitrate treatment has been strictly limited to , nitrate treatment of potable water - and pilot - scale studies. A significant body of research has explored th e use of zero valent iron bench (ZVI) in denitrification. Several patented granular media options have also been developed including sulfur modified iron (SMI) media, granular clay media , and powdered metal media. Key factors in the consideration of chemical de nitrification are the reliability and consistency of nitrate reduction, the lack of full - scale installations, the type of media, and the dependence on temperature and pH. Chemical denitrification has the potential to become a feasible full - scale nitrate t reatment alternative, with the advantage of reducing nitrate to other nitrogen species and avoiding the need to dispose of a concentrated waste stream. However, currently this option is an emerging technology in and full - scale te sting. Due to the potential benefits, further research and - need of additional pilot optimization of chemical denitrification systems will likely make this a competitive option in the future, especially for multiple contaminants (e.g., arsenic and chromium). Conclusions  Current fu ll - scale nitrate treatment installations in the United States consist predominantly of IX and RO. While EDR is a feasible option for nitrate removal from potable water, the application he use of biological of EDR is generally limited to waters that have high TDS or silica. T denitrification to address nitrate contamination of drinking water is more common in Europe scale systems are - than in the U.S. However, this option is emerging in the U.S. and two full n may become a feasible nitrate treatment expected in a few years. Chemical denitrificatio option in the future; however, the lack of current full - scale implementation suggests the need for further research, development and testing.  Brine reuse and treatment are vital to the continued reliance on IX fo r nitrate treatment of potable water. The low brine technologies offer a minimal waste approach and current research and development of brine treatment alternatives seem to be lighting the path toward future progress.  ity and insufficient water quantity, the need to address In regions with declining water qual multiple contaminants will increase in the future, suggesting the future dominance of technologies capable of multiple contaminant removal. In this context, for any individual water , the most appropriate technology will vary with the contaminants requiring source or system 9 Technical Report 6: Drinking Water Treatment for Nitrate

26 mitigation. Although complex, analysis of the optimal treatment option for pairs and groups of contaminants will assist in the treatment design and selection. In such scenarios, the best treatment option for nitrate may not be the most viable overall.  Currently and into the future, selection of the optimal and most cost - effective potable water treatment options will depend not only on the specific water quality of a given water so urce, but also on the priorities of a given water system. If land is limited, the typical configuration required for biological treatment may not be feasible. If brine waste disposal options are costly or limited, implementation of denitrification treatm ent or development of brine recycling and treatment may be the most suitable option. When deciding on nitrate treatment, the characteristics of the water system must be taken into  ll water systems account as well. With consideration of economies of scale, many rural sma cannot afford to install treatment. Even with financial assistance to cover capital costs, the long term viability of a treatment system can be undermined by O&M costs that are simply not sustainable. For such systems, treatment can beco me more affordable through consolidation of multiple small water systems into larger combined water systems that can afford treatment as a treatment options alone, like conglomerate. With a continued decline in water quality, non - blending groundwater sour ces or drilling a new well, may become insufficient measures for a water system to provide an adequate supply of safe and affordable potable water. Especially in rural small communities, perhaps the most promising approach will be consolidation of multipl e Alternatively, nearby water systems and the installation of a single centralized treatment plant . For additional separate small treatment facilities can be consolidated under a single agency. discussion on the comparison of alternative water supply opt ions and associated costs see Technical Report 7 (Honeycutt et al. 2012) .  While current cost considerations are commonly the driving force in selecting nitrate treatment, it is essential to consider the long term implications of current industry decisions. For example, it may be cost - effective for a particular system to utilize conventional IX currently, but future water quality changes (e.g., increasing nitrate levels, co - contamination, high salt loading), discharge regulations, or disposal fees may lead to an unmanageable increase in costs. Environmental sustainability in drinking water treatment is being addressed with brine treatment alternatives and denitrification options. It is important to approach the future of indset that environmental sustainability and economic drinking water treatment with the m sustainability are tightly interwoven.  Centralized treatment may not be feasible for widespread rural communities ; another approach to consider is to centralized management (e.g., design, purchasing, and maintenance) minimize costs. of - Entry (POE) treatment equipment is an important option to  Point - of - Use (POU) and Point - consider, especially for the provision of safe drinking water from private wells. Unless ecomes an option, users relying on domestic wells connecting to a nearby public water system b 10 Technical Report 6: Drinking Water Treatment for Nitrate

27 have two main alternatives: drilling a new well to attain safe drinking water or installing a POU or POE device for the treatment of contaminated water. The use of POU and POE treatment equipment by small public water systems is currently only a temporary option in California and term would require regulation changes. - While POU and reliance on these devices for the long POE treatment equipment has been shown to effectively address nitrate and other properly maintain nants, it is important to contami these devices to ensure the supply of consistently safe drinking water.  Within the drinking water community, the options typically considered to address nitrate ies are available or emerging (EDR, BD, CD) contamination are IX and RO. Alternative technolog because, under some circumstances, they offer advantages over IX and RO. New technologies will continue to be investigated and developed because no single option is ideal for all situations. There is not a nitr ate treatment option currently available that can affordably address all possible scenarios. The following diagram is a rough guide for treatment technology selection based on water quality concerns and possible priorities for a given water sou rce or syst S.3). This diagram includes generalizations and is not intended to be definitive. In em (Table the selection of nitrate treatment technologies the unique needs of an individual water system must be assessed by professional engineers to optimize treatment selection and design. 1 . Comparison of m ajor t Table S. t ypes . 3 reatment Concerns Priorities EDR BD CD RO IX IX RO EDR BD CD High Nitrate High Hardness Not a Major Concern Removal High TDS Reliability Removal Arsenic Training/ Ease of Removal operation Radium and Minimize Capital Uranium Cost Removal Chromium Minimize Ongoing Removal O&M Cost Perchlorate Minimize Removal Footprint Industry Experience Unknown Ease of Waste  Poor Good (blank) Management 1 Ion Exchange (IX), Reverse Osmosis (RO), Electrodialysis Reversal (EDR), Biological Denitrification ( BD), Chemical This table offers a generalized comparison and is not intended to be definitive. There are Denitrification (CD). notable exceptions to the above classifications. 11 Technical Report 6: Drinking Water Treatment for Nitrate

28 1 Introduction of the most important water quality Nitrate contamination of potable water sources is becoming one concerns in . The maximum contaminant level (MCL), 45 mg/L as California and across the United States - is currently being approached or exceeded in potable water ) (10 mg/L as nitrogen (N)), nitrate (NO 3 t locations throughout the United States (Nolan et al. 2002; Chen et al. 2009 supply sources a U . S . EPA ; N.D. ). A major source of nitrate contamination is fertilizer. Application of fertilizer in excess of the amount taken up by crops leads to leaching into the groundwat er. Leakage from livestock feedlots and nitrate problem (LLNL 2002). Additional sources include waste storage also contributes to the , and various industrial applications. Due to the wastewater treatment discharge, faulty septic systems nitrate contamination is more common in rural agricultural areas. The major health typical sources, concern of nitrate exposure through drinking water is the risk of methemoglobinemia, especially in infants and pregnant women. Due to the nature of the infant digestive s ystem, nitrate is reduced to nitrite which can render hemoglobi n unable to carry oxygen (SWRCB 2010). 1.1 Management Options for Nitrate in Potable Water To meet the nitrate MCL in the provision of potable water, both non - treatment and treatment options are considered. Source management with non - treatment can sometimes provide less costly solutions through wellhead protection, land use management, well abandonment, source modification, tion to a nearby system), or development of alternative sources (including consolidation or connec 5 treatment options can be limited by various factors including location, - The feasibility of non blendin g. budget, source availability, and variability of water quality (i.e., fluctuations in nitrate levels), resulting in eed for treatment to remove or reduce nitrate. the n Current treatment methods include ion exchange (IX), reverse osmosis (RO), electrodialysis/ , and chemical denitrification (CD). These electrodialysis reversal (ED/EDR), biological denitrification (BD) te management options are examined in detail to assess research findings, capital and O&M costs, nitra typical limitations , and the latest improvements. Design and cost considerations will be addressed with the development of guidelines for determining the most appropriate treatment option based on source water quality and other water system characteristics. equipment should also be considered as part of Point - of - Entry (POE) and Point - of - Use (POU) treatment e Safe Drinking Water Act (SDWA) [Se comprehensive examination of nitrate treatment. Th ction a U 1998) identifies both POE and POU treatment units as options for compliance S . EPA . 1412(b)(4)(E)(ii)] ( technologies for small systems ; California regulations governing the use of POU and POE devices for water sys tem compliance currently restrict their use to a temporary basis and only for systems having particular characteristics (Section 3.7) . 5 For the purposes of this discussion, blending has been categorized as a “non - treatment” option; however, in p ractice, blending is sometimes referred to as “treatment.” Treatment options throughout this report refer to treatment technologies - effectively ad dress available for the removal or reduction of nitrate in drinking water. Blending can sometimes be used to cost the nitrate problem through dilution, but has been categorized separately from treatment options for simplicity. 12 Technical Report 6: Drinking Water Treatment for Nitrate

29 Lastly, hybrid systems are explored. The combination of multiple treatment technologies, including several developing brine treatment alternatives, can maximize the advantages of each option. The goal of this investigation is to provide an overview of management strategies and treatment options, highlighting the most recent advances and elucidating costs and common probl ems in application. Technical Report 6: Drinking Water Treatment for Nitrate 13

30 2 Non Treatment Options for Nitrate Contaminated Potable - Water 2.1 Well Abandonment, Inactivation, and Destruction With adequate capacity from other sources, the simplest option for management of nitrate ces is well abandonment contaminated potable water sour . However, the lack of and proper destruction sufficient alternative water supplies often rules out well abandonment as an option. Based on a recent survey conducted by the American Water Works Association (AWWA), 30.4% (17/56) of survey participants with wells impacted by nitrate selected well abandonment as the implemented option for addressing nitrate contamination (Weir & Roberson 2010 ; Weir & Roberson 2011). It is important to determine the local requirements for safely removi ng a well from service. For proper abandonment, local requirements can include covering, sealing , and plugging of the well to prevent contamination and to avoid hazardous conditions. Inactivation or abandonment of a well differs from well destruction. T hrough inactivation or temporary abandonment, the well can be brought back online in the future (e.g., when treatment is installed). In contrast, well destruction involves the filling of a well, making it no longer viable. The costs for proper well aband onment and destruction can be substantial and vary with , well public supply . Analysis of and local standards for well destruction well depth, diameter , location abandonment and destruction in the Tulare Lake Basin and Salinas Valley, as well as across Cal ifornia, is included below in the Section 5.1 Well Abandonment, Destruction , and Inactivation . Additional information on private well abandonment and destruction is provide d in Technical Report 2, Section 9 (Viers et al. 2012) . 2 .2 Wellhead Protection and Land Use Management While limiting current nitrate contamination of groundwater will not immediately remove the need for treatment, over time, load reduction will minimize source water nitrate levels. Agricultural practices, man agement of dairies, control of wastewater treatment plant discharge, and monitoring and for a full discussion remediation of septic tank s can be improved to minimize nitrogen loading ( discharge see Technical Report 3, Dzurella et a l. 2012) of source load reduction . For example, a project addressing well head protection and land use management performed by the University of Waterloo (Rudolph 2010) successfully decreased groundwater nitrate levels within a two year travel time from 17 to 7 mg/L total stored nitrogen. Reduced nitrogen loading was accomplished by purchasing agricultural land and implementing Best Management Practices (BMPs). 2.3 Development of Alternative Sources and Source Modification n and movement in the subsurface, a new well With adequate information about the nitrate distributio can potentially be developed to access higher quality source water. Due to the anthropogenic nature of . If suitable 2002) (Nolan et al. the contamination, nitrate concentration typically decreases with depth 4 Technical Report 6: Drinking Water Treatment for Nitrate 1

31 w ater quality exists, drilling a deeper well can remove the need for nitrate treatment. However, the quality improvements must be balanced by a potential decrease in source capacity. Due to drilling and pumping requirements, capital and operational costs increase with the depth to uncontaminated water. When considering the installation of a deeper well to avoid nitrate contamination, it is important to be aware of the risk of encountering other water quality concerns at greater depths (e.g., arsenic) (see Section 4.2 Water Quality - Co - contaminants ) . Connecting to a nearby water system that is not impacted by nitrate or to a larger system that can e, since 1995, the City afford nitrate treatment is often the best option for smaller systems. For exampl of Modesto, CA, has been in charge of providing compliant water to the residents of Grayson , using an ion exchange plant for nitrate removal (Scott 2010). Similarly, consolidation of multiple nearby small systems can decrease the c ost of treatment per customer to more reasonable levels. Additional or temporarily , alternative source options include purchasing water rights, trucking in potable water Reliance on hauled and/or bottled water is only an interim solution for use in relying on bottled water. emergencies or while an effective compliance option can be implemented. Technical Report 7 provides (Honeycutt et al. a comprehensive discussion of alternative water supply options and associated costs 2012) . Modification of impacted source wells can allow for withdrawal of water with lower nitrate levels by limiting screened intervals to regions of better water quality. Down hole remediation requires of the higher water characterization of the water quality profile to determine the screening depth range quality. Specialized monitoring equipment and techniques are available that can be used wit hout 2008). With water profile characterization, existing wells can be selectively removing pumps (BESST Inc. ( screened using a packer/plug to limit withdrawal from unwanted regions Figure 1 ) . The effective application of such well modification techniques is dependent on the subsurface characteristics in the porous media can lead to surrounding he vicinity of the well. Vertical migration of nitrate through t increasing nitrate levels in the water withdrawn from the well. 15 Technical Report 6: Drinking Water Treatment for Nitrate

32 Figure 1 . Selective well screening using a packer/plug. The City of Ceres, CA, is in the process of drilling new wells , in p art to avoid the need for nitrate treatment; well modification has also been implemented to avoid water wit h high nitrate levels (Cannella 2009). 2.4 Blending ctive The dilution of a nitrate impacted source with an alternate low nitr ate source can be a cost - effe option to produce compliant water; this is known as blending and can be applied independently or with 6 treatment . Blending is a common practice for the production of compliant water, but relies on the High sistency of nitrate levels to avoid exceedances. availability of a low nitrate source and the con nitrate groundwater can also be blended with surface water when a surface water source is available; One drawback of implementing however, surface water treatment requirements would increase costs. bl ending to address nitrate contamination is that reliance on blending can limit operational flexibility. If the source used for dilution were compromised, then production would need to be stopped from both wells. Water can also be trucked in for blending purposes when a low nitrate source is unavailable locally ; however, hauling water for blending purposes is a temporary solution . Based on the recent AWWA survey, 51.8% (29/56) of respondents with nitrate impacted sources selected blending as the Weir & Roberson option to address nitrate contamination (Weir & Roberson 2010 ; Likewise, 2011). 6 treatment” option; however, in practice, For the purposes of this discussion, blending has been categorized as a “non - blending is somet imes referred to as “treatment.” Treatment options throughout this report refer to treatment technologies - effectively address available for the removal or reduction of nitrate in drinking water. Blending can sometimes be used to cost the nitrate problem through dilution, but has been categorized separately from treatment options for simplicity. 16 Technical Report 6: Drinking Water Treatment for Nitrate

33 nitrate contamination of drinking water in Germany is often addressed by blending, avoiding the costs of le alternative to treatment that avoids treatment (Dördelmann 2009). When feasible, blending is a simp disposal concerns and the certification re 2005). However, quirements of treatment (WA DOH disadvantages include the capital investment for accessing an alternative source and monitoring pply of compliant water (WA DOH 2005). Analysis of water requirements to ensure consis tent su systems utilizing blending to address nitrate contamination in the Tulare Lake Basin and Salinas Valley, tems . as well as across California, is included below in the Section 5.2 Survey of Blending and Treating Sys 17 Technical Report 6: Drinking Water Treatment for Nitrate

34 3 Treatment Options for Nitrate Contaminated Potable Water IX, RO , and ED/EDR transfer nitrate ions from water to a concentrated waste stream that requires ) lists these three processes as tion Agency ( U . S . EPA disposal. The United States Environmental Protec U 2010). In contrast, through accepted potable water treatment met hods for nitrate removal ( EPA . S . biological and chemical denitrification, nitrate is converted to reduced nitrogen species, rather than displaced to a concentrated waste stream that requires disposal. A survey of nitrate treatment systems was conducted to assess the current state of nitrate treatment. f utilities with The list of surveyed utilities was developed with the intention of covering a range o respect to geographic location, treatment type, population size , and residual handling techniques ( Table - scale ). Detailed case studies have been compiled for each of the treatment technologies where full 1 ties have been in operation or are moving ahead with design. This survey was conducted through facili collaboration with Jacobs Engineering in the completion of the associated assessment of nitrate treatment alternatives for AWWA and is complemented by a paralle l survey of nitrate treatment systems in California. Details from the initial survey are included as examples following a discussion of each of the treatment technologies. Details of the complementary survey of California systems are included in the seco nd half of this report. A brief comparison of fundamental design considerations, and advantages and disadvantages of the treatment options examined herein is listed in Table 2 . It is important to note that the contents of Table are not intended to provide a comprehensive set of criteria for treatment options. Other important 2 criteria in determining the best treatment option, which are site specific and cannot be broadly O&M costs, system size (capacity), and system footprint. generalized, include capital and IX is the most commonly used nitrate treatment method, with full - scale systems in use throughout the - United States. Full scale application of biological denitrification in potable water treatment is mainly limited to Europe and chemical denitrification methods have been investigated only at the pilot - scale. Others have provided thorough reviews of available nitrate treatment tech nologies (Kapoor & 2001); however, a recent comprehensive review of 1997; Soares 2000; Shrimali & Singh Viraraghavan the state of nitrate treatment is absent from the literature. 18 Technical Report 6: Drinking Water Treatment for Nitrate

35 1 . Utilities i ncluded in the c ase s tudies. Table rate Avg. Influent Nit Location Capacity (gpm) Case # Treatment Type - mg/L as NO (mg/L as N) 3 Ion Exchange – 1 Conventional ion exchange with blending California 400 31 – 53 (7 12) 2 Conventional ion exchange with blending California 400 ~45 (~10) ® 71 3 Counter Current Ion Exchange (MIEX 16) ) Indian Hills, CO 50 53 – – (12 4 Multiple vessel ion exchange California 500 – 900 35 – 89 (8 – 20) 40 5 Multiple vessel ion exchange Chino, CA 5000 – 200 (9 – 45) Reverse Osmosis 6 Reverse osmosis and blending Bakersfield, CA 120 75 – 84 (17 – 19) Reverse osmosis, e xploring biological – 20) 7 Brighton, CO 4600 49 – 89 (11 reduction 89 (10 8 Arlington Desalter, Riverside, CA 4583 44 – Reverse osmosis and blending – 20) Combined Reverse Osmosis and Ion Exchange Reverse osmosis, ion exchange and blending Chino Desal ter I, Chino, CA 4940 (RO), 3400 (IX) 148 9 – 303 (33 – 68) 10 Reverse osmosis, ion exchange and blending Chino Desalter II, Mira Loma, CA 4167 (RO), 2778 (IX) 70 – 224 (16 – 51) Electrodialysis /Electrodialysis Reversal/Selective Electrodialysis rodialysis Reversal (EDR) ~80 (~18) Spain 11 Elect 3,260 (each, 2 systems) – (SED) Israel 310 84 Selective Electrodialysis 89 (19 – 20) 12 Biological Denitrification Implementing fluidized bed biological 13 ) 5 Rialto, CA 2000 – 4000 17 – 19 ( ~4 – reduction 20) – 14 Implementing fixed bed biological reduction Riverside, CA 1670 44 – 89 (10 19 Technical Report 6: Drinking Water Treatment for Nitrate

36 m 2 . Potable w ater t reatment o ptions for n itrate Table anagement (adapted from WA DOH 2005). Reverse Osmosis Electrodialysis Ion Exchange Biologica l Denitrification Chemical Denitrification No Yes Full - scale Systems Yes Yes Yes Removal to waste stream Removal to waste stream Removal to waste stream Treatment Type Biological reduction Chemical reduction lfate, iron, manganese, total Turbidity, iron, manganese, Turbidity, iron, manganese, Su Common Water TSS, hydrogen sulfide, suspended solids (TSS), metals Temperature and pH, anoxic SDI, particle size, TSS, Quality Design Temperature and pH (e.g., arsenic), hardness, conditions hardness, organic matter, hardness, metals (e.g., Considerations organic matter metals (e.g., arsenic) arsenic) pH adjustment, nutrient and Pretreatment Pre Pre - filter, address hardness Pre - filter, address hardness pH adjustment - filter, address hardness substrate addition, need for Needs an oxic conditions pH adjustment, iron Post treatment - Filtration, disinfection, possible pH adjustment pH adjustment pH adjustment removal, potential ammonia substrate adsorption Remineralization Needs Remineralization control Waste/Residuals Sludge/biosolids Waste media, Iron sludge Waste brine Concentrate Concentrate Management Initial plant startup: Days to weeks Minutes Minutes Minutes - Minutes up Time Start After reaching steady state: Minutes Conventional (97%) Water Recovery scale - Not demonstrated full Up to 85% Up to 95% Nearly 100% Up to 99.9%) Low brine ( No waste brine or concentrate, No waste brine or Multiple contaminant nitrate reduction rather than concentrate, nitrat e ter Nitrate selective resins, Multiple contaminant removal, higher wa reduction rather than transfer to a waste stream, high recovery Advantages removal, desalination (TDS common application, transfer to a waste stream, water recovery, and potential multiple contaminant removal removal) (less waste), desalination, for multiple contaminant and potential for multiple unaffected by silica contaminant removal removal Substrate addition, potentially Inconsistency of nitrate more complex, high monitoring reduction, risk of nitrite Potential for nitrate peaking, needs, possible sensitivity t formation (potential o Membrane fouling and scaling, lower water recovery, high chemical use (salt), brine Energy demands, i ncomplete denitrification), environmental conditions, risk of operational complexity, Disadvantages waste disposal, potential for operational complexity, nitrite formation (potential reduction to ammonia, lack energy demands, waste of full disinfection byproduct (DBP) - scale systems, pH and waste disposal incomplete denitrification), disposal - temperature dependence, formation (e.g., NDM A) post treatment to address log - turbidity standards and 4 possible need for iron removal virus removal (state dependent) 20 Technical Report 6: Drinking Water Treatment for Nitrate

37 3.1 Ion Exchange (IX) As the most commonly used method for the removal of nitrate in potable water treatment, IX h as been widely researched, with numerous full - scale installations in operation. With the potential for multiple contaminant removal, IX can also be used to address other water quality concerns including arsenic, perchlorate, selenium, chromium (total and chromium - 6), and uraniu m (AWWA 1990 ; Boodoo 2004). Selected IX installations used for nitrate treatment in the United States are listed in Table 3 . Table 3 . emoval. Selection of f ull - scale i on e xchange i nstallations for n itrate r Year Influent nitrate Capacity Locations Reference - ) (MGD) Installed (mg/L as NO 3 Ellsworth, MN 1994 - 0.047 MN Dept. of Ag. (N.D.) 1995 - 0.047 MN Dept. of Ag. (N.D.) Clear Lake, MN 1998 Adrian, MN - 0.129 MN Dept. of Ag. (N.D.) Edgerton, MN 2002 - 0.137 MN Dept. of Ag. (N.D.) Contaminant Removal News (2007) McCook, NE 2006 Up to 125 6.8 McFarland, CA, Guter (1995), See also Guter (1982) 60 1 1983 Well 2 McFarland, CA, 1987 64 1 Guter (1995), See als o Guter (1982) Well 4 La Crescenta, CA 1987 70 – 100 2.7 Guter (1995) Grover City, CA, Guter (1995) - 80 – 130 2.3 3 wells Up to 55 Des Moines, IA 1992 Des Moines Water Works, Rash (1992) 10 Glendale, AZ: full (2010), See also Clifford et - Meyer et al. Spiked up to 177 10 2010 scale pilot al. (1987) Indian Hills, CO 2009 53 – 0.072 (design) 71 See Case Study Due to its common application, the investigation of IX for the removal of nitrate is prevalent in the literature (Yoon et al. 2001; Chabani et al. 2006; Samatya et al. 2006; Clifford 2007; Meyer et al. 2010; Clifford et al. 2010). Kapoor & Viraraghavan (1997) provide an extensive review of IX research up to 1997. Modifications of conventional IX have led to the emergence of more efficient IX processes i ncluding multiple vessel configurations, counter current configurations, the use of specialized resins, improved hydraulics, and weak base anion exchange (WBA IX). 3.1.1 Conventional Ion Exchange Figure 2 , raw Conventional IX utilizes a strong base anion (SBA) exchang e resin. In accordance with water passes through pretreatment to remove suspended solids and to address other constituents ge vessel. capable of fouling the resin. The nitrate laden treatment stream then enters the ion exchan Upon contacting the resin, nitrate displaces chloride at surface sites, removing nitrate from the water. 21 Technical Report 6: Drinking Water Treatment for Nitrate

38 . Conventional ion exchange schematic. Figure 2 2+ the divalent cations of hard water (Mg and This technique is similar to a water softener, which replaces 2+ + Ca ) with a monovalent cation (Na ). Eqn. 1 depicts the transfer of ions, with R as the resin surface site. - treatment for Leaving nitrate behind, treated water exits the ion exchange vessel and passes on to post stabilization and disinfection. - - -  R R NO Cl + NO + Cl (Eqn. 1) - 3 3 To prevent nitrate breakthrough, regeneration is necessary when the resin is exhausted of chloride ions ackwashed with a high salt (chloride has been displaced at the majority of surface sites). The media is b solution (0.5 – 3 M, Clifford 2007) to reverse the process, resulting in a brine waste stream high in nitrate and other concentrated ions (Eqn. 2). - - (Eqn. 2) R - NO + Cl  R - Cl + NO 3 3 2 - - - - The relative affinity of common anions for > NO conventional anion exchange resin is SO > Cl > HCO 3 3 4 (Bae et al. 2002 ; Clifford et al. 2010). If generic resins are not regenerated soon enough, sulfate displacement of nitrate in the resin can lead to nitrate release from the resin to the tre atment stream , chromatographic peaking and is further or (Eqn. 3). This is known as nitrate dumping, nitrate peaking discussed below. 2 - - (Eqn. 3) 2R - NO + SO + 2NO  R SO - 2 4 4 3 3 22 Technical Report 6: Drinking Water Treatment for Nitrate

39 ins, nitrate selective resins Due to the stronger affinity of the sulfate ion for generic anion exchange res - 2 - - - have been developed for which the order of affinity is NO Guter > Cl (Guter > HCO ; > SO 1982 3 3 4 7 8 the exchange capacity and selectivity coefficient 1995). Important factors in resin selection are of the 9 resin and the rat e of ion transfer (kinetics ). . Case Studies 3.1.7 Ion Exchange Detailed case studies of conventional IX plants are included in Section - Design Considerations - 3.1.2 Ion Exchange assist with IX system design including Dow’s CADIX (Computer Assisted Various tools are available to Design for Ion Exchange) (Dow 2010b) and Lenntech’s Ion Exchange calculator (Lenntech 2009b.) Table of conventional IX to nitrate removal from 4 summarizes key design considerations in the application potable water. 7 Exchange capacity: The exchange capacity is a measure of how many ions the resin can capture per unit volume. 8 Selectivity coefficient: The selectivity coeffic ient of a resin refers to the relative affinity of resin surface sites for a particular ion, in this case nitrate. 9 Kinetics: The term kinetics refers to the rate of a reaction. The rate that nitrate displaces chloride on the resin is impor tant for effic ient treatment and can be affected by competing ions. 23 Technical Report 6: Drinking Water Treatment for Nitrate

40 Table . Summary of d esign c onsiderations for c onventional IX. 4  Generic SBA resins for maximum exchange capacity (for low sulfate) o Less expensive than nitra te selective resins o - Less frequent regeneration due to higher capacity (in the absence of co contaminants) o Nitrate dumping potential Resin Selection  Nitrate selective resins to avoid nitrate dumping (for high sulfate) o More expensive than generic resins Longer bed life o More nitrate removed per unit of waste brine o and organic matter to prevent resin  Filtration to remove iron, manganese, TSS , fouling Pretreatment Water softening (anti - scalant, acid or water softener) to prevent scaling  , 10 Dechlorination to prevent resin oxidati on  11  Chloride : alkalinity ratio and dezincification 12 - Treatment Post Chloride : sulfate ratio and galvanic corrosion   Potential pH adjustment and restoration of buffering capacity to avoid corrosion  pH adjustment (caustic soda or soda ash) Chemical Usage  Regenerant brine, salt consumption  Frequency of regeneration depends on water quality and resin type  Fresh brine preparation and waste disposal O&M 2010c) Dow ;  Resin loss and replacement: 3 – 8 year lifetime (WA DOH 2005  of nitrate levels Continuous or frequent monitoring Backwashing to dislodge solids  Fixed bed versus continuous regeneration  Key system configuration parameters are system flow rate, bed swelling, bed depth,  System backwash flow rate, and rinse requirements Components o Vessels in parallel o r in - series Co - o - current regeneration current or counter  Significant cost of waste brine disposal is of greatest concern for inland systems Close proximity to coastal waters is beneficial for brine disposal  Waste - an include sewer or septic system, drying beds, trucking off  Management options c Management site, coastal pipeline, deep well injection , and advanced treatment and Disposal  Disposal options can be limited by waste brine water quality (e.g., volume, salinity, metals , and radionuclides)  of recycling and treatment of waste brine is desirable Optimization Need to manage resin fouling  o Hardness, iron, manganese, suspended solids, organic matter, and chlorine Limitations Competing ions (especially sulfate)   Disposal of waste brine  Possible role of resin re siduals in DBP formation 10 The resin can be degraded by oxidation; the functional amine groups on the resin surface are susceptible to oxidation which ad to diminished capacity (Dow 2010d). can le 11 As nitrate and other anions d isplace chloride on the resin, chloride is released to the product water, leading to the potential for taste issues and dezincification (Kapoor & Viraraghavan 1997). Dezincification refers to the ability of product water to dissolve zinc from brass and is dependent on the ratio of chloride to alkalinity (> 0.5 can be problematic). By restoring alkalinity, the dezincification potential can be minimized. 12 Galvanic corrosion can result in the release of lead from brass and galvanized solder - copper connectio ns and is associated ). ; Edwards & Triantafyllidou 2007 1999 with a high ratio of chloride to sulfate (> 0.58 can be problema tic) (Edwards et al. 24 Technical Report 6: Drinking Water Treatment for Nitrate

41 Water Quality Raw water quality is a key factor in the efficiency of an IX system, impacting resin selection, pretreatment and post treatment needs, regeneration efficiency, chemical usage, and waste disposal. - lity parameters include alkalinity, hardness, iron, manganese , and potential Important water qua . competing ions (predominantly sulfate) Selection of the appropriate resin for a given system depends directly on source water quality. In the presence of high levels of co - con taminants, nitrate selective resins may be necessary rather than generic resins. Both strong base anion exchange (SBA) resin and weak base anion exchange (WBA) resin can be suitable for nitrate removal from potable water. The latter will be addressed sep arately. The two standard types of SBA resins deviate in their functional groups. Anion exchange is dependent on the trimethylamine groups of the SBA Type I resin and the dimethylethanolamine groups of the SBA Type II resin (Helfferich 1995). Nitrate sel ective resin was invented in the earl y 1980’s by Gerald Guter (Guter 1982, see related patent: Guter 1983). Clifford & Weber (1978 and 1983) contributed to the development and characterization of these resins with their research on res in selectivity (Clif ford et al. 2010). Nitrate selective resins rely on different functional groups than Type I and Type II SBA resins. Nitrate selectivity is attributed to the and t , increased hydrophobicity and site spacing of exchange sites due to the triethyl, tripropyl ributyl functional groups of nitrate se lective resin (Clifford & Weber 1978; Guter 1982; Clifford et al. 2010). The use of nitrate selective resin avoids the problem of nitrate peaking (i.e., nitrate dumping or y the greater affinity of generic resins for sulfate. chromatographic peaking), typically caused b It is important to note the distinction between nitrate peaking and nitrate breakthrough. As nitrate displaces chloride on the resin, the nitrate capacity of the resin is gradually exhausted leading t o increasing effluent nitrate levels until influent levels are reached. This is known as breakthrough and can occur regardless of resin type. Nitrate peaking can also occur upon exhaustion of the resin capacity for nitrate. However, with nitrate peaking , the nitrate on the resin is displaced by sulfate, thereby increasing the effluent nitrate concentration to levels above that of the raw water (Clifford et al. 2010). The peak in nitrate concentration is due to the combination of the influent nitrate ion s and the nitrate ions that are coming off of the resin via sulfate displacement. Under low sulfate conditions, the use of generic SBA resins is preferred, due to their larger exchange capacity. As the ratio of sulfate to nitrate increases, the use of n itrate selective resins avoids possible nitrate peaking, minimizing the risk of MCL exceedance and the need for more frequent regeneration. However, with the highest nitrate selectivity, regeneration efficiency can decrease, increasing chemical use. With a stronger affinity for nitrate, the removal of nitrate from the resin in regeneration is more difficult because nitr ate is more strongly bound (Dow 2010c). Nitrate selective resins are more costly than the generic option (Water Quality Products 2003), b ut under appropriate conditions the use of regenerant can be minimized and bed life can be increased significantly with their use. Different system configurations have been implemented as an alternative to the use of nitrate selective resins to address , for more 2010) et al. th e problem of co - contaminants and the risk of nitrate peaking (Clifford 25 Technical Report 6: Drinking Water Treatment for Nitrate

42 information, ee System Layout and Site Considerations , below . In the consideration of IX for multiple s resin depends in part on the type and contaminant removal, the selection of the most appropriate concentration of co - contaminants. Modeling and column or pilot testing is important to determine appropriate design parameters and to design treatment based on a full life cycle analysis of costs. Upstream of the IX vessels, pretreatment of the source water may be necessary to avoid resin fouling ( 2003 ; WA DOH Water Quality Products 2005 ). Potential constituents of concern include organic matter, turbidity, total suspended solids (TSS), sand and metals, primar ily iro n and manganese ( Kapoor & Viraraghavan filtration is typically used to - 1997 ; Water Quality Products 2003; WA DOH 2005 ). Pre remove these constituents; however, additional pretreatment methods may be used. For example, coagulation/flocculation and filtrat ion may be necessary for surface waters. Pretreatment may be needed for waters with a total concentration of metals (e.g., iron and manganese) above 0.1 mg/L (Ten , States Standards 2007). Hard, alkaline water can lead to resin scaling magnesium due to calcium and accumulation; pH adjustment or water softening can be used to prevent resin scaling (Water Quality Products 2003). Created by Wolfgang Holl, the carbon dioxide regeneration of ion exchange (CARIX) 1995). The CARIX process enables simultaneous process combines anion and cation exchange (Holl removal of cations (for hardness reduction) and anions through the combination of a weak acid cation exchanger and a strong base anion exchanger. Resin exposure to disinfectants (chlorine and chloramines) s hould be avoided to prevent resin oxidation and the possible release of disinfection byproducts, specifica lly nitrosamines (Kemper et al. 2009). - Disinfection byproducts (DBPs) are potentially cancer causing compounds that can be formed through a reaction of disinfection chemicals like chlorine and chloramines with organic matter. Due to the amine functional groups of anion exchange resins, “these resins may serve as precursors for nitrogenous disinfection byproducts, such as nitrosamines, nitramines, and lonitromethanes” (Kemper et al. 2009 , ha p. 466 ). Recent research suggests that the risk of DBP formation is higher with the use of new resin; however, precursors can be a problem with downstream chloramine use or with upstream d isinfection ( s ee Kemper et al. 2009 in Table A.1 of the Appendix). Magnetic ion exchange resin is an exception as its primary purpose is to remove organic carbon and limi t DBP formation (Boyer & Singer 2006) ( s ee the ® 3.1.7 Ion Exchange MIEX - process in S ection ). Case Studies IX can reduce alkalinity due to removal of bicarbonate. To prevent corrosion in downstream and pH 13 . pipes, the product water pH and buffering capacity may need to be increased Soda ash can be added to the regenerant brine to load a portion of the resin sites with bicarbonate rather than chloride. This can restore some alkalinity to the water as bicarbonate is released from the resin when displaced by nitrate and other anions in the treatment stream (Wa ter Quality Products 2003). To minimize the need treatment, an upstream atmospheric degasifier for carbon dioxide removal - for caustic addition in post added during pretreatment (Dow 2010). can be 13 acity or the It is important to note the relationship between alkalinity and pH. Alkalinity is a measure of buffering cap resistance to changes in pH. Demineralized water or water with a low buffering capacity will be susceptible to more dramatic of pH changes and is considered unstable. The pH of acidic product water should be adjusted and the buffering capacity demineralized product water should be restored to avoid corrosion downstream and the potential for lead and copper challenges. 26 Technical Report 6: Drinking Water Treatment for Nitrate

43 product water, As nitrate and other anions displace chloride on the resin, chloride is released to the leading to the potential for taste issues, dezincification, and galvanic corrosion. Dezincification refers to the ability of product water to dissolve zinc from brass and is dependent on the ratio of chloride to alkalinity (as CaCO can be problematic, according to Kapoor & Viraraghavan (1997)). By restoring ) (> 0.5 3 alkalinity, the dezincification potential can be minimized. Galvanic corrosion can result in the release of lead from brass and galvanized solder - copper connections and is assoc iated with a high ratio of chloride 1999 to sulfate (> 0.58 can be problema tic) ( Edwards et al. ). It is ; Ed wards & Triantafyllidou 2007 important to consider the potential downstream impact of subtle water quality changes caused by treatment. System Layou t and Site Considerations The IX system can be operated using a fixed bed or as a continuous system. Details of modifications to conventional fixed bed systems are provided below. Key parameters in vessel sizing and system configuration are system flow r ate, bed swelling, bed depth, backwash flow ra te, and rinse requirements current configuration. Vessels (Dow 2010). Regeneration can be designed in a co - current or counter - - can be operated in parallel or in iency per regeneration series for redundancy, to maximize removal effic cycle, to address nitrate peaking and to consistently produce water with limited variation in water ity parameters (Clifford et al. 2010). qual Residuals Management and Disposal Management of waste brine can be costly. Options incl ude discharge to a sewer or septic system, waste volume reduction using drying beds, trucking to an off - site approved disposal location, ocean discharge through a coastal pipeline, deep well injection, and advanced treatment. Water quality characteristics of the waste brine (e.g., volume, salinity, metals , and radionuclides) can affect the feasibility and costs of disposal options. Proximity and access to coastal waters can be a significant factor in determining the isposal is of greatest concern to inland communities. Although burden of brine disposal. Generally, d other removal technologies (RO and ED) also require concentrate disposal, this is an issue of particular concern with IX. Because IX requires the addition of salt for resin regeneration, the waste stream consists of not only the nitrate and other ions that have been removed from the water, but also the spent brine solution used in regeneration. The high cost of nitrate laden brine disposal has led to research into optimization of recycling an d treatment of this waste stream. Partial regeneration and reuse of treated brine for multiple regeneration cycles can minimize the volume of waste while maximizing re generation efficiency (Clifford ; 2007 Cliffor d et al. 2010). Application of IX systems coupled with biological, chemical, or catalytic denitrification enables removal of nitrate from waste brine, with reduction to nitrogen gas. The electrochemical destruction of nitrate in waste brine is also 3.6 Brine being explored. Several combined configurations of interest are discussed below in Section Treatment Alternatives and Hybrid Treatment Systems . 27 Technical Report 6: Drinking Water Treatment for Nitrate

44 Maintenance, Monitoring and Operational Complexity , ity and Regeneration frequency will depend on pretreatment measures, water qual the type of resin , 2005). The typical amount of brine waste compared to water produced can range from used (WA DOH ~3.0% for conventional systems (Clifford 2007) to ~0.5% for low brine systems (Calgon Carbon Corporation 2007, discussed below). A constant supply of brine must be available for resin regeneration and waste brine requires appropriate storage and disposal. Resin loss can be controlled by adjusting the backwash flow rate and adding screens (Keller 2000). Resin life will also depend on water quality and pretreatment measures; resin replacement may be required after 3 – 8 years (WA DOH 2005 ; Dow 2010c). To ensure the production of compliant water, continuous or frequent monitoring of nitrate levels is necessary. In addition to resin re generation, backwashing is used to dislodge solids and “resin fines” (Dow 2010). In comparison with alternative treatment options, IX requires limited O&M with high feasibility of automation and low operational complexity. Cost Con - siderations 3.1.3 Ion Exchange For optimal operation of an IX system, the fundamental objective is to maximize regeneration efficiency, while meeting necessary potable water guidelines. Factors affecting system cost include facility size ncluding nitrate concentration), environmental factors (temperature), (flow rate), source water quality (i and target effluent nitrate concentration. Disposal of waste brine is commonly a significant portion of O&M costs. equipment, vessels, resin, Capital costs for IX include land, housing, piping, storage tanks, O&M preliminary testing (pilot studies), permits, and training. O&M costs include resin replacement (due to - scalant, pH loss or degradation), resin disposal, brine disposal or treatment, chemical use (salt, anti and maintenance, power , adjustment), repair and labor. Published cost information from existing IX installations is listed in Table 5 . Costs have been adjusted to 2010 dollars, unless indicated otherwise. Costs can be difficult to assess due to inconsistencies in how cost information is reported. Comparison of IX costs is not always valid due to differences in influent water quality parameters, system size, waste management options, and system configuration. Published costs do not always include comparable information. It would be inappropriate to compare the O&M costs of a facility that excludes disposal costs with others that include this information. The listed cost information is provided as an approximate range of costs for specific facilities. Costs for implementing IX may be very different from those listed here. A thorough cost analysis of design parameters for specific locations would be required for accurate cost estimation. The information gathered through the questi onnaire includes detailed costs associated with the unique case studies included in this analysis. A detailed discussion of treatment costs is included below in Section 6 Treatment Cost An . alysis 28 Technical Report 6: Drinking Water Treatment for Nitrate

45 Table . Selected p ublished c osts* of i on e xchange s ystems for n 5 r emoval . itrate System Flow** < 0.5 MGD 0.5 – 5 MGD 5+ MGD Annualized Capital Cost ($/1000 gal) 0.37 – 1.21 [1] 0.28 – 0.94 [2, 3] 0.28 – 0.61 [3, 4, 5] – O&M Cost ($/1000 gal) 0.60 – 4.65 [1] 0.46 1.25 [2, 3] 0.37 – 0.87 [3, 4, 5] 1.44 [3, 4, 5] – 2.19 [2, 3] Total Annualized Cost ($/1000 gal) 0.97 – 5.71 [1] 0.74 – 0.65 *Costs have been adjusted to 2010 dollars with 7% interest over 20 years, unless indicated otherwise (below). , costs are based on actual system flow rather than design capacity. **When available [1] Minnesota Department of Agriculture (N.D.), not adjusted to 2010 dollars, 20 year amortization without . [5] Drewry (2010). interest. [2] Guter (1995). [3] Conlon et al. (1995). [4] Meyer et al. (2010) 3.1.4 Ion Exchange Selected Research - A large body of research has focused on IX. Table A.1 of the Appendix is a list of recent studies relevant - to nitrate removal from potable water. Given the many years of successful full e operating scal experience with IX, current applied research is focused on brine recycling efficiency, the optimization of waste management, and improvements in resin capacity and selectivity, to improve efficiency and reduce costs. - 3.1.5 Ion Exchange y of Advantages and Disadvantages Summar A summary of advantages and disadvantages of IX in comparison with other treatment options is listed in Table A.6 of the Appendix. Significant advantages of IX include years of industry experience, multiple contaminant re moval, selective nitrate removal, financial feasibility, use in small and large systems, and the ability to automate. Disadvantages include the costly disposal of waste brine, the potential for nitrate dumping and resin fouling, the possible need for pH a djustment, the potential for hazardous waste generation (i.e. brine with traces of co - contaminants like arsenic and chromium), and the , in DBP formation (Kemper et al. 2009). possible role of resin residuals 3.1.6 Modifications to Conventional Ion Exchange Modifications of conventional IX have led to the emergence of low - brine IX processes including ® magnetic ion exchange (MIEX ), ion exchange separation (ISEP®), Envirogen (formerly Basin Water) systems, and weak base ion exchange (WBA IX). Despite their po tential advantages, it is important to note that proprietary technologies may have inherent disadvantages, such as a lack of flexibility to use better technology when it becomes available, vulnerability if the manufacturer goes out of business or discontin ues supporting the product, and a lack of competition to keep O&M costs down. Counter Current Flow with Specialized Resin ® Magnetic ion exchange technology (MIEX ), developed by Orica Watercare, offers a low brine alternative ® Figure process ( ) differs from to conventional IX, using a uni que SBA Type I resin. The MIEX 3 29 Technical Report 6: Drinking Water Treatment for Nitrate

46 conventional IX in that the resin is fluidized in a contactor with spent resin removed from the contactor for regeneration outside of the process wat er stream and then returned to the contactor. This is in ® contrast to conventional IX, where the resin is stationary. Unlike conventional IX, the MIEX fluidized bed process is tolerant of suspended solids and low levels of oxidants. reprinted with permission, Source: Orica Figure 3 . ( Process flow diagram for counter current MIEX® process . b Watercare 2008 . ) The minimization of waste brine is accomplished through frequent batch regeneration. The magnetized resin encourages aggl omeration of loaded resin particles, resulting in faster settling. Loaded resin is removed from the bottom of the IX vessel and is passed to regeneration, while regenerated resin is added at the top of the IX vessel. This configuration removes the risk o f nitrate peaking because clean resin, added at the end point of the system, captures any displaced nitrate, while competing ions, such ® process has been proven to effectively as sulfate can be removed early on in the resin vessel. The MIEX address variou s water quality concerns including nitrate, organic carbon, arsenic, chromium - 6 and 2005; perchlorate ( Seidel et al. 2004; Humbert et al. Boyer & Singer 2006; and Watercare 2008 ). ® A detailed case study of a MIEX treatment plant in Indian Hills, CO is in cluded in S ection 3.1.7 Ion Case Studies - . Exchange 30 Technical Report 6: Drinking Water Treatment for Nitrate

47 Improved Hydraulics and Nitrate Selective Resins The Layne Christensen Company and Rohm and Haas offer their Advanced Amberpack® system which utilizes nitrate se lective resins and their patented Fractal Distribution Technology to increase the treated water volume and decrease the waste brine. Nitrate selective resins have a greater affinity for nitrate - selective resins , resulting in an improvement in rem oval efficiency, especially in waters where than non the sulfate to nitrate ratio is greater than one . The distributor system is designed to provide uniform flow through the ion exchange vessels in both treatment and regeneration modes. As a result of the unifor , the exchange capacity of the resin m flow can be maximized while the brine and rinse values can be minimized, thus increasing the water efficiency of th e system (Rohm and Haas Company 2007). Multiple Vessel Carousel Configuration ® Calgon Carbon’s Continu ous Ion Exchange Separation System (ISEP System) utilizes a carousel configuration. This configuration has the potential to avoid downtime for regeneration, minimize the amount of resin needed, and maximize regeneration efficiency. Illustrated in Figure 4 , multiple resin vessels are rotated from active treatment to resin regeneration and rinsing and back to active - The vessels rotate in the opposite d irection of the water movement ( treatment. 5 ). The counter Figure and counter current system can provide consistent product water, operating uninterrupted with flow - four - flow treatment. This configuration results in flow regeneration and down zones within the - up ® system (Calgon Car bon Corporation ISEP for Nitrate Removal Brochure 200 3). In the Adsorption Zone, for single pass flow. Nitrate and other anions are the feed stream is passed through 14 ports in parallel removed ed water is the Displacement Zone, soften In from the feed water as they transfer to the resin. - up in the regeneration zone. In the used to displace the hard feed water to prevent scale build Regeneration Zone, a combination of brine and rinse is directed through the beds for a true counter - current regeneration to maximize the regenerat ion efficiency. In the Rinse Zone, a small amount of softened feed water is used to prevent any of the salt from the regeneration zone from reaching the product water. ection 3.1.7 Ion Exchange A detailed case study of an ISEP® treatment plant in Chino, CA is included in S - Case Studies . 31 Technical Report 6: Drinking Water Treatment for Nitrate

48 3. 200 ) Figure 4 . Vessel rotation in Calgon Carbon countercurrent ISEP® sy stem. (Source: Calgon Carbon Corporation Example of flow through the ISEP® system . 3 200 ) Figure 5 . Calgon Carbon Corporation . ( Source: 32 Technical Report 6: Drinking Water Treatment for Nitrate

49 Multiple Vessel Staggered Configuration systems using multiple beds operated ion exchange ( IX Envirogen Technologies, Inc. offers proprietary ) in a staggered design ( Figure 6 ). Such a configuration has the potential to maximize resin capacity and minimize waste and chemical use. reprinted with 6 . Example of an Envirogen multiple bed proprietary a Figure . ( Sou rce: nion exchange system 2009 . ) permission, Envirogen Envirogen’s low brine IX system has been successfully implemented for nitrate and uranium removal in 60 ernardino County, CA (Envirogen 2010). Delivering 2 MGD with nitrate levels reduced from 50 – San B mg/ L as nitrate (11 – 14 mg/L as N) to 35 mg/L as nitrate (8 mg/L as N) and a system footprint of 50’ X 50’, the installation resulted in recovery of a well that had been previously decommissioned. Envirogen is contracted to handle all operation and maintena nce, including waste disposal. In Yucca Valley, CA, nitrate treatment was required due to over pumping and the resulting intrusion of septic system contaminated waters. An Envirogen IX system was installed to provide 2.8 MGD. With treatment, nitrate co ncentrations are decreased from 58 mg/L as nitrate (13 mg/L as N) to 20 mg/L as nitrate (4.5 mg/L as N) with 50% blending and a 0.3% waste rate (0.15% net) (Envirogen 2002 ). ection n virogen low - brine IX treatment plant in California i s included in S A detailed case study of an E . - Case Studies 3.1.7 Ion Exchange 33 Technical Report 6: Drinking Water Treatment for Nitrate

50 Weak Base Anion Exchange (WBA IX) Weak base anion exchange (WBA IX), an emerging technology, can be an effective option for nitrate removal from potable wate r. Highly pH dependent, nitrate removal using WBA IX is governed by Eqn. 5 (Applied Re search Associates N.D.). While SBA IX resin can remove nitrate by splitting nitrate salts, WBA N.D. ). First, acid addi tion protonates the WBA resin IX resin removes strong acids (Remco Engineering (Eqn. 4). Next, the positively charged resin sites remove anions, like nitrate, from the treatment stream (Eqn. 5). + + NH (Eqn. 4) + H R  R - NH - 3 2 - + R - + NO  R - NH - NO (Eqn. 5) NH 3 3 3 3 g the resin, in accordance with Eqn. 6. Rather than the high salt Resin regeneration occurs by neutralizin solution used to regenerate SBA resins, weak bases are used to neutralize the WBA resin. + - - + NO + Na - OH NH  R - NH (Eqn. 6) + HOH + Na - NO R 3 2 3 3 As discussed in previous sections, the use of S nitrate brine waste stream. - BA IX resin results in a high Due to the high salt content, the nitrate in the waste stream generally cannot be beneficially reused. However, with the use of alternative weak bases for regeneration (Eqns. 7 and 8), the waste s tream NO and Ca(NO does not have a high salt content and could potentially be recycled as fertilizer (NH ) ) 3 3 4 2 (Clifford 2007). R N - HNO (Eqn. 7) + NH OH  R N - HOH + NH NO 3 4 4 3 3 3 2 R N - HNO + Ca(OH)  2R N - HOH + Ca(NO ) (Eqn. 8) 3 2 3 2 3 3 The use of WBA resins is more operationally complex than the use of SBA resins. Chemical use includes , and pH adjustments are more significant. acids and bases, the system is susceptible to corrosion duct water pH Adjustment of influent pH to between 3 and 6 is necessary, with subsequent pro adjustment as well (Clifford 2007). WBA resins can also be more sensitive to temperature, with one o C (Dow 2010), but this should not impact their use with resin having a maximum operating level of 95 municipal drinking water treatment. Rege neration of WBA resin is more efficient than that of the SBA resin of conventional IX; regenerant waste volumes are minimized and waste products can sometimes ecycled as fertilizer (Clifford 2007 and ARA & Purolite N.D.a). be r Weak Base Ion Exchange for Nitrate, or the “WIN” Process, was developed by Applied Research Associates, Inc. (ARA) and The P urolite Company (ARA & Purolite N.D.b) as a treatment option that can be less costly and more efficient, with significantly lower waste volumes, than conventio nal SBA IX ( ). 7 Figure 34 Technical Report 6: Drinking Water Treatment for Nitrate

51 Figure 7 . Process schematic of weak base ion exchange for nitrate ("WIN" process) . ( Source: reprinted with ) permission, N.D.b . ARA & Purolite As an emerging technology, available information is limited to that provided by the manufacturers. The process consists of a pretreatment step to decrease pH, followed by ion exchange vessels in series, and aining effluent cont - treatment to increase pH. The manufacturer states, “The volume of nitrate - post from the WIN process is typically less than 0.2% of the treated water and, in some cases, may be land N.D.b). As discussed above, the use of SBA IX resin results in a high as fertilizer” (ARA & Purolite applied - nitrate brine waste stream. Due to the high salt content, the nitrate in the waste stream generally cannot be beneficially reused. However, with the use of alternative weak bases for regeneration, the waste stream does not have a high salt content and could potentially be recycled as fertilizer. - Case Studies 3.1.7 Ion Exchange Case Studies Conventional Ion Exchange - scale The following case studies provide detailed information on the design and operation of full - IX is also used by the Chino Basin conventional IX treatment plants for nitrate removal. Conventional Desalter Authority in combination with RO. Detailed case studies for the Chino Desalter are listed separately in the RO case study section. 35 Technical Report 6: Drinking Water Treatment for Nitrate

52 System Location: California System Type: Communi CASE #1 ty Water System Treatment Type: Conventional Ion Exchange Startup Date: 2007 System Description Treatment Type Ion Exchange and Blending 400 gpm System Capacity A California utility is resp onsible for a system that has two groundwater supplies, one of which has Raw Water Nitrate 53 mg/L as nitrate – 31 nitrate at levels that exceed the MCL. The – 7 12 mg/L as N impacted well has a production capacity of 400 - gpm and historical nitrate concentrations ranging from 31 mg/L to 53 mg/L of nitrate as NO – 12 mg/L as N). (7 3 The utility has implemented a blending program and installed a conventional IX treatment system in 2007 for nitrate control and treatment. ed the decision to The nitrate impacted sources also have arse g/L, which influenc nic levels above the MCL of 10 u install IX. IX can simultaneously remove nitrate and arsenic. The treatment system is comprised of three pressure vessels. Two vessels contain Purolite A300E for arsenic removal and the third vessel contains the nitrate selective lite A520E resin. The system was originally installed in 2007 and was further modified in 2009. The system is Puro designed to decrease nitrate levels to less than 22 mg/L as nitrate (5 mg/L as N) prior to blending. The maximum distribution system water goal for nitrate is 35 mg/L as nitrate (8 mg/L as N). To assure the system maintains the treatment goals, online nitrate analyzers have been installed at two locations, after the IX system and after blending. Source Water Quality - g/L as N)  mg/L as NO -  Co - contaminants Nitrate (m 3 Average – 35 (8) o Arsenic (15 u g/L) o Sulfate (66 mg/L) o Minimum – 31 (7) o o 53 (12) – Maximum Treatment Technology Selection nced by The conventional IX system was selected to treat both nitrate and arsenic. The decision was further influe the ability to discharge the waste brine to a municipal sewer, a cost effective disposal option. Often technology - selection is limited by the costs of brine management. If a utility has the option to dispose of the waste brine to a r it can significantly decrease the operations and maintenance cost of the system. No other municipal sewe technologies were pilot tested prior to installation of the IX system. Operational Notes Since nitrate is an acute contaminant, the reliability of the treatment system is an utmost concern. The system is equipped with online nitrate analyzers which causes a shutdown of the treatment system when nitrate is at or eakthrough above 35 mg/L as nitrate (8 mg/L as N). Additionally, the treatment system has experienced arsenic br resulting in concentrations above the MCL. The treatment system has also experienced premature nitrate breakthrough as a result of faulty distributors in the ion exchange vessels which have since been replaced. Technical Report 6: Drinking Water Treatment for Nitrate 36

53 Treatment System Parameters  Des Bed volumes prior to regeneration  ign Capacity – 345 o 470 (approximately o 400 gpm maximum capacity 220,000 300,000 gallons –  Pretreatment treated) o None  Resin Type: Purolite A200E and Purolite Treatment system footprint  A520E nitrate selective o Treatment system: 30’ x 35’ o Building footprint: None Previous resins used: None  – Salt Consumption: 1,700 lb/week (May   Ion exchange pressure vessels week (Oct Sept) and 600 lb/ – April) Number of vessels: 3 o Diameter of vessels: 6’ o Volume of brine generated  o Height of vessels: 6’ 800 gal/vessel/backwash o 99.7% water efficiency Design Loading Rate  o 2 o 5.2 gpm/ft  Monitoring: Online nitrate analyzer o Laboratory samples o Residuals Management The waste brine is discharged to the sewer and sent to a municipal wastewater treat ment facility, with an annual cost of $12,000. There have not been any unexpected residuals that have impacted the disposal option. Technology Benefits and Drawbacks Drawbacks Benefits  Ease of regeneration  System has potential for breakthrough of nitrate or arsenic - - sewer brine disposal  Direct to  Time intensive operations Simple, manually operat  ed system  Required increase in operator certification (California T - 3) Treatment Technology Costs Capital Costs (Total with explanation or component costs) Total: $350,000 Annual O&M Costs (Total with explanation or component costs) Total: $66,500 Resin: $13,000 Brine Disposal or $12,000 Treatment: Chemicals: $5,500 Repair/Maintenance (not $4,500 including Labor): $2,500 Power: $28,000 Labor ($): 37 Technical Report 6: Drinking Water Treatment for Nitrate

54 System Location: California CASE #2 System Type: Community Water System Treatment Type: Conventional Ion Exchange Startup Date: 2006 System Description Treatment Type Ion Exchange 400 gpm System Capacity A California water utility operates a 44 mg/L as nitrate Raw Water Nitrate system that has three groundwater N 10 mg/L as supplies with varying amounts of nitrate contamination. Two of the wells require treatment as the nitrate concentration is at or above the 45 mg/L as nitrate ( 10 mg/L as N) MCL. The third well has nitrate ranging from 22 mg/L as nitrate (5 mg/L) as N to 31 mg/L as nitrate (7 mg/L as N) and does not require treatment, but is blended with the IX treated water prior to distribution. In 2006, a conventional IX sys tem was installed. The treatment system is comprised of three pressure vessels that contain the Rohm and Haas HP 555 ion exchange resin. Since the nitrate concentration in the impacted wells is typically at the MCL, but above the established water qualit y goal of 35 mg/L as nitrate (8 mg/L as N), a sidestream treatment approach is utilized. In sidestream treatment a portion of the flow is passed through the treatment ncentration and allows the system while the remainder is bypassed. The IX treatment results in very low nitrate co two streams to be combined to achieve the water quality goal. Sidestream treatment offers capital and operational costs savings. Source Water Quality –  Co - contaminants mg/L as nitrate (mg/L N)  Nitrate 44 (10) Sulfate Average – o o Treatment Technology Selection IX was selected for this system because the utility was familiar with the technology from other installations. No technologies were pilot tested prior to the installation of the IX system. Treatment System Parameters Bed volumes prior to regeneration   Design Capacity o 309 BV o 400 gpm maximum capacity Regeneration occurs every 12.4 o Pretreatment  hours o 100 mesh strainer 10 micron screen o Resin Type: Rohm and Haas HP 555   Previous resins used: None  Treatment system footprint o Treatment system: 20’ x 100’ Salt Consumption: Estimated to be 32,0 00  Residual handling : 40’ x 100’ o lbs of salt per month Total footprint: 60’ x 100’ o Volume of brine generated   Ion exchange pressure vessels 1,000 gal/vessel/backwash o o Number of vessels: 3 o 99.2% water efficiency o Diameter of vessels: 4’  Monitoring: o Height of vessels: 6’ o Online nitrate analyzer  Design Loading Rate Laboratory samples o 2 10 gpm/ft o Technical Report 6: Drinking Water Treatment for Nitrate 38

55 Residuals Management The spent brine is held in a storage tank and ultimately disposed off site . Brine volumes are minimized by using a partial flow treatment strategy where only 400 gpm is treated by the IX system while the remaining 500 gpm bypasses the system. The untreated flow is recombined with the treated water prior to entering the distrib ution system. Technology Benefits and Drawbacks Benefits Drawbacks  No onsite brine disposal  Familiarity with technology Simple, manually operated system   Design flaws  Inconsistent operations  Potential for nitrate breakthrough Treatment Technology Costs Cost information was not included with the survey response. Operational Notes This particular IX system has had a series of operational challenges, many of which can be attributed to faulty engineering. Sections of exposed Schedule 80 PVC faile d due to freezing, while other sections of pipe failed due to prolonged UV exposure. The brine reclaim tank experienced algal growth and was ultimately replaced with an - resistant tank. opaque, UV The plant has experienced shutdowns due to an exceedance of the nitrate MCL. Routine sampling showed distribution system nitrate concentrations above the MCL. The system utilizes an online nitrate analyzer to prevent MCL violation; however, a calibration error prevented the system from shutting down as progra mmed. It is believed a second failure occurred as a result of the brine saturator having a low salt concentration resulting in incomplete regeneration of the resin. It should be noted that, due to concerns of nitrosamine release (which can be common for IX systems), the effluent of each vessel and the entry point to the distribution system are tested before the system is placed in service. 39 Technical Report 6: Drinking Water Treatment for Nitrate

56 Modifications to Conventional Ion Exchange Case Studies - ion on the design and operation of full scale low - The following case studies provide detailed informat - ® ® , ISEP , and Envirogen (formerly Basin Water) systems. brine IX systems including MIEX System Name: Indian Hills System Location: Indian Hills, CO PWSID: CO 0130065 CASE #3 System Typ e: Community Water System Treatment Type: Counter Current Magnetic Ion Exchange Questionnaire completed by: Operations staff and Orica Water Care Representatives Startup Date: 2009 System Description Counter Current Ion Exchange Treatment Type 50 gpm System Capacity The Indian Hills Water District – 53 Raw Water Nitrate 71 mg/L as nitrate (District) utilizes a groundwater well 16 mg/L as N – 12 with a production capacity of approximately 50 gpm. The well has historical nitrate concentrations ran ging from 12 mg/L to 16 mg/L of nitrate as N. The District has implemented i c e xchange process developed by Orica Water Care. Unlike traditional packed bed IX, the on the urrent ounter - c ® of fluidized beds which allows for a reduction Orica process uses a Magnetic Ion Exchange (MIEX ) resin in a series in brine generation. The minimization of waste brine is accomplished through frequent batch resin regeneration. The magnetized resin ettling. Loaded resin is removed from the encourages agglomeration of loaded resin particles, resulting in faster s bottom of the IX vessel and is passed to regeneration, while regenerated resin is continuously added at the top of the IX vessel. This configuration reduces the risk of nitrate spiking/chromatographic peaking bec ause clean resin, added at the end point of the system, captures any displaced nitrate, while competing ions, such as sulfate, can be removed early on in the resin vessel. Regeneration is performed continuously in small batches. Loaded resin is passed t o regeneration tanks through the bottom of the IX vessel. The loaded resin is regenerated and then returned to the contactor vessel, thus maintaining a consistent ion exchange capacity in the contactor vessel. Numerous regenerations are performed on a da ily basis, with the actual number of regenerations depending on the system’s flow rate. Source Water Quality  Nitrate – mg/L as nitrate (mg/L N) 62 (14) o Average – Minimum – 53 (12) o o Maximum – 71 (16) contaminants - Co  None noted in survey o 40 Technical Report 6: Drinking Water Treatment for Nitrate

57 Treatment Tec hnology Selection The counter - current magnetic ion exchange system was selected based on the expected low levels of brine when ® - process was pilot tested at compared to conventional packed bed IX. Prior to full scale implementation, the MIEX Indian Hills. Indian Hills did not pilot test any other technology prior to implementation. Treatment System Parameters  Design Loading Rate  Design Capacity 2 o o 7.27 gpm/ft at design flow 50 gpm maximum capacity 25 gpm typical o  Bed volumes prior to regeneration o 14 gpm actual – 11 o 125 BV treatment - Pretreatment/Post  Resin Type: MIEX DW 1401  Strong base - 1 micron bag filters o anion exchange resin in chloride form Chlorination o  Previous resins used: None Treatmen t system footprint  Salt Consumption: 3,500 lbs/MG treated  o Treatment system: 9.75’ x 5.5’ water Regeneration system: 13.3’ x 6.5’ o  generated Volume of brine o Residuals handling system: 2,800 gal brine/MG treated water o 2,000 gal storage tank  99.7% water efficiency o o Total system footprint: 20.5’ x 9.75’  Monitoring: Building footprint: 30’ x 50’ o o Online nitrate analyzer  Number of contactors o Laboratory samples o t (2) – 3’ diameter; 6.75’ heigh o Field colorimeter Residuals Management The waste brine is sent to a 2,000 gallon underground cally pumped from the storage tank. The brine is periodi storage tank and ultimately land applied. Indian Hills is investigating deep well injection as an alternative disposal mechanism. The waste brine was analyzed in the pilot portion of the project and no unforeseen residuals were identified that would further limit the brine disposal options. While not specifically analyzed at this site, waste brine ® from other MIEX installations has not had detectable nitrosamine concentrations. ® System, Bed Volumes are define d as the volume of For the MIEX water treated per volume of resin regenerated. BV = gal water treated/gal resin regenerated. For example, if 5 gal of resin were regenerated for every 1000 gal of water treated, the regeneration rate would be: BV = 1000gal/5gal = 200B V. ® The design regeneration rate for the MIEX System is 125 BV, meaning that 8.0 gal of resin are regenerated per 1000 gal of water treated. The system regeneration rate can be adjusted through the control system. There is – 2 gallons of resin attrition per 1 MG of treatment. Small amounts of resin are periodically added approximately 1 to the regeneration system manually. 41 Technical Report 6: Drinking Water Treatment for Nitrate

58 Technology Benefits and Drawbacks Benefits Drawbacks Consistent treatment performance Generation of waste brine    Relatively low volumes of waste brine  Resin lost in the treatment must be removed prior to distribution  No nitrate dumpin g Treatment Technology Costs Capital Costs (Total with explanation or component costs) Operating and Monitoring Approximately $150,000 Equipment: (IX treatment system, including initial resin fill) O&M Costs (Total with explanation or component costs) $0.15/1000 gallons treated – Resin: $0.08 N/A Resin Disposal: N/A Brine Disposal or Treatment: Chemicals: Salt (based on an estimated salt cost of $ 100/ton): $0.15 – $0.20/1000 gallons treated * Sources August, 2010. Vaughan , F., Orica Watercare. ( 2010) Personal Communication . September, 2010. Martin, B . (2010) Completed questionnaire. * Unpublished sources used in the development of the case stud ies are not reflected in the References section of this report. 42 Technical Report 6: Drinking Water Treatment for Nitrate

59 System Location: California CASE #4 System Type: Community Water System Treatment Type: Ion Exchange, Well Modification (proposed), Well Abandonment – Startup Date: 2003 2009 System Description Treatment Type Ion Exchange and Blending A California district operates 27 900 gpm – 500 System Capacity groundwater sources, 15 of which – 35 Raw Water Max 89 mg/L as nitrate contain nitrate at or n ear the MCL. 8 Nitrate – 20 mg/L as N The district has considered and implemented a variety of solutions and well destruction. The nitrate impacted wells range in capacity from 500 gpm , including IX, well modifications to 1000 gpm with nitrate ranging from 35 – (8 89 mg/L as nitrate – 20 mg/L as N). It should be noted that a water - (< 8 mg/L of nitrate as N) has been implemented in the district, and all quality goal of 35 mg/L of nitrate as NO 3 sources are treated to below this level. itrate contaminated sources in 2003 and the most recent system was The district began actively treating the n installed in 2009. The district currently has 6 wells with active IX systems, 7 wells have been destructed or made inactive, 2 wells are being considered for well modifications , and one we ll has an enhanced control system where there will be an automatic shut down if the nitrate levels exceed a predetermined level. The range of historical nitrate concentrations of the wells is shown below. Source Water Quality mg/L as nitrate – Nitrate  (mg/L N) 15.0 to 40.4 (3.4 to 9.1) - Well I o o 7 to 58 (1.5 to 13.1) - Well A - Well J o 6 to 33.1 (1.4 to 7.5) - Well B o 15 to 48 (3.4 to 10.8) - Well K o 51 (11.5) 4 to 47 (0.9 to 10.6) - Well C o 87 (19.6) - Well L o Well D - 13 to 88.4 (2.9 to 20.0) o - 37 (8.4) o Well M Well E ND to 40.7 (ND to 9.2) - o Well N - 37.4 (8.4) o - Well F o 11 to 53.8 (2.5 to 12.1) 64 (14.5) - Well O o o Well G - 2 to 52 (0.5 to 11.7) 16.3) 72.4 (1.9 to o – 8.5 - Well H Treatment Technology Selection as in 2003. At the time of installation IX was deemed the The first nitrate treatment system the district installed w best available technology as it was the most economical with respect to the well’s nitrate concentrations and flow s that required treatment in an rates. When possible, the district elected to install similar systems on the well effort to establish operational parallels between their systems. When the footprint of the system or excessive nitrate concentrations made IX treatment infeasible, the district has elected for well destruction. In recent c ases where treatment is necessary, the district has evaluated well modifications to determine if it is feasible to reduce the nitrate concentrations without the need for active treatment. Technical Report 6: Drinking Water Treatment for Nitrate 43

60 The following section shows the typical parameters of the district ’s individual IX systems with the exception of the salt consumption which represents the total salt use for the entire district. Treatment System Parameters Bed vol   Design Capacity umes prior to regeneration 500 900 gpm o – o 300 BV Pretreatment   Treatment System Manufacturer Envirogen Technologies o None o  Treatment system footprint Resin Type: Conventional type 1 ion  Two 35’ x 20’ cemen t slabs exchange resin o o Treatment system: 35’ x 10’  Previous resins used: None Housed in a cargo container o  Salt Consumption: 25 tons/ week  Residuals handling system combined for all systems Three 12’ x 12’ dia. tanks o Volume of brine generated  Ion exchange pressure vessels  o volumes per vessel 1.3 bed Number of vessels: 16 o regenerated Diameter of vessels: 3’ o o Approximately 99.6% water Height of vessels: 6’ o efficiency Design Loading Rate  Monitoring  2 o 12 gpm/ft Online nitrate analyzer o o Laboratory samples Residuals Management The waste brine is disposed of at an offsite facility. This decision has been impacted by elevated selenium and NDMA in the waste brine. Technology Benefits and Drawbacks Drawbacks Benefits   Treatment system failure poses an acute Allows wells that would normally be offline due to nitrate contamination to be health risk to the potable water system used for potable purposes  Time intensive operations if there is a treatment disruption - The treatment system vendor provides on  ns site technical support and operatio  The systems have numerous valves and moving parts. If there is a mechanical failure it can be diffi cult to identify the source High operating and brine disposal costs   The district pays the system vendor for stand - by fees and service charges if the system is in stand - by mode during times of low water use 44 Technical Report 6: Drinking Water Treatment for Nitrate

61 Treatment Technology Costs Capital Costs (Tot al with explanation or component costs) – Total: $360,000 per unit up, and Sampling/testing Electrical, Pi ping, Set (construction costs) The district does not own the treatment plants, tanks, resins, etc. O&M Costs (Total with explanation or component cos ts) $59,239.41 per month, per unit Total: Brine Disposal or $34,016.75 per month, per unit Treatment: $3,525.00 (service fee) per month, per unit Repair/Maintenance (not including Labor): $13,541.41 (salt) per month, per unit Salt: $7,500.00 (s tand by fee) per month, per unit Other: $656.25 (tax) per month, per unit Technical Report 6: Drinking Water Treatment for Nitrate 45

62 System Name: City of Chino System Location: Chino, CA PWSID: CA3610012 CASE #5 System Type: Community Water System Treatment Type: Ion Exchange (IX) Questionnaire c ompleted by: Gilbert Aldaco, Water Utilities Supervisor, City of Chino Public Works Startup Date: 2006 System Description Ion Exchange and Blending Treatment Type The City of Chino, CA, operates 13 groundwater 5000 gpm System Capacity sources and 3 GWUDI (groundwater under direct influence of surface water) sources. All of the GWUDI and 8 of the groundwater sources 200 mg/L as nitrate – ~40 Raw Water Nitrate are impacted by nitrate contamination. IX and 45 mg/L as N – ~9 s high nitrate and blending are used to addres ® perchlorate levels. One of the wells is inoperable due to perchlorate contamination. IX using the ISEP system is th used for the treatment of 3 wells ranging in capacity from 1100 to 2300 gpm and is being considered for a 4 well. Addi tionally, an Envirogen (formerly Basin Water) IX system is used for the treatment of an 800 gpm well. Treated surface water from the Metropolitan Water District of Southern California (MWDSC) and treated groundwater ® is used for blending. The blend ratio is 3:1. The ISEP from the Chino Basin Desalter Authority system was built in 2005, with a capacity of 5000 gpm (7.2 MGD), and approved for operation in 2006. Source Water Quality contaminants mg/L as nitrate (mg/L N)  Nitrate –  Co - 40 to 100 (9 to 45) o Perchlorate o Treatment Technology Selection The costs and feasibility of several alternatives for nitrate treatment, including reverse osmosis, biological processes, and conventional (fixed bed) IX were investigated. Biological treatment and fixed bed IX were pilot ® system was selected based in part on the potential to tested prior to installation of the current system. The ISEP simultaneously address perchlorate contamination and due to the efficiency of operation. Additional Information were not required to increase their level of certification to operate the treatment plant.  Operators  When asked about the areas in which the nitrate treatment technology has exceeded expectations, the response was , aminants at varied flow rates (i.e., 1500 “Reliability and ability to effectively remove cont gpm to 5000 gpm).” incidence of a plant shutdown due to an alarm or exceedance of an MCL; however, the There has been no  plant was shut down for approximately 1 month due to theft of computer control equipment. 46 Technical Report 6: Drinking Water Treatment for Nitrate

63 Tre atment System Parameters   Design Capacity Previous resins used: None 5000 gpm o  Bed volumes prior to regeneration o Continuous regeneration Pretreatment: Filtration   Footprint Salt Consumption (@ 2400 gpm)  o Treatment system: 4400 sq.ft. o 4.9 tons/day – Nitrate mode o Perchlorate mode 18.6 Residuals handling: 2300 sq.ft. o – o Total system: 6700 sq.ft. tons/day  Volume of brine generated (@ 2400 gpm)  Ion exchange vessels Nitrate mode: 12.7 gpm o o Number of vessels: 30 Diameter of vessels: 3’ ency 99.5% water effici  o Perchlorate mode: 31 gpm o Height of vess o els: 6’  98.7% water efficiency  Max. nitrate concentration goal for delivered water Monitoring  - ~36.3 mg/L NO o (~8.2 mg/L N) Weekly effluent testing for o 3 nitrate and perchlorate Nitrate concentration goal for the  o Monthly effluent testing for treatment system (before blending) - and total Coliform , nitrite, sulfate (~4.3 mg/L N) ~19.0 mg/L NO o 3 Monthly raw water testing for o Manufacturer: Calgon Carbon Corporation  nitrate, perchlorate, nitrite,  Resin Type: Purolite SBA sulfate, tot al Coliform, HPC  Grab Samples for resin byproduct testing were negative for nitrosamines Residuals Management - reclaimable waste pipeline leading to the LA County Sanitation District, with a Waste brine is discharged to a non total disposal cost of ~$50, 000/yr. Technology Benefits and Drawbacks Benefits Drawbacks  Flexibility of operation and High efficiency None listed   Reliability and Ease of O & M Treatment Technology Costs Capital Costs (Total with explanation or component costs) llion Total: ~ $4.6 mi Housing: ~ $492 K Piping: ~ $1.1 million Resin: ~ $350 K Testing: ~ $20 K Other: ~ $2.4 million for ISEP Equipment and Engineering; ~ $350 K for electrical upgrade; ~ $200 K for pumps and associated equipment. Annual O&M Costs (Total with expla nation or component costs) Total: Not reported ~ $50 K Brine Disposal or Treatment: Chemicals: ~ $364 K for salt, ~ $50 K for hydrochloric acid 47 Technical Report 6: Drinking Water Treatment for Nitrate

64 3.2 Reverse Osmosis (RO) - of - Use Reverse osmosis can be a feasible option for nitrate removal in both municipal and Point applications ( 1995; Black 2003 ; Howe 2004 ). The first commercial application of RO for Cevaal et al. potable water treatment was in Coalinga, CA , in 1965 ( National Academy of Engineering 2008). RO can be used to address multiple contaminant e . g ., nitrate, arsenic, sodium, s simultaneously including ionic ( chloride , and fluoride), particulate ( ., . g ., asbestos and Protozoan cysts) , and organic constituents ( e . g e membrane under permeable - som e pesticides) (Dvorak & Skipton 2008). Water is forced through a semi pressure such that the water passes through, while contaminants are impeded by the membrane. A Figure 8 . typical process schematic of RO for nitrate removal from potable water is illustrated in Figure 8 . Reverse osmosis schematic. The required pressure will be dependent on the concentration of solute in the feed water. The collected l. concentrate is high in nitrate and other rejected constituents (salts) and requires appropriate disposa The extent to which the RO membrane removes constituents from the water is called the rejection rate. Rejection rates for sodium chloride and sodium nitrate can be as high as 98% and 93%, respectively ate of an RO system refers to the “maximum (Elyanow & Persechino 2005). The water recovery r percentage of permeate produced from a given feed flow,” and the flow/flux rate is “the maximum flow of the membrane” (Cevaal et al. 1995 , p. 107 ). Modifications and of permeate through a sq. ft. improvements to standard RO have led to the emergence of more efficient RO processes including High ™ Efficiency Reverse Osmosis (HERO ) and Ultra - low Pressure Reverse Osmosis (ULPRO) systems. 3.2.1 Reverse Osmosis - Design Considerations 6 summarizes key design considerations in the application of RO to nitrate removal from potable Table water. 48 Technical Report 6: Drinking Water Treatment for Nitrate

65 o . Summary of d esign c onsiderations for r everse 6 smosis . Table  Thin film membranes - Higher rejection rates, lowe r pressures than CTA membranes - Cellulose triacetate m embranes (CTA)  T olerant of low chlorine levels RO Membranes  Hollow fiber membranes - Compact configuration  Ultra - low pressure RO membranes (ULPRO)  Consider rejection rate, water recovery, and frequency of cleaning  Multiple contaminant removal  Prevent membrane damage, scaling and biological, colloidal , and organic fouling  Scaling o Acid (e.g., sulfuric acid) and/or anti - scaling agents (e.g., poly - acrylic acid) o Water softening  Biological fouling Pretreatment o Upstream disinfection, but dechlorination to prevent membrane damage  Reducing agent (e.g., sodium bisulfate) or activated carbon  Colloidal fouling o Pre - filtration to remove suspended solids Chemical treatment to keep suspended solids in solution o Av oid corrosion  Adjust pH, restore alkalinity for buffering capacity (see o - remineralization) and/or add corrosion inhibitor (e.g., poly ortho phosphate blend ) Post Treatment -  Remineralization o Blending, pH adjustment, addition of caustic soda, bicarbonate, sodium carbonate, ph osphates, and/or silicates  Blending, disinfection, and storage pH adjustment, up and down (acids and bases)  Chemical Usage  Anti - scalants  Cleaning chemicals (acids and bases) Frequency of membrane cleaning depends on water quality and membrane  used o pically once a month for 1 hour Ty filtration system Management of chemicals and pre -   Waste storage and disposal  Membrane replacement/membrane life O&M o Up to 20 years or more with appropriate pretreatment and maintenance lux rate Monitoring of nitrate levels and membrane f   Automation can be feasible  Low operational complexity (though higher than IX depending on pretreatment needs)  Maximize water recovery while minimizing energy use o Pressure range of 100 to 200 psi System er quality o Based on system size and feed wat Components Key system configuration parameters are system flow rate, number of  membranes/stages, system footprint, flux rate, water recovery rate, pump selection and sizing, pressure requirement, cleaning frequency 49 Technical Report 6: Drinking Water Treatment for Nitrate

66 Significan t cost of waste brine/concentrate disposal of greatest concern for  inland systems  Management options include sewer or septic system, drying beds, trucking off - , and advanced treatment site, coastal pipeline, deep well injection Waste Management and  Disposal options can be limit ed by waste brine/concentrate water quality (e.g., Disposal volume, salinity, metals , and radionuclides)  Optimization of recycling and treatment of waste brine/concentrate  Higher water recovery (more costly) to minimize waste volume (tradeoff between energy costs a nd disposal costs)  Need to prevent membrane scaling, fouling , and damage o Hardness, iron, manganese, suspended solids, silica , and chlorine Limitations  High energy demands  Disposal of waste concentrate  Complete demineralization (no control over target cons tituents) Water Quality While regulated by operational pressure, the water recovery rate depends largely on feed composition. d Problematic constituents include sulfate, calcite, calcium sulfate dehydrate (gypsum), silica, colloids, an vaal et al. microorganisms (Ce 1995; Elyanow & Persechino 2005; Tarabara 2007). Filtration upstream of the RO membranes is required to remove suspended solids. The life of the RO membranes and n water quality and prefilters, and the frequency of membrane cleaning are also directly dependent o the efficiency of pretreatment measures. Treatment efficiency can be compromised by membrane fouling. Anything that decreases available membrane surface area can limit the passage of water through the membrane and decrease water very. The four main types of membrane fouling are scaling, colloidal fouling, biological fouling , and reco organic fouling. When the salt concentration in the feed water exceeds the saturation point at the membrane surface, precipitation of solids on the memb rane can diminish the removal effici ency 14 2005). Scale forming constituents , (Elyanow & Persechino such as precipitates of silica, calcium, barium , and strontium salts pose a significant threat to RO by limiting the membrane surface area 14 Pretreatment options to prevent scaling include By addition of acid and/or anti the scaling agents and water softening. - d of ecreasing the pH, the prevalent form of the carbonate cycle is bicarbonate rather than the carbonate ion. The precipitation calcium carbonate will therefore be limited by the concentration of the carbonate ion. Note: the addition of acid helps only if the scale forming constituent is calcite, due to the speciation of carbonate (Lenntech, 2009c). Anti - scaling chemicals can function in three ways: threshold inhibition, crystal modification , and dispersion (Lenntech, 2009c). Threshold inhibition occurs when the anti - scalant increases the solubility of a potential scalant to super saturation, allowing for a greater concentration to remain in solution. Crystal modification refers to the interference of negatively charged functional groups on the anti scal ant - with salt crystal formation and membrane attachment. Anti - scaling chemicals can also promote dispersion of crystals by attaching to them and increasing their negative charge. The most economical means of membrane scale control will depend on the syst em, but typically the use of anti - scaling agents alone or in combination with acid addition is the most financially prudent option. The third alternative to prevent membrane scaling is to remove the problematic constituents from the water entirely. softening can be used to replace calcium and magnesium cations with sodium ions. Generally the most expensive Water option, this requires the addition of a water softener upstream of the membranes and will result in an additional brine waste stream. 50 Technical Report 6: Drinking Water Treatment for Nitrate

67 15 . through which wat Development of biomass on the membrane surface can have a similar er can pass 16 . Additionally, RO membranes have limited to zero tolerance for negative effect on performance membrane perform 2009e). organic species, like grease and oil; organic fouling inhibits ance (Lenntech Lastly, suspended solids not removed by pretreatment filtration can inhibit membrane performance 17 . Silica can be a particularly problematic constituent for RO membranes due through colloidal fouling ilica fouling and silica scale formation, which can be very difficult to to the potential for colloidal s remove. Modifications to conventional RO have emerged to manage high silica source waters (Section - Improvements and Modifications , b elow). 3.2.5 Reverse Osmosis Common pretreatment measures used to address membrane fouling are included in Table 6 . With acid addition in pretreatment, the permeate will need to be neutralized in post - treatment to avoid a corrosion in the distribution s such as , poly - orthophosphate ystem. Alternatively a corrosion inhibitor blend, can be used ( U . S . EPA 2003). Additionally, because RO is not selective in the removal of ions, treated water is demineralized. Thus, alkalinity may need to be added to rest ore minerals and buffering 18 2004 (WHO capacity to avoid corrosion in the distribution system ; Lenntech 2009f). In post - treatment, blending (if used) follows the RO modules, after which water is disinfected and stored (Cevaal et al. 1995). Considerations nts and Site System Compone RO systems are operated in stages. Following pretreatment, water is pumped through the membranes 1995). Water recovery can be improved by passing the concentrate th booster pumps (Cevaal et al. wi through the membranes more th an once, but higher water recovery comes at the expense of increased scaling potential. As the water becomes more concentrated, saturation can lead to precipitation on the 2005). The membrane flux rate can be decreased to l imit scaling and membrane (Elyanow & Persechino se membrane life (Cevaal et al. 1995). Key aspects of the system are pressure pumps, membrane increa configuration, membrane flux rate, number of stages/number of membranes, flow rate, and cleaning - scalant requirements. Pump sizi and anti ng is based on system size and pressure requirements. The water recovery rate can be regulated by the operational pressure. The necessary pressure is dependent /cm, on the concentration of solute in the feed water. For example, with a conductivity of 1550 μS 15 The Lang elier Saturation Index (LSI) and other corrosion indices are used to characterize the scaling potential of calcium carbonate (calcite), a commonly problematic constituent. 16 Upstream disinfection can limit membrane biofouling; however, additional measures must be taken to avoid membrane d carbon” exposure to chlorine “by dosing with a reducing agent (such as sodium bisulfate) or by contacting with activate 2005 ). (Elyanow & Persechino , p. 6 17 The potential for colloidal fouling is typically characterized by the silt density index (SDI). A n SDI greater than 3 can indicate the need for further pretreatment to minimize cleaning frequenc y and membrane damage (Lenntech 2009e ; Elyanow & Persechino 2005). Chemical treatment can keep suspended solids in solution. Alternatively, a prefilter can be used to remove solids from the feed water (Remco Engineering N.D. ). 18 It is important to note the relationship between alkalinity and pH. Alkalinity is a measure of buffering capacity or the resistance to changes in pH. Demineralized water or water with a low buffering capacity will be susceptible to more dramatic pH changes and is considered unstable. The pH of acidic product water should be adjusted an d the buffering capacity of demineralized product water should be restored to avoid corrosion downstream. Options for the stabilization and remineralization of demineralized water include blending, pH adjustment, and addition of caustic soda, bicarbonate, sodium 2009f). Corrosion indices and models can be used to determine osphates, and/or silicates (WHO 2004; Lenntech carbonate, ph appropriate solutions for specific scenarios. 51 Technical Report 6: Drinking Water Treatment for Nitrate

68 Panglisch et al. (2005) determined the suitable pressure to be 145 – 174 psi. Similarly, at start - up, the oval was 170 psi (Cevaal et al. operational pressure at an RO facility in Brighton, CO , used for nitrate rem 1995). Two types of spiral wound R O membranes are commonly used: polyamide thin film composite Cevaal et al. membranes (TF) and cellulose triacetate membranes (CTA) ( 1995 and Remco Engineering lower ). While TF membranes are capable of slightly higher rejection rates and can be operated at N.D. re tolerant of low chlorine levels (AWWA 2011 ). Hollow fiber RO pressures, CTA membranes a membranes are also available, which can minimize system footprint , but can be more susceptible to suspended solids (Hydranautics fouling from 2001). Using rec ently developed ultra - low pressure RO (ULPRO) membranes, operational pressures can be reduced, decreasing power costs (Drewes et al. Improvements and Modifications . 2008). For additional information see S ection 3.2.5 Reverse Osmosis - Disposal Residuals Management and The volume of the waste stream can be considerable, ranging from 15% to 50% of the starting volume depending on the operational parameters (Howe 2004). The waste stream, or concentrate, can be discharged to a wastew lant or a septic system (Bilidt 1985 ; Howe 2004), as long as the ater treatment p system can accommodate an increased salt concentration. Additional disposal options include drying beds, infiltration basins, trucking off - site, a coastal pipeline, deep well injection, advanced treatment, and most commonly, discharge to nearby surface salt - waters (i.e., oceans), when available (Howe 2004). Important water quality characteristics of the waste brine (e.g., volume, salinity, metals , and radionuclides) can affec t the feasibility and costs of disposal. Options for inland communities are more limited and costly. Proximity to coastal power plants can be advantageous. Power plants using ocean to ocean waters (Black 2003). water for cooling can provide a pre - existing infrastructure for disposal The high cost of nitrate laden concentrate disposal has led to research into optimization of recycling and treatment of this waste stream. Coupling of RO systems with biological, chemical, or catalytic denitrification enab les removal of nitrate from the waste concentrate, with reduction to nitrogen gas. Additionally, the Vibratory Shear Enhanced Process (VSEP), from New Logic Research, has been explored in the context of treatment of RO concentrate from wastewater treatmen t. By applying a shear force across the membrane, pore clogging by colloidal particles is minimized, leading to the potential for N.D.). Several combined configurations of interest are discussed improve d water recovery (Lozier et al. below in . S ection 3.6 Brine Treatment Alternatives and Hybrid Treatment Systems Maintenance, Monitoring , and Operational Complexity RO systems are typically highly automated, accommodating the greater operational complexity of RO ison with IX. Several operational decisions will be dictated by operator availability operation, in compar and training. For instance, chemical addition in pretreatment can be quite effective, but will require more intensive maintenance. In contrast, opting for the more exp ensive choice, installing a water softener, will require less operator time. Membrane cleaning frequency varies widely and depends on the efficiency of pretreatment measures and water quality. Interruption of operation is not always 52 Technical Report 6: Drinking Water Treatment for Nitrate

69 ranes can be isolated and cleaned in place (CIP) in stages. Membranes are necessary as the memb typically cleaned with ac id or caustic solutions (WA DOH 2005). Cleaning agents are selected based on the cause of membrane fouling. Cleaning and rinsing can take an hour and with effective pretreatment, monthly membrane cleaning should be sufficie nt (Remco Engineering N.D. ; Bates N.D.). With effective - using anti pretreatment, cleaning frequency can be significantly minimized. A n RO plant in Milan, Italy , ing only once every 18 months (Elyanow & Persechino 2005). Filters should be scalants, requires clean checked weekly and, if used, the water softener should be maintained with sufficient salt every day. Effluent nitrate concentrations require monitoring to ensure compliance and to assess membrane performance. Over time, membrane degradation will lead to a gradual decrease in removal efficiency. Membrane life varies and can range from 5 to 20 years or more (Remco Engineering N.D. ). Waste concentrate management consists of appro priate storage and disposal. More operationally complex than IX, operators of RO systems will typically require more training and system maintenance will demand more time and chemicals. However, with the implementation of appropriate pretreatment measure s and the ability for system automation, operational complexity can be minimized. Considerations 3.2.2 Reverse Osmosis - Cost For the efficient operation of an RO system, the fundamental objective is to maximize water recovery with the minimum amount of en ergy and chemical usage, while meeting necessary potable water guidelines. Factors affecting system cost include facility size (how much water), source water quality effluen t (including nitrate concentration), environmental factors (temperature and pH), and target nitrate concentration (Bilidt 1985). Lower operating pressures are less costly, but result in decreased water recovery. High operating pressures maximize water recovery (decreasing disposal costs), but for “ specialized pumps” (WA DOH 2005). Thus, there is a trade increase energy demands and the need - off between the costs of increasing water recovery (increased pretreatment and operational pressure) and the costs of disposal (pumping, storage , and disposal expenses). In pretreatment, the use of anti - scalants rather than acid or a water softener is generally the least expensive. The use of a water softener is the least cost competitiv e option (Lenntech 2009c). Regarding small water systems, “reverse osmosis is one of the most expensive forms of be cost effective centralized treatment and will likely not unless there are multiple contam inants needing removal” (WA DOH 2005 , p. 27 ). Capital costs for RO include land, housing, piping, storage tanks, O&M equipment, membranes, preliminary testing (pil ot studies), permits, and training. O&M costs include membrane and filter - replacement, membrane and filter disposal, concentrate disposal or treatment, chemical use (anti scalant, pH adjustment, disinfection , etc.), repair, maintenance, power, and labor. . 7 Table Published cost information, from existing RO installations used for nitrate treatment, is listed in Costs have been adjusted to 2010 dollars, unless indicated otherwise. Costs can be difficult to assess due to inconsistencies in how cost information is reported. Comparison of treatment costs is not always valid due to differences in influent water quality parameters, system size, waste management options, always include comparable information. It would be and system configuration. Published costs do not inappropriate to compare the O&M costs of a facility that excludes disposal costs with others that include this information. The listed cost information is provided as an approximate range of costs for 53 Technical Report 6: Drinking Water Treatment for Nitrate

70 specific facilities. Costs for implementing RO may be very different from those listed here. A thorough cost analysis of design parameters for specific locations would be required for accurate cost estimation. The information gathered through the questi onnaire includes detailed costs associated with the unique case studies included in this analysis. A detailed discussion of treatment costs is included below in Section 6 Treatment Cost An alysis . 7 . Selected Table osts* of r everse o smosis s ystems for n itrate r emoval. c System Flow** < 0.5 MGD 0.5 – 5 MGD 5+ MGD – Annualized Capital Cost ($/1000 gal) – 5.17 [1, 2] 1.00 3.51 1.30 [3, 4] 0.95 [3] 1.46 – 16.16 [1, 2] 1.22 – 2.01 [3, 4] 1.63 [3] O&M Cost ($/1000 gal) 3.21 [3, 4] 5.73 2.58 [3] Total Annualized Cost ($/1000 gal) – 19.70 [1, 2] 2.52 – *Costs have been adjusted to 2010 dollars with 7% interest over 20 years, unless indicated otherwise. ther than design capacity. **When available, costs are based on actual system flow ra [1] Walker (N.D.), costs not adjusted to 2010 dollars. [2] Personal communication with two representatives of small water systems (2010). [3] Conlon et al. (1995). [4] Cevaal et al. (1995). 3.2.3 Reverse Osmosis - Selected Research Much research has focused on RO; Table A.2 of the Appendix is a list of recent studies relevant to nitrate removal from potable water and several examples of RO application. Current RO research focuses on improvements of membranes and waste manag ement, and decreasing energy use. 3.2.4 Reverse Osmosis - Summary of Advantages and Disadvantages A summary of advantages and disadvantages of RO in comparison with other treatment options is listed in Table A.6 of the Appendix. Advantages of RO include h igh quality product water, multiple and application for POU contaminant removal, desalination (TDS removal), feasible automation , applications. According to Elyanow & Persechino (2005) , in their comparison of RO and EDR, “... the best 3 economical choice for s mall capacity systems (<110 gpm or <25 m /hr) are simple RO plants, which have (p. 7). less electrical and hydraulic complexity than EDR and other technologies” In waters where salinity is a problem, RO can be better suited than IX due to the ability to r emove multiple contaminants (including trihalomethane formation potential prec ursors (THMFPs)) (Cevaal et al. 1995). Disadvantages of RO include high capital and O&M costs, membrane fouling susceptibility, high pretreatment and energy demands, and potentia lly large waste volume (lower water recovery) requiring disposal. The high cost of disposal from inland locations can result in RO treatment becoming cost e prohibitive. Howe (2004) presents several alternatives to conventional disposal measures of RO wast brine, including reuse for industrial processes, processing (e.g., for salt production), or use in energy generation (“solar brine pond”). 54 Technical Report 6: Drinking Water Treatment for Nitrate

71 - Improvements and Modifications 3.2.5 Reverse Osmosis Process Modification ™ High Efficiency Reverse Osmosis (HERO ) is a patented multi - step process enabling increased water recovery (greater than 90%) and minimizing cleaning requirements. This process limits scaling by ted 2010b). Raw water is subjec incorporating hardness reduction, CO tripping, and pH adjustment (GE s 2 to intensive pretreatment before passing through the RO membranes as follows (Engle 2007):  Weak acid cation exchange (WAC) is used to remove hardness ions, CO stripping is used to remove carbonate and increase pH, and  2 Base addition is used to increa se the pH to a level of 10.5.  . An example flow diagram is illustrated in Figure 9 With such pretreatment, water fed to the RO membranes is softened and pH is adjusted high enough to ca. The high chemical usage and multiple steps result in a more significantly increase the solubility of sili complicated process than conventional RO. However, benefits include increased water recovery, decreased waste volume, and the ability to treat severely impaired and poor quality source water containi ng multiple contaminants (Engle 2007). ™ The HERO process was initially designed to produce ultra - pure water for use in electronics applications and was patented by Debasish Mukhopadhyay with licensing rights fo r different applications (Engle ™ 2007 ). The HERO process has been implemented for drinking water treatment in the small community , ica and to produce high quality drinking water from brackish groundwater high in sil of Yalgoo, Australia 2009 ; Thomson et al. 2009 ). Higher removal rates nitrate (Water Corporation 2007; Water Corporation ™ result in decreased waste volume. Using the HERO process in Yalgoo, waste volumes are as low as “one - tenth of a conventional plant’s concentrat ed brine residue for disposal, eliminating the need for 2007 ). big evapo r ation ponds” (Water Corporation , p. 8 55 Technical Report 6: Drinking Water Treatment for Nitrate

72 Central Arizona Salinity Study Figure 9 . Flow chart of the HERO™ process . ( Source: repr oduced with p ermission, . ) 2006 Membrane Modification – Low Pressure Membranes Research and development in membrane technology has resulted in the emergence of Ultra - Low Pressure Reverse Osmosis (ULPRO). In contrast to the high pressures required for conventional RO, use ates. Energy demands of ULPRO membranes allows for lower operating pressures and improved flux r can be reduced due to lower operating pressures; however, pretreatment practices to prevent membrane scaling and fouling are similar to those necessary for conventional RO membranes (Drewes et al. 2008). ULPRO membranes are available from several manufacturers. Operating pressures are in be over the range of 50 to 125 psi, while the pressures required for conventional RO membranes can Drewes et al. Excel Water 2007; 200 psi ( 2008; Koch Membrane Sy stems 2008). Drewes et al. (2008) wherein co mpared the performance of ULPRO membranes and conventional RO membranes , filtration for both RO options. Findings indicate that the ULPRO - included nano pretreatment capable of successfully removing nitrate and multiple additional membranes included in the study were contaminants to potable water standards. “With regard to operating costs, operating pressure is the only TMG [ULPRO membrane] operating parameter considered to deviate from the benchmark ESPA2 the were [conventional RO membrane] membrane. Pret reatment requirements and recovery rate rating pressure” (Drewes same. Electrical consumption will be directly proportional to the required ope 2008, p. 93). However, in the cost comparison between the two membranes, the benefits of lower et al. operating pressures were overshadowed by the poor recovery of the ULPRO membranes after cleaning. 56 Technical Report 6: Drinking Water Treatment for Nitrate

73 The authors suggest that the cleaning of the ULPRO membranes would need to be optimized for an improved cost comparison. - 3.2.6 Reverse Osmosis Case Studies scale RO The following case studies provide detailed information on the design and operation of full - treatment plants used for nitrate removal. Chino I Desalter and Chino II Desalter are combination ntional IX. systems using both RO and conve 57 Technical Report 6: Drinking Water Treatment for Nitrate

74 System Location: California CASE #6 System Type: Community Water System Treatment Type: Reverse Osmosis (RO) Startup Date: 2002 System Description Reverse Osmosis and Blending Treatment Type 120 gpm System Capacity A California water utility operates a Raw Water Nitrate 84 mg/L as n itrate 75 – system that has three groundwater 17 19 mg/L as N – supplies, one of which has nitrate at levels that exceed the MCL. The impacted well has nitrate concentrations that range from 75 to 84 mg/L as nitrate (17 to 19 mg/L as N) and has a typical production capacity of 100 gpm. In 2002, the utility implemented a blending program and installed a n RO system for nitrate control and treatment. In RO, raw water is forced through a semi - permeable membrane under pressure such that the water passes through, while contaminants are impeded by the membrane. The required pressure will be dependent on the ejected concentration of solute in the feed water. The collected concentrate is high in nitrate and other r constituents (salts) and requires appropriate disposal. The extent to which the RO membrane removes constituents from the water is called the rejection rate. Rejection rates for sodium chloride and sodium nitrate can be as high as 98% and 93%, re s pectively (Elyanow & Persechino 2005). The high nitrate supply is blended with one of the other two groundwater sources prior to RO treatment. The RO o consumers system reliably removes nitrate to below 35.4 mg/L as nitrate (8 mg/L as N) and the water delivered t typically has nitrate levels below 13.3 mg/L as nitrate (3 mg/L as N). This system utilizes a leach field type system to land apply the RO concentrate. Source Water Quality mg/L as nitrate (mg/L N) - Nitrate Co - contaminants   Average - 75 (17) o Fluoride - 3.3 mg/L o o Minimum - 80 (18) o Arsenic 84 (19) - o Radium o Maximum Treatment Technology Selection RO was selected as the most appropriate treatment system as the technology can reliably remove nitrate in addition to the co - occurring contaminants that are present, specifically fluoride, arsenic , and radium. Technical Report 6: Drinking Water Treatment for Nitrate 58

75 Treatment System Parameters  Design Capacity Clean - in - place (CIP)  o CIP frequency: Quarterly 120 gpm maximum capacity o (4x/year) Pretreatment  o Initiated when there is a 15% o - scalant (Hyposperse MCD Anti decrease in permeat e flow or salt 150) rejection or a 15% increase in Treatment system footprint  - trans membrane pressure o Treatment system: 15’ x 30’ CIP chemicals: Dilute phosphoric o m footprint: 40’ x 100’ Total syste o acid RO System   Water efficiency: 80% System manufacturer: Aria™ o  Monitoring o Membrane manufacturer: o Laboratory samples Osmonics Number of stages: 4 o  Service life of membranes o Number of RO elements per Approximately 8 years o stage: 4 Residuals Management - posed to an on site leach field. The concentrate is dis Technology Benefits and Drawbacks Benefits Drawbacks  Effectively removes nitrate and other co -  Energy intensive contaminants  Relatively low water efficiency (80%)  On - site concentrate disposal  Consistent operations eatment Technology Costs Tr Treatment technology costs are not available for this system. Operational Notes The RO system has never had any extended unplanned shut downs or been shut down as the result of an alarm. oride MCL that occurred near the end of the useful life of the membranes. There has been an exceedance of the flu The membranes have since been replaced, resolving this operational issue. References – GE Elyanow, D. and Persechino, J. (2005) mpany, Water & Advances in Nitrate Re moval. General Electric Co Process Technologies. Accessed June 11, 2010 via < http://www.gewater.com/pdf/Technical%20Papers_Cust/Americas/English/TP1033EN.pdf>. Technical Report 6: Drinking Water Treatment for Nitrate 59

76 System Name: City of Brighton System Location: Brighton, CO PWSID: CO 0101025 CASE #7 Syst em Type: Community Water System Treatment Type: Reverse Osmosis (RO) Questionnaire completed by: Dave Anderson, City of Brighton RO Chief Plant Operator Startup Date: 1993 System Description Reverse Osmosis Treatment Type The City of Brighton (City) utilizes six 600 gpm) 6.65 MGD (4, System Capacity groundwater wells with production – Raw Water Nitrate 89 mg/L as nitrate 49 capacities ranging from 900 to 1500 11 20 mg/L as N – gpm and one groundwater source which has been designated as groundwater under direct influence of surface water (GWUDI), as an emergency well, with a production capacity of 700 gpm. These seven sources are impacted by nitrate with average concentrations ranging from 49 to 89 mg/L of nitrate as nitrate (11 to 20 mg/L of nitrat e as N). The City has implemented RO with blending. The design capacity of the RO system is 6.65 MGD of permeate at 80% recovery (1150 gpm/train). Green sand and cartridge filters (Graver) are used to treat the GWUDI source, primarily for the removal of manganese. (Additional sources operated by the City that are not impacted by nitrate are purchased treated surface water and additional GWUDI wells.) Raw water enters the system with 40% of feed water bypassing the RO system and 60% of feed water passin g to pretreatment. After anti scalant addition, pretreated water is pressurized with boost pumps and passed to the RO - skids. Waste concentrate exits the system for disposal and post - treatment of the permeate includes CO stripping 2 treated water is blended with bypassed water and sent to storage orine and caustic. Post and the addition of chl - and ultimately distribution. Source Water Quality Nitrate - mg/L as nitrate (mg/L N)  46.9 to 90.3 (10.6 to 20.4) - o Average o Minimum - 20.02 to 69.97 (4.52 to ) 15.8 Maximum 78.8 to 112.9 (17.8 to o - 25.5) contaminants  Co - TDS: 580 to 1000 mg/L, RO TDS ~34 o mg/L, Finished TDS ~280 mg/L o Fluoride: 1.3 mg/L TOC: < 2 mg/L o o Hardness: 370 480 mg/L as CaCO – 3 (historically, Cevaal et al. 1995) 60 Technical Report 6: Drinking Water Treatment for Nitrate

77 Treatment Technology Selection IX and EDR were also considered and pilot tested prior to installation of the RO system. RO was selected due to nitrate levels and hardness. IX could have been less costly; however, the lower salt levels in RO concentrate make ste to the South Platte River. “By selecting RO, the City hoped to actually reduce the salt it possible to discharge wa - load on the river with RO since many Brighton residents currently using home ion exchange softening units would 1995 , p. 102 Biological treatment is also being explored for the treatment of no longer use them” (Cevaal et al. ). nitrate in the waste brine. Residuals Management The waste concentrate is continuously discharged via a brine line to the South Platte River. Biological treatment is being explored for the treatment of the waste concentrate. The biological system would be located on the West side of the RO treatment system and would allow for reduction of nitrate in the waste stream. As mentioned to decrease salt loading to the South Platte River. above, the use of RO rather than ion exchange was an effort home ion exchange units to soften water. - Historically, Brighton residents used in Technology Benefits and Drawbacks Benefits Drawbacks Constant gener   Consistent treatment performance ation of waste stream Ease of operation   High power consumption 61 Technical Report 6: Drinking Water Treatment for Nitrate

78 Treatment System Parameters   Design Capacity Flux rate of the RO membranes 13 gpd/sf 6.65 MGD of permeate at 80% o o recovery (1150 gpm/train) System Manufacturer  - treatment o Hydranautics and Hydrocode  Pretreatment/Post scalant: - Anti o Membrane Type  King Lee Technologies CPA2 (no others used in past) o Pre Treat Plus 0100 phosphonate  Membrane Life 5 micron car tridge filters o o Unknown, none have required 2.5 inch diameter  replacement (5 yrs. ago, the  90 day replacement manufacturer said the pH adjustment o membranes should last 3 more  caustic soda (NaOH) yrs.).  removal) Air stripping (CO 2 Membrane Cleaning   Treatment system footprint o Clean in Place initiated by time Treatment system: o rather than decrease in flux. 11,000 sq. ft. at installation Every ~157 million gals treated 1995) (Cevaal et al. (~2x/yr)  Number of contactors Chemicals: o 2 stages, 5 trains o Nalco Product and Citric acid RO elements/stage o  Waste 36 x 18 array  o Discharge via Brine line to South 6 membranes/vessel  Platte River  324 total Recovery Rate: 80% o Max. Concentration goal for delivered water   toring Moni o 35.4 mg/L as nitrate (8 mg/L as N), o Ion chromatography (always produce lower) At Source  Rejection Rate   At Point - of - Exit o – 95 98% rejection Grab ISE (HACH) o Nitrate goal (before blending): o At Blending Point  ~4.4 mg/L as nitrate (~1 mg /L as N) - Exit  At Point - of o Testing once per year is required for compliance Treatment Technology Costs Capital Costs (Total with explanation or component costs) Operating and M onitoring $8,253,000 (1993) 4MGD RO facility Equipment: Annual O & M Costs (Total with explanation or component costs) $2,873,293.00 Total: Membrane: 0 Membrane Disposal: 0 Brine Disposal or Treatment: 0 Chemicals: Approx. $100,000 year Power: Appro x . $210,000 year for RO Labor (Hours per Year): 10 hr/day, 7 day/wk 2 MGD Thornton treated : $3.60/1000 gallons COMPLETE Cost (including treatment, distribution, $3.16/1000 gallons : everything) Technical Report 6: Drinking Water Treatment for Nitrate 62

79 Additional Information  d for plant operators is Colorado A treatment. The level of certification require During a power outage, there is a pause before the generators start. This requires a manual restart of the  system. This system has never produced water exceeding the nitrate MCL and has never had an unplanned  shutdown exceeding one week.  The major benefit of the RO system is the rejection rate allowing for removal of regulated contaminants.  The most significant disadvantages are the high power consumption and the continuous brine discharge. noted that there has been a decreasing trend in nitrate levels in their sources. The operator also  References Cevaal, J.N., Suratt, W.B., and Burke, J.E. (1995) Nitrate removal and water quality improvements with reverse – 111. osmosis for Brighton, Colorado. Desalination , 103 , 101 Anderson, D. (2010) In - person interview and tour of the facility. July 19, 2010. September 15, 2010. Anderson, D. (2010) Completed questionnaire. 63 Technical Report 6: Drinking Water Treatment for Nitrate

80 System Name: Western Municipal Water District Arlington Desalter - e, CA System Location: Riversid PWSID: CA3310049 CASE #8 System Type: Community Water System Treatment Type: Reverse Osmosis (RO) and Blending Questionnaire completed by: Joseph Bernosky, Director of Engineering Startup Date: 1990 strictly for desalting, Upgraded to drinking water treatment in 2002 System Description The Western Municipal Water Reverse Osmosis and Blending Treatment Type District (District) operat es a system 6.6 MGD (4,600 gpm) System Capacity comprised of seven wells, five of Raw Water Nitrate 89 mg/L as nitrate 44 – which contain nitrate above the – 10 20 mg/L as N MCL. Three of the nitrate impacted wells are treated by a 6.6 MGD RO facility. The permeate, or treated water, from the RO system is blended with the remaining two wells prior to distribution. The RO and blending facilities are collectively referred to as the Arlington Desalter. Approximately 60% of the total flow is treated by the RO system and the remaining 40% is blended with the treated water. The te concentration of 22 mg/L as nitrate (5 mg/L as N) in the distribution system. District targets a nitra The Arlington Desalter facilities were originally installed in 1990 to address the salt imbalance in the Upper Santa South Arlington Basin were treated by the Arlington Ana Watershed. High salinity waters withdrawn from the Desalter and subsequently discharged to the Santa Ana River for downstream use (and downstream drinking water treatment). The system was upgraded to a drinking water treatment facility in 2002 with the addition of disinfection, a clear well and a pump station used to pump drinking water into the distribution system for the city of Norco, CA. Source Water Quality - mg/L as nitrate (mg/L N)  Nitrate contaminants - Co  - TDS - 1200 mg/L o Average o 75 (17) o Minimum - 44 (10) - 89 (20) o Maximum Treatment Technology Selection Due to the original intent of the system and the 2002 conversion for drinking water production, no other cilities. However, biological technologies were pilot tested prior to the installation of the RO and blending fa scale implementation anticipated. A case study treatment of RO bypass water was recently pilot tested with full - about this fixed bed biological pilot study is listed separately. Technical Report 6: Drinking Water Treatment for Nitrate 64

81 Treatment System Parameters   Design Capacity: 6.6 MGD Flux rate of the RO membranes 2 o 16 gpd/ft -  Pretreatment/Post treatment - scalant Anti - scalant: Y2K Anti o  System Manufacturer o Hydranautics  Treatment system footprint  Treatment system: o Membrane Type 2 Approximately 7,500 ft o excluding Koch HR400 clear well and pump station Membrane  Life o > 10 years Stages  of Number o Stage 1: 36 vessels each with 6 Membrane Cleaning  units o Occurs 2 times per year Stage 2: o h with 6 12 vessels eac Chemicals: Low pH solution, o units hydrofluorosilic acid, high pH solution Max. Concentration goal for delivered  water Waste: The RO concentrate is disposed of  o 22 mg/L as nitrate (5 mg/L as N) offshore via the Santa Ana Regional Interceptor (SARI) brine line Water Recovery  Original design: 75 o 76% –  Monitoring o Current: 80% o Ion chroma tography o Online nitrate analyzers Residuals Management Waste is discharged to the Santa Ana Regional Interceptor (SARI) (Brine Line). The SARI line prevents degradation of natural waters caused by increased salinity. Managed by the Santa Ana Watershe d Project Authority (SAWPA), the SARI line is a dedicated interceptor line built to help users meet discharge requirements. In addition to the Arlington Desalter, the SARI line is used by other dischargers including industrial and domestic sources. The istrict’s contribution to the total flow of the SARI line is approximately 5%. The SARI line carries water to the D Orange County Sanitary District for wastewater treatment with ultimate offshore discharge. Having access to the SARI line for brine disposal is a benefit of this system; however, there have been complications with the SARI line. Tremendous scaling problems in Arlington Desalter’s reach of the SARI line have resulted in the need to address calcium carbonate buildup and to consider additional c leanout points. Due to multiple discharge sources, the SARI line combines waters having very different water quality characteristics. Reactions within the mixed water can vary based on water chemistry. Several cleaning and maintenance options considered. are being Technology Benefits and Drawbacks Drawbacks Benefits Costly brine disposal   Reliable  C omplications with SARI line (scaling, etc.)  Access to SARI line for disposal existing - The treatment potential of the pre  RO system has been maximized through conversion for drinking water treatment Technical Report 6: Drinking Water Treatment for Nitrate 65

82 Treatment Technology Costs Capital Costs (Total with explanation or component costs) Total Capital Costs: Unavailable for initial installation in 1990 Annual O&M Costs (Total with explanation or component costs) (2009/2010) Total: $2,931,228 Labor and Overhead: $836,530 Outside Services: $200,000 Emergency Repairs: $25,000 $85,000 General & Administrative: Vehicle and Equipment: $9,000 $151,800 SARI Fixed Cost: Materials and Supplies: $5,500 Perm its and Fees: $27,000 SARI Variable Cost: $470,000 Chemicals: $150,000 $971,398 Energy: Additional Information Regarding water recovery, the original design recovery rate was 75 – 76%. With modifications to anti - scalant use, the recovery rate incre ased to 78% and then to the current operational water recovery rate of 80%. To further improve the water recovery rate, additional testing is anticipated which will require engineering work and a cost - benefit analysis. Currently used membranes are 10 yea rs old and are still performing adequately with respect to operational parameters (flux rate, rejection rate, etc.). The District has budgeted for membrane replacement in this fiscal year; however, due to adequate performance, the current membranes may be used for an additional year. Membranes are actually attaining better water recovery than manufacturer specifications. Membrane life is also exceeding initial expectations. Sources * 2010) Bernosky, J. ( . November 5, 2010 Personal communication. October, 2010 . Bernos ky, J. (2010) Completed questionnaire. * Unpublished sources used in the development of the case studies are not reflected in the References section of this report. 66 Technical Report 6: Drinking Water Treatment for Nitrate

83 System Name: Chino Basin Desalter Authority (CDA) Chino I Desalter – tem Location: Chino, CA Sys PWSID: CA31610075 CASE #9 System Type: Community Water System ge (IX) Treatment Type: Reverse Osmosis (RO) and Ion Exchan Questionnaire completed by: Timothy Mim Mack, CDA Coordinator, City of Ontario René Cruz, Engineering Project Manager, Project Partners Inc., serving the (CDA). Startup Date: RO in 2000, IX added in 2005 System Description Treatment Type Reverse Osmosis, Ion Ex change The Chino Basin Desalter Authority (CDA) in And Blending southern California is a conglomerate of the System Capacity RO: 4940 gpm following agencies: Inland Empire Utilities IX: 3400 gpm Agency (IEUA), Jurupa Community Services District (JCSD), City of Chino, City of Chino Hills, 147 Raw Water Nitrate 303 mg/L as nitrate – City of Ontario, City of Norco, Santa Ana River 33 68 mg/L as N – Water Company (SARWC), and Western r treatment facilities include two desalters: Chino I Municipal Water District (WMWD). The CDA drinking wate Desalter (discussed here) and the Chino II Desalter (discussed in the next case study) to address high TDS levels as well as nitrate contamination. The Chino I Desalter operates 14 source wells, 11 of which have raw nitrate levels well above the MCL. Treatment consists of a combination of RO, conventional anion exchange and blending. Sixty stripping prior to percent of total flow is treated with RO, 27% with IX and 13% passes only through VOC/Air - blen ding. The RO system was installed in 2000 and the IX system was added in 2005. Source Water Quality   Nitrate - mg/L as nitrate (mg/L N) (of contaminants - Co TDS: 1100 mg/L o nitrate impacted wells) 303 (~33 o Average: ~147 – – 68) ~114 65) – o Minimum : 289 (~26 – – 351 (~36 – 79) : Maximum o ~161 Treatment Technology Selection , they were determined to be the best mode of technology to , RO and IX were selected because combined - - adequately treat the high TDS, high re pilot tested or considered nitrate source water. No other technologies we prior to the installation of the system. Technical Report 6: Drinking Water Treatment for Nitrate 67

84 Treatment System Parameters  Design Capacity System Manufacturer  line - RO: Code o RO: 4940 gpm o o IX: Hungerford and Terry o IX: 3400 gpm Membrane Type: Dow 400 BW  Pretreatment  30 - Anti o - scalant: threshold inhibitor Membrane Life: 5 years  o Filtration: 1 micron pre - filters  Membrane Cleaning o pH adjustment: sulfuric acid Flux decrease initiates CIP o  treatment - t Pos E o very 6 month pH adjustment: sodium hydroxide o o Chemicals: for pH adjustment based on manufacturer  Treatment system footprint recommendation RO system: 143’ X 80’ o o IX system: 190’ X 60’ Resin Type: Rohm and Hass Amberjet  4400 CL SBA  Number of contactors 4 trains, 2 stages/train o Volume treated prior to regeneration  o RO elements/stage o 700,000 gallons stage 1: 196 elements  o Regeneration once every 12 hrs  stage 2: 98 elements Salt consumption: 75 tons per week   Ion exchange pressure vessels  Volume of b rine/backwash o Nu mber of vessels: 4 o 53,000 gallons o Diameter of vessels: 12’ 92.4% water efficiency o Height of vessels: 11’ o  Resin life: Has not been replaced (online 2 Design Loading Rate: 1.66 gpm/ft  for 5 years) Max. nitrate concentration goal for   Monitoring delivered water: 36 mg/L as nitrate (8.13 o Online nitrate analyzers mg/L as N) Treatment train  Nitrate concentration goal for the  Blending point  ng): 10 treatment system (before blendi POE  mg/L as nitrate (2.25 mg/L as N) o Laboratory samples Quarterly testing for NDMA o RO recovery rate: 80%  RO membrane flux rate: 0.9 gfd/psi  Residuals Management Concen trate/brine is discharged into a regional brine line called the Inland Empire Brine Line (IEBL) and formerly known as the Santa Ana Regional Interceptor (SARI). Technology Benefits and Drawbacks Benefits Drawbacks  RO provides better removal  RO is expensive IX is inexpensive   High waste rate of RO has very low energy demands IX does not address TDS  IX   Resin replacement will be costly Technical Report 6: Drinking Water Treatment for Nitrate 68

85 Treatment Technology Costs Capital Costs (Based on projected costs in 2004) Treatment Plant Expansion Total (5000 $6,379,530 expansion): afy Ion Exchange Treatment (4.9 MGD): $4,031,900 Onsite Modifications : $1,735,000 $612,630 SARI Discharge Upgrades & Storm Drain: Additional SARI Capacity Purchase (not $4,140,000 included in above total): O & M Costs (Based on CD A 2010/11 Budget, for complete plant operation, not just the treatment system) Total: $7,496,315 Chemicals: $662,257 Electricity , Plant Total: $2,843,000 $1,353,439 (includes SARI fees, permits and other fees) Operating Fees: $1,370,698 Labor ($): Additional Information  The RO treatment system is described as falling short of expectations with respect to the high waste rate. 15% of all incoming water is sent to the brine line and delivered to a treatment plant outside of the local watershed at the Orange County Sanitation District. and high TDS.  Plant shutdown has been required in the past due to high or low pressure, high nitrate ,  In the event of insufficient treatment and the production of water in exceedance of an MCL, the plant has n MOV that closes automatically, sending water to a storm drain. a Sources * Listed costs are based on: Chino Basin Desalter Authority. (2005) Presentation: Chino I Desalter Expansion & Chino II Des alter Project Update. Year 2010/11 Budget Adoption. Chino Basin Desalter Authority. (2010) F iscal Unpublished sources used in the development of the case studies are not reflected in the References section of this * report. 69 Technical Report 6: Drinking Water Treatment for Nitrate

86 – System Name: Chino Basin Desalter Authority (CDA) Chino II Desalter System Location: Mira Loma, CA PWSID: CA3310083 CASE #10 System Type: Community Water System Treatment Type: Reverse Osmosis (RO) and Ion Exchange (IX) Questionnaire completed by: Timothy Mim Mack, CDA Coordinator, City of Ontario René Cruz, Engineering Project M anager, Project Partners Inc., serving the (CDA). Startup Date: 2006 System Description Reverse Osmosis, Treatment Type The Chino Basin Desalter Authority (CDA) in Ion Exchange southern California is a conglomerate of the And Blending following agencies: Inland Empire Utilities System Capacity RO: 4167 gpm Agency (IEUA), Jurupa Community Services District (JCSD), City of Chino, City of Chino Hills, IX: 2778 gpm City of Ontario, City of Nor co, Santa Ana River 224 mg/L as nitrate – Raw Water Nitrate 70 Water Company (SARWC), and Western 16 51 mg/L as N – Municipal Water District (WMWD). The CDA drinking water treatment facilities include two desalters: Chino I Desalter (discussed above) and the Chino II vels as well as nitrate contamination. The Chino II Desalter Desalter (discussed here) to address high TDS le operates 8 source wells, all of which have raw nitrate levels well above the MCL. Treatment consists of a , and blending. The combined RO/IX syste combination of RO, conventional anion exchange m was installed in 2006. Source Water Quality   Nitrate - mg/L as nitrate (mg/L N) (of contaminants - Co nitrate impacted wells) TDS o 224 (~16 51) o – Average: ~70 – Minimum – 43) o : ~53 190 (~12 – 260 (~18 – 59) : Maximum o ~81 – Treatment Technology Selection , they were determined to be the best mode of technology to , RO and IX were selected because combined - nitrate source water. No other technologies were pilot tested or considered - TDS, high adequately treat the high prior to the installation of the system. Technical Report 6: Drinking Water Treatment for Nitrate 70

87 Treatment System Parameters   Design Capacity System Manufacturer RO: PROTEC Bekaert o RO: 4167 gpm o IX: Hungerford and Terry o o IX: 2778 gpm Pretreatment   Membrane Type: Dow/Filmtec Model - o Anti BW30 scalant: threshold inhibitor - 400 o Filtration: 1 micron pre - filters Membrane Life: ~5 years  o pH adjustment: sulfuric acid  Membrane Cleaning  Post treatment - Flux decrease initiates CIP o o pH adjustment: sodium hydroxide Every 6 months to 1X per year o o - Chemicals: King Lee (anti  Treatment system footprint scalant ), high/low pH, Silica o RO system: 30’X 188’ Cleaner o IX system: 30’ X 188’ Total system: 60’ X 300’ o Resin Type: Rohm and Hass Amberjet  4400 CL SBA Number of contactors  3 trains o  Volume treated prior to regeneration o 48 vessels/train – 0.8 o 1.4 MGD o RO elements/stage: Regeneration is based on nitrate o  7 elements per stage levels Ion exchange pressure vessels  Salt consumption: 50 tons per week  Number of vessels: 4 o  Volume of brine/backwash 2 Design Loading Rate: 10.1 gpm/ft  o NA  concentration goal for Max. nitrate Resin life: NA  delivered water: 25 mg/L as nitrate (5.7  g Monitorin mg/L as N) (goal), 35 mg/L as nitrate (7.9 Online nitrate analyzers o mg/L as N) (max.)  At source Nitrate concentration goal for the   Blending point treatment system (before blending): 4 mg/L  POE as nitrate (0.9 mg/L as N) o Laboratory samples 1.70 gfd/psi  RO membrane flux rate: 0.3 0 – Residuals Management Concentrate/brine is discharged to an industrial sewer that drains to the Inland Empire Brine Line (IEBL). Waste is transported 45 miles to the Orange Cou nty Sanitation District. Technology Benefits and Drawbacks Benefits Drawbacks  RO works well for nitrate removal  RO is expensive IX is less expensive IX does not accomplish contaminant   removal as well as RO Technical Report 6: Drinking Water Treatment for Nitrate 71

88 Treatment Technology Costs Capital Costs (Ba sed on projected costs in 2004) Treatment Plant Total (10,400 AFY): $19,171,837 $4,346,900 Ion Exchange Treatment: Chino II Desalter: $14,284,500 RO Membranes: $540,437 Ion Exchange Land (not included in $1,730,138 above total): (not included in Ion Exchange SARI Fee $10,105,000 above total): O & M Costs (Based on CDA 2010/11 Budget, for complete plant operation, not just the treatment system) Total: $6,111,799 Chemicals: $615,000 Electricity Plant Total: $2,331,000 – Operating Fees: ncludes SARI fees, permits and other fees) $706,154 (i $1,133,615 Labor ($): Additional Information Operator certification levels range from T  - 3 to T - 5.  Plant shutdown has been required in the past due to chemical pump failure and a high clearwell. mbined system is described as working well but at a high price. Overall the co  Sources * Listed costs are based on: alter Project Chino Basin Desalter Authority. (2005) Presentation: Chino I Desalter Expansion & Chino II Des Update. Chino Basin Desalter Authority. (20 10) Fiscal Year 2010/11 Budget Adoption. * Unpublished sources used in the development of the case studies are not reflected in the References section of this report. 72 Technical Report 6: Drinking Water Treatment for Nitrate

89 3.3 Electrodialysis (ED/EDR/SED) he use of electrodialysis (ED), including electr odialysis reversal (EDR) and selective electrodialysis (SED), T in potable water treatment ( 10 ) has increased in recent years, offering the potential for improved Figure water recovery, the ability to selectively remove nitrate ions , and the minimization of chemical and energy requirements (Kneifel & Luhrs 1988; K apoor & Viraraghavan 1997 ; Hell et al. 1998; Koparal & Ogutveren 2002; Midaoui et al. 2002; Sahli et al. 2008; Banasiak & Schafer 2009 ). Electrodialysis reversal schematic. . Figure 10 Nitrate removal is accomplished by passing an electrical current through a series or stack of anion and cation exchange membranes, resulting in the movement of ions from the feed solution to a , nitrate ions (and other anions) move through the concentrated waste str eam. Illustrated in Figure 11 anion exchange membrane toward the anode. Continuing toward the anode, nitrate is rejected by the anion - impermeable cation exchange membrane and tra pped in the recycled waste stream. Cations can be removed in a similar manner, migrating toward the cathode through the cation exchange membrane - and rejected by the cation impermeable anion exchange membrane. Nitrate selective membranes allow nt without significantly altering the balance of other ions in the water. for treatme The electrical current is passed through the system with the migration of ions across the membranes. For every anion that leaves a compartment, a cation of equivalent charge also le aves, maintaining the charge balance in each compartment. Across the system, the flow of electrons, moving from the cathode to the anode (negative to positive), is governed by the movement of ions through the membrane stack and by the reactions in the ele ctrode compartment. Small levels of gaseous byproducts must be removed. Electrolysis of water generates oxygen at the anode and hydrogen gas at 1995). The the cathode and chloride can be reduced at the anod e, producing chlorine gas (AWWA rtment is rinsed to restore ions for current transfer and to remove unwanted reaction electrode compa products. Technical Report 6: Drinking Water Treatment for Nitrate 73

90 Illustration of electrodialysis membrane stack. 11 Figure . - pressure water, ED has i nherent energy demands. Requiring constant electrical current and low However, voltage adjustment enables selective demineralization. “Plants can be designed to remove from 50 to 99 percent of source water contaminants or dissolved solids. Source water salinities of less than 100 mg/L up to 12,000 mg/L TDS can be successfully treated to produce finished wa ter of less than 1995 , p. 7 10 mg/L” (AWWA ). Case Studies . A detailed case study of an EDR plant in Spain is included in section 3.3.6 Electrodialysis - 74 Technical Report 6: Drinking Water Treatment for Nitrate

91 3.3.1 Electrodialysis Design Considerations - f 8 summarizes key design considerations in Table ED o r nitrate removal from potable water. the use of Table 8 . Summary of d esign c onsiderations for e lectrodi alysis/ e lectrodialysis r eversal.  Use of anion and cation exchange membranes  Selective membranes Membranes o Monovalent versus multivalent  Consider water recovery and frequency of cleaning irect filtration Lower pretreatment requirements because this is not d   EDR systems can avoid or limit chemical use  Prevention of scaling and fouling Pretreatment o Filtration to remove suspended solids o Treatment for iron and manganese removal o Water softening or use of anti - scalants or acid to prevent scaling  pH adjustment to avoid corrosion (if acid used to prevent scaling) Post - Treatment  Disinfection  Possible pH adjustment (acids and bases) Chemical Usage Possible anti scalants -   Possible cleaning chemicals Highly automated   Frequency of membrane cleaning depends on w ater quality and membrane used Polarity reversal (electrodialysis reversal) multiple times per hour o minimizes fouling ED systems can require weekly cleaning o O&M  - filtration system Management of chemicals and pre o Including electrode compartment rinse solution W aste storage and disposal  High monitoring demands  Potentially higher operator demand than IX and RO, due to system complexity  Maximize water recovery while minimizing energy use  umber and Key aspects of the system are pretreatment requirements, n  System configuration of electrodialysis stacks and stages, membrane selection and Components configuration, operating voltage and pressure, reversal frequency (for EDR), gas venting of anode and cathode compartments, disinfection, “brine loop, electrode , concentrate discharge, a nd dosing station” (Hell et al. 1998 , p. 178 ). rinse loop  Concentrate disposal of greatest concern for inland systems o Close proximity to coastal waters is beneficial for brine/concentrate disposal tions include sewer or septic system, reuse for irrigation, drying  Management op Waste site, coastal pipeline, deep well injection and advanced - Management and beds, trucking off Disposal treatment Disposal options can be limited by waste brine/concentrate water quality (e.g.,  volume, salinity, metal s , and radionuclides)  Optimization of recycling and treatment of waste concentrate  Need to prevent membrane scaling and fouling and suspended solids o Hardness, iron, manganese , Limitations Disposal of waste concentrate  High system complexity  75 Technical Report 6: Drinking Water Treatment for Nitrate

92 y Water Qualit Membrane life, cleaning frequency, and pretreatment needs are dependent on feed water quality. Pretreatment may be needed for iron levels above 0.3 mg/L, manganese levels above 0.1 mg/L, and hydrogen sulfide le vels exceeding 0.3 mg/L (WA DOH 2005). Spe cifications for an example EDR system from GE indicate feed water turbidity levels should be < 0.5 nephelometric turbidity units ( NTU ) with typical TDS levels between 100 and 3,000 mg/L (maximum 12,000 mg/L) (GE 2008). SDI limits are generally higher for EDR than for RO, with typical limits of 12 and 4 – 5, respectively (Elyanow & filtration diminishes - Pereschino 2005). Softening may be necessary to reduce hardness, while pre suspended solids. The potential for scaling increases with increasing TDS and i s exacerbated by increased solids precipitation with higher water recovery goals. To minimize fouling/scaling, scaling chemic - als and cleaned with acid (AWWA membranes can be treated with anti 1995). However, in comparison with other membrane processes, f ouling is minimal because the membrane is subjected to the transfer of ions (directed by the electrical current), rather than the transfer of the entire feed stream. Unfortunately, because ED does not serve as a filter (the water does not pass through the filtration in pretreatment membrane), ED fails to remov e microbial contamination (AWWA 1995). Pre - treatment address these concerns. and disinfection in post - To further minimize fouling and thus the need for chemical addition, the polarity of the system can be reversed with electrodialysis reversal (EDR). By reversing the polarity (and the solution flow direction) several times per hour, ions move in the opposite direction through the membranes, minimizing buildup and the need for chemical addition to c ontrol scaling. Biological fouling concerns are lower than other separation processes due to development of membranes that are “more organic resistant and chlorine ( Elyanow & Persechino 2005 , p. 8 ). ED depends on the transfer of an electrical c urrent and is tolerant” therefore more efficient when used for brackish waters. In low conductivity feed waters, the ion removal efficiency declines. In contrast to conventional RO, EDR is unaffected by silica. Considerations System Components and Site ED and EDR systems are operated in stages. Water recovery can be improved with stages operated in series while capacity can be increased with stages operated in parallel. Key aspects of the system are pretreatment requirements, the number and configuration of elect rodialysis stacks and stages, membrane selection and configuration, operating voltage (based on desired removal), reversal frequency (for EDR), gas venting of the anode and cathode compartments, disinfection, “brine loop, electrode rinse loop, concentrate discharge, a nd dosing station” (Hell et al. 1998 , p. 178 ). The membranes used in ED/EDR are anion and cation exchange membranes. Membranes have been designed for selective removal based on valency (monovalent versus multivalent) to screen for particular c onstituents (AWWA 1995). Alternating different selective membranes in the membrane stages can avoid precipitation in the concentrate stream. For example, one stage can remove calcium am), this prevents calc and a second stage can remove sulfate (to an alternate concentrate stre ium 1995). sulfate precipitation (AWWA 76 Technical Report 6: Drinking Water Treatment for Nitrate

93 Residuals Management and Disposal Waste management requirements are similar to RO and IX; however, the burden of disposal in ED/EDR systems is not as significant due to higher water recovery, se lective removal, and the lack of direct s , septic system s , drying beds, off - site trucking, filtration (Reahl 2006). Disposal options include sewer coastal pipeline, deep well injection, reuse for irrigation, and advanced treatment. Important water quality characteristics of the concentrate (e.g., volume, salinity, metals , and radionuclides) can affect the feasibility and costs of disposal options. Several combined configurations of interest are discussed in Section 3.6 Brine Treatment Alternatives and Hybrid Treatment Systems . Maintenance, Monitoring and Operational Complexity , Although ED/EDR systems are amenable to automation, operator demands can be higher than other separation processes ( AWWA 1995) . While ED systems have grea ter pretreatment demands and can require membrane cleaning once a week, EDR systems minimize pretreatment demands and scaling issues, but can still have higher maintenance demands than RO, due to the complexity of the system (Kapoor & Viraraghavan 1997). Appropriate gas venting is important to avoid hazardous conditions ( AWWA 1995 ). Membrane life will depend on water quality and pretreatment measures. However, due to the lack of direct filtration and operation under low pressure, membranes are long lasti ng, and do not require frequent replacement. 3.3.2 Electrodialysis - Cost Considerations For the efficient operation of an ED system, the fundamental objective is to maximize water recovery ing necessary potable water with the minimum amount of energy and chemical usage, while meet guidelines. Factors affecting system cost include facility size (how much water), source water quality (including nitrate concentration and other contaminants), target effluent nitrate concentration, and disposal options. Cap ital costs for ED/EDR systems include land, housing, piping, storage tanks, O&M equipment, cation and anion exchange membranes, preliminary testing (pilot studies), permits, and training. O&M costs include membrane replacement, membrane disposal, concentr ate disposal or treatment, chemical use (limited : anti - scalant , etc.), repair, maintenance, power, and labor. Very little published cost information from existing ED systems used for nitrate removal is available in the literature, due to the limited number of full scale systems. Costs of ED systems are most - comparable to RO. However, in some instances, ED can be the less costly choice due to the greater pretreatment and post - treatment demands (higher chemical use and post - treatm ent pH adjustment) of RO (R eahl 2006). EDR can be chosen over RO when high water recovery is a priority, especially if land must be purchased for concentrate ponds. “ New technology has also reduced the capital and operating cost of EDR nitrate removal by increasing the hydraulic e fficiency of the EDR stacks and pumpin g listed here Costs have been adjusted to 2010 dollars, unless system” (Elyanow & Persechino 2005 , p. 8 ). In a technical paper from GE Water & Process Technologies (the primary supplier indicated otherwise. Technical Report 6: Drinking Water Treatment for Nitrate 77

94 of EDR sy Werner & Gottberg (2005) present O&M costs of an electrodialysis plant in stems in the U.S.), Table ). It is unclear how disposal costs were Suffolk, VA (not specifically for nitrate treatment) ( 9 included in this study. High water recovery in comparison with other removal processes and disposal of waste concentrate to a nearby estuarine tributary would maintain low di sposal costs (Werner & 2005). According to Ameridia, the American division of Eurodia In Gottberg dustrie (a manufacturer of EDR systems), the capital investment for a nitrate treatment EDR unit for a ~0.5 MGD system (in 2005) was $475,000 or $0.94 per gallon of daily capacity ($559,653 or $1.11 in 2010 dollars, respectively) (Ameridia). However, addi tional capital costs are likely not included in this figure. A detailed discussion of treatment costs is included in Section 6 Treatment Cost An alysis . Table 9 . Sample EDR O&M c osts (from Werner & Gottber g 2005). 1 O&M Category EDR (/1000 gallons) Fixed $0.72 ($1.07 adjusted) Professional Services $0.06 ($0.09 adjusted) Chemicals $0.02 ($0.03 adjusted) Utilities $0.21 ($0.31 adjusted) Maintenance $0.17 ($0.25 adjusted) 34 adjusted) $0.23 ($0. Membrane Replacement Production (1997) 827,339,440 gallons $1.41 ($2.09 adjusted) Total O&M Cost 1 Costs adjusted from 1998 dollars to 2010 dollars. The listed cost information is provided as an approximate range of costs for specific facilities. Costs for i mplementing ED may be very different from those listed here. A thorough cost analysis of design parameters for specific locations would be required for accurate cost estimation. - Selected Research 3.3.3 Electrodialysis n desalination applications. Table A.3 of the Appendix list s recent Much research on ED has focused o studies relevant to nitrate removal from potable water and several examples of ED application. - contaminants on system performance and improveme nts in Research is focused on the influence of co exchange membranes, including nitrate selectivity. 3.3.4 Electrodialysis - Summary of Advantages and Disadvantages A summary of advantages and disadvantages of ED in comparison with other treatment options is listed vantages of ED/EDR systems include low chemical usage, long lasting in Table A.6 of the Appendix. Ad membranes, selective removal of target species, flexibility in removal rate (through voltage control ) , good water recovery rate , feasible automation, and multiple contaminant removal ( Prat o & Parent ). With the ability to selectively remove multiple 1993 ; AWWA 1995; Hell et al. 1998; WA DOH 2005 contaminants, ED/EDR systems can be used to address the following constituents: TDS, total chromium, chromium ury, chloride, copper, sulfate, uranium, fluoride, 6, arsenic, perchlorate, sodium, merc - 78 Technical Report 6: Drinking Water Treatment for Nitrate

95 nitrate/nitrite, iron, selenium, hardness, barium, bicarbonate, cadmium and strontium (AWWA 1995 , and GE 2010). Using current reversal, EDR offers additional advantages, improving system performance by “d etaching polarization films, breaking up freshly precipitated scale or seeds of scale before they can cause damage, reducing slime formations on membrane surfaces, reducing problems associated with the use of chemicals, and cleaning electrodes with acid au tomatically during anodic operation” (AWWA , p. 9, 10 1995 ). Additionally, in comparison with RO systems, EDR can treat waters with higher SDI, silica , and chlorine levels (Elyanow & Persechino 2005). Disadvantages of ED/EDR systems include the possible ne ed for pretreatment to prevent membrane scaling and fouling, waste disposal, high maintenance demands, costs (comparable to RO systems), the need to vent gaseous byproducts, the potential for precipitation (especially for high water recovery), high system complexity, and limited manufacturers with U.S. experience (e.g., GE is the primary source of EDR systems for drinking water in the U.S.). Additionally, unlike RO, ED does not remove uncharged constituents in the water. sis 3.3.5 Modifications to Electrodialy Electrodialysis (SED) Selective Since 1997, selective electrodialysis (SED) , d eveloped by Shikun & Binu i , formerly Nitron, Ltd. , has been successfully implemented throughout Israel , reducing national water costs by 55%. The manufacturer minimizing waste volume (Nitron SED offers high water recovery (up to 95%), thereby indicates that as a nitrate treatment 2010). The SED system is accepted by the S . EPA . option for large plants (Nitron U 2009). While similar to traditional ED processes, SED utilizes n itrate selective membranes which have been shown to increase operational performance when used for nitrate treatment. The nitrate selective membranes used in the SED process have been shown to remove up to 70% of sulfate ions and carbonate ions, which have a tendency to nitrate from solution. At the same time, cause scaling issues in the concentrate stream of traditional ED/EDR and RO technologies, are more readily rejected by the nitrate selective membranes used in the SED process. As a result, the sca ling potential is reduced in the concentrate stream. Since scaling problems are minimized, membrane cleaning frequency, maintenance costs, and down time are reduced compared to traditional EDR installations. Another important aspect of membrane selectivi ty is the energy efficiency of the process. Energy efficiency is related to the extent of ion transfer in ED/EDR and SED technologies. In traditional ED/EDR , energy use is less focused, resulting in the removal of many ions, including ions that do not ne - ed to be addressed. SED specifically targets nitrate ions, avoiding energy use for the removal of non target ions and improving energy efficiency. Pretreatment can be limited to filtration, energy efficiency is maximized due to low pressure operation (2 – 4 bars , ~30 – ~60 psi ), chemical use is limited to concentrate treatment, and low maintenance demands are possible due to automation, remote monitoring and control, and infrequent cleaning. In the SED process there is no change in the pH of the product water. This avoids the need for pH ; 2010). Membranes are cleaned Nitron adjustment or remineraliz ation in post - treatment (Nitron 2009 79 Technical Report 6: Drinking Water Treatment for Nitrate

96 – 6 months and membrane life is typically 7 – in place (CIP) for 1 hour every 4 10 years (Nitron 2010). Additional adva ntages of SED include constant membrane performance, no chemical contact with potable water, the simplicity of the system consisting of pre - filtration and membrane stacks (UV can be drawbacks of SED include the 2009b). Potential (Nitron added for disinfection), and a small footprint lack of full - scale application in the U.S. for nitrate removal from drinking water and, unlike RO, ED does not remove uncharged constituents in the water. in Israel is included in the A detailed case study of the use of SED for nitrate removal at locations following section. 3.3.6 Electrodialysis - Case Studies - scale EDR The following case studies provide detailed information on the design and operation of full and SED treatment plants used for nitrate removal. Technical Report 6: Drinking Water Treatment for Nitrate 80

97 System Name: Ga ndia EDR System Location: Valencia, Spain System Type: NA CASE #11 Treatment Type: Electrodialysis Reversal (EDR) Questionnaire completed by: GE Water & Process Technologies Startup Date: 2007 System Description Treatment Type Electr odialysis Reversal 2 systems at 4.7 MGD each System Capacity Gandia is a tourist area on the Raw Water Nitrate 80 mg/L as nitrate Mediterranean coast of Spain. The 18 mg/L as N area sees peak demand during the e summer months when th Spanish Legislature released royal Decree 140/2003 which changed the nitrate limit population almost triples. The to 50 mg/L as nitrate (11.3 mg/L as N). This new law required treatment of the existing system to achieve the new nally, the existing nitrate limits. Additio well systems had deteriorated over time, forcing the municipality to find alternate wells to feed the community. Upon analysis of the wells (old and new), it was determined that the nitrate levels were too high to meet the drinking wat er standard. The well samples had up to 80 mg/L as nitrate (18.1 mg/L as N). Treatment was necessary to produce acceptable levels of nitrate in the product water. An evaluation was conducted and EDR was selected as the technology of choice for a treatment plants. EDR the Gandi offered high recovery while effectively reducing the nitrate levels below 25 mg/L as nitrate (5.6 mg/L as N). EDR was piloted on the wells to verify the nitrate removals and operating cost estimates for power requirements and chem ical consumption. The pilot study was successful, and the final systems were designed around 90% water recovery with the overall nitrate removal of 73%. Source Water Quality Nitrate (mg/L N)  - o < 80 mg/L as NO 3 Treatment Technology Selection ective ED, and Reverse Osmosis were considered for treating the Gandia Wells. EDR was eventually EDR, Sel selected for the high recovery and reduced operating costs compared to RO. SED was ruled out based on the high capital costs of the system. 81 Technical Report 6: Drinking Water Treatment for Nitrate

98 Treatment System Parameters Design Capacity  Water recovery rate  o 94.3% o 3,260 gpm  EDR System No. of modules: 4 o Lines per module: 5 o Stages per line: 2 o Water Quality Results The table below summarizes the water quality of the raw water, finished water, concentrate s tream, and the total waste from the Gandia EDR facility. Total waste includes concentrate blowdown, electrode waste, and off - spec product from the system. Total Concentrate Ion Raw Percent Treated Water Removal Stream Water Waste Ca 82 24.9 44.3 5 772.6 70% 151.0 213.7 Mg 24 8.3 65% 10 Na 23 180 128.1 57% 9.2 K 6.5 1.0 0.3 70% 2074.1 1471.2 HCO 250 99.1 60% 3 SO 58 12.7 424.4 605.2 78% 4 29 Cl 289.3 203.3 7.5 74% NO 60 16.6 411.1 584.4 72% 3 527 TDS 179.3 3339.8 4728.6 66% pH 7.5 7.1 8.1 8.3 Te chnology Benefits and Drawbacks Drawbacks Benefits  EDR membranes are chlorine tolerant,  Higher capital cost than RO . providing means to control biological  System footprint larger than competitive growth. . technologies Relatively low operational expenditures :  o Membrane life expectancy is 15 years. o Low chemical consumption ies. compared to other technolog o Lower energy consumption compared to RO. High water recovery, small concentrate  stream for disposal compared to other technologies. 82 Technical Report 6: Drinking Water Treatment for Nitrate

99 Operating Costs Capital costs for the EDR system were not provided. O & M Costs (Total with explanation or component costs) Unit Cost Labor: $/1,000 gal 0.17 Energy: $/1,000 gal 0.15 Maintenance: $/1,000 gal 0.03 Chemicals: $/1,000 gal 0.10 0. 19 $/1,000 gal Consumables: Overhead: $/1,000 gal 0.04 Total: $/1,000 gal 0.67 Operational Notes Operating costs were based on estimates prior to plant start - up. After four years of operating, the plant has not as commissioned in the Valencia area for 2.9 MGD replaced any membranes. In 2010, another facility (L’Eliana) w production rate for nitrate removal using the EDR technology. Sources * , V.S., Carbonell Cháfer Nitrate and Hardness Removal with Electrodialysis Reversal , J.S., and de Armas Torrent , J.C. (ED R) in Gandia , (Valencia, Spain). Unpublished sources used in the development of the case studies are not reflected in the References section of this * report. 83 Technical Report 6: Drinking Water Treatment for Nitrate

100 System Name: Weizmann Institute System Location: Rechovot, Israel CASE #12 Sy stem Type: NA Treatment Type: Selective Electrodialysis (SED) Questionnaire completed by: Shikun & Binui Environmental Group Startup Date: 2008 System Description Selective Electrodialysis Treatment Type 310 gpm System Capacity The Weizmann Institute of Science Raw Water Nitrate 89 mg/L as nitra te 84 – (Institute), located in Rechovot, – 19 20 mg/L as N Israel, is one of the top - ranking multidisciplinary research institutions in the world. In 2007, the nitrate MCL in the Israeli National drinki ng water regulations changed from change, two of the Institute's wells were is th 90 mg/L (20.3 mg/L as N) to 70 mg/L (15.8 mg/L as N). Because of removed from the drinking water supply and the Institute had to rely on external water suppliers. Over time, municipal and national water costs increased. The Institute’s management looked for solutions to solve nitrate problems and enable them to reopen the Institute's wells. Prior to treatment the wells were used for potable purposes and the Institute’s ir rigation needs. Selective , n SED system was Electrodialysis (SED) was identified as the Institute’s most appropriate treatment technology. A implemented for the 310 gpm well. The Institute opted for treatment of one well and uses the second well as a dedi cated irrigation supply source. The nitrate enriched concentrate from the SED process is fed into the non - potable irrigation system where the nitrate enhances plant growth. SED has been successfully implemented throughout Israel since 2007. Developed b y Nitron, Ltd., SED offers high water recovery (up to 95%), thereby minimizing waste volume. Pretreatment can generally be limited to filtration, energy efficiency is maximized due to low pressure operation (30 60 psi), chemical use is limited to concen – trate treatment (no need for chemical addition to feed or product water), and low maintenance demands are possible due to automation, remote monitoring and control and infrequent cleaning. Membranes are cleaned in place (CIP) 6 months and membrane life is typically 7 – 10 years. for 1 hour every 4 – Source Water Quality  Nitrate – mg/L as nitrate (mg/L N) – 84 (19) Average o Minimum – 84 (19) o – 89 (20) o Maximum Treatment Technology Selection SED and RO were considered for treating the Weizmann Instit ute well. A 10 year life cycle cost analysis that included capital and operations costs identified SED as the more cost - effective solution. Technical Report 6: Drinking Water Treatment for Nitrate 84

101 Treatment System Parameters Design Capacity Water recovery rate   94.3% o 310 gpm o Pretreatment  SED unit information:  Number of SED units: 1 o o Cartridge filtration o Membrane pairs: 240 Post  treatment - o 59% Nitrate reduction nation Chlori o 30% TDS reduction o – Acid addition (pH 4.5 o 5) to Energy consumption: 2.3 o concentrate to prevent the KWh/1,000 gal precipitation of calcium carbonate and calcium sulfate in Monitoring:  Online nitrate analyzer o the concentrate cells o Laboratory nitrate samples Treatment system footprint  Online o pH meters o Treatment system: 8’ x 5’ Online turbidity o o Building footprint: Butler building Online conductivity meters o - lient request 50’ x 13’ as per c Water Quality Results The table below summarizes the water quality of the raw water, finished water, and concentrate stream from the Weizmann Institute. Ion Raw Concentrate Treated Percent Water Remo val Water Stream 2051 45.9% Cl 194 105 SO 92 3.0% 140 95 4 HCO 232 165 28.7% 442 3 44 92 52.2% NO 835 3 109 76 30.4% 627 Na Ca 118 73 37.8% 808 160 Mg 24 15 36.6% 20.5 K 30.6% 3.6 2.5 Ba 0.149 0.093 37.8% 1.020 Sr 0.84 0.52 37.8% 5.7 5 869 574 TDS 5090 33.9% 6.5 7.8 7.8 pH Residuals Management The concentrate from the SED process is fed into the Weizmann Institutes non - potable system which is used for irrigation purposes. Since the SED process selectively removes nitrate, the Total Dissolved Solids (TDS) of the RO system would be if it were treating the same water, which allows the concentrate to n concentrate is less than a s also be used for irrigation without the salinity adversely affecting plant growth. This management approach i beneficial since the concentrated nitrate solution has limited the amount of fertilizer applied by the Weizmann Institute. 85 Technical Report 6: Drinking Water Treatment for Nitrate

102 Technology Benefits and Drawbacks Benefits Drawbacks   Ease of regeneration Capital intensive technology Fewer chemicals than comparable Requires specific operator training   technologies 10 year life span  Membranes 7 – Energy consumption less than that of RO   High water recovery  Concentrate solution has relatively low TDS which may increase disposal options Treatment Technolog y Costs Capital Costs (Total with explanation or component costs) Comments $ 650,000 Total: Light building 50X13 feet. According to customer demands. Housing: Pipes, electric valves, storage tanks (the largest would be with a 50,000 Piping: volum . e of about 180 cu. ft. for product water) Storage Tanks (include See above. description of uses): Operating and Monitoring Control system, remote assistance for the control system, on line 90,000 Equipment: nitrate measurement, conductivity, pH, turbid ity. 300,000 Complete SED membrane stack. Membranes Modules: According to local regulations. Permits: ; manufacturing & , design Design installation; Electricity boards – 210,000 Other: Pumps and blowers. t costs) O & M Costs (Total with explanation or componen Unit Comments %/year 10% per year is the replacement rate . Membrane: 5% of the SED system 4 ; Site specific – Gallons/year Concentrate Disposal or Treatment: capacity. Chemicals: Total, [lb/1000 gallon] 0.77 Specific acid consumption [lb/10 00 gallon] 0.72 Specific chlorine consumption [lb/1000 gallon] 0.04 0.01 Specific caustic soda consumption [lb/1000 gallon] Repair/Maintenance (not including $/year 7,000 Labor): Specific power consumption [kWh/1000 gallon] 1.4 Labor ($): $ 16, 000 (40$ per hour basis) hours 400 Labor (Hours per Year): 86 Technical Report 6: Drinking Water Treatment for Nitrate

103 Operational Notes The SED system is highly instrumented and the PLC has over 200 monitored inputs. As a result there have not been any failures that have resulted in MCL violation. While ther e have been failures resulting in alarms and shutdowns, the control system has been robust enough to shut down the system and prevent water with high mal nitrate entering the distribution system. Typical shutdowns can be rectified in a matter of hours and nor operation is resumed. Sources * (2010) Completed questionnaire. Merhav, Neta. . October, 2010 Unpublished sources used in the development of the case studies are not reflected in the References section of this * report. 87 Technical Report 6: Drinking Water Treatment for Nitrate

104 3.4 Biological Denitri fication (BD) Commonly used in wastewater treatment, biological reactors are emerging as a method for denitrification of potable water with the potential to address multiple contaminants including nitrate, trace organic chemicals (Brown 2008). Biological denitrification (BD) in chromate, perchlorate, and in Europe since 1804 (Lenntech 2009), with recent full - potable water treatment has been implemented scale systems in France, Germany, Austria, Poland, Italy , and Great Britain (Meyer 2009 ; Dördelmann 2009) . To date, full scale drinking water applications in the United Sates have been limited to a single - - plant in Coyle, OK (no longer online). However, two full scale biological denitrification systems are anticipated in California within the next couple of years. Denitrification occurs naturally in the environment as part of nitrogen cycling. Application of biological denitrification to potable water treatment ( Figure 12 ) utilizes denitrifying bacteria to reduce nitrate to innocuou s nitrogen gas in the absence of oxygen (anoxic conditions). Figure 12 . Biological denitrification schematic. The reduction of nitrate proceeds stepwise in accordance with Eqn. 9. In contrast with the separation , RO processes of IX and ED/EDR, nitrate is reduced and thereby removed from the system rather than , simply being displaced to a concentrated waste stream. - -  NO NO  NO  N O  N (Eqn. 9) 2 2 2 3 Denitrifying bacteria require an electron donor (substrate) for the redu ction of nitrate to nitrogen gas. In conventional wastewater treatment, substrate addition is not typically needed, because the wastewater contains sufficient carbon for denitrification to occur. However, depending on the source, substrate addition is of ten required for the biological denitrification of potable water. The addition of a carbon substrate in potable water treatment is somewhat counter intuitive. In fact, one principal objective of potable water treatment is to minimize dissolved carbon in the water to minimize growth of microbes (e.g., biofilms) and production of disinfectant byproducts (e.g., THMs). Feed water composition may need to be further augmented with the addition of nutrients required for cell growth (phosphorus for example). Au totrophic bacteria utilize sulfur or hydrogen as an electron donor and inorganic carbon (typically carbon dioxide) as a carbon source for cell growth (Eqns. 10 and 11), while heterotrophic bacteria consume an organic carbon substrate, like methanol, ethano l , or acetate ( Eqn. ; ). Kapoor & Viraraghavan 1997 12) (Mateju et al. 1992 88 Technical Report 6: Drinking Water Treatment for Nitrate

105 Eqns. 10 through 12 illustrate the overall denitrification reaction defining the stoichiometric relationship between electron donor, carbon source , and nitrate in the production of cells and the conversion of nitrate to nitrogen gas. Not all nitrogen is converted to nitrogen gas. Some nitrogen is required for cell growth. The governing stoichiometric equation indicates the necessary dose and varies with the substrate used. For e xample, the stoichiometric factor for acetic acid is 0.82 moles of acetic acid per mol e of nitrate (Dördelmann et al. 2006). 0 + - S + 0.5 CO 11 + 10 NO NH + 2.54 H 1.71 O + 2 3 4 2 2 - + 0.92 C H O N + 11 SO (Eqn. 10) + 5.4 N + 9.62 H  7 2 5 2 4 - + + 0.35 NO 0.010 C + 0.35 H (Eqn. 11) + 0.052 CO  H O H + 1.1 H O N + 0.17 N 2 2 7 2 5 2 3 2 - + O OH + NO + 2.44 H (Eqn. 12) + H 1.08 CH  0.065 C H + 0.76 CO O N + 0.467 N 2 7 5 3 2 3 2 2 Thiobacillis de Various species of bacteria are responsible for denitrification including nitrificans, Pseudomonas maltophilia and P seudomonas putrefaciens (Kapoor & Micrococcus denitrificans, Viraraghavan 1997). Due to slower bacterial growth rates, autotrophic denitrification offers the advantage of minimizing biomass accumulation; however, autotrophic denitrification requires alkalinity cell growth (Della Rocca et al. 2006). to supply the inorganic carbon source (carbon dioxide) for BIODEN® and DENICARB® are heterotrophic biological denitrification processes, while DENITROPUR® is phic option. Selected full - scale biological denitrification systems are listed in Table 10 an autotro (Dördelmann 2009). 1 t . Full - scale b iological d enitrification s ystems for p otable w ater 10 r eatment. Table Flow rate Location Reactor Configuration Substrate, Denitrification type 3 m /h (MGD) Germany 150 (0.95) Acetic acid, Heterotrophic Neuss Fixed bed, down - flow Frankfurt Airport flow, DENICARB® Ethanol, Heterotrophic - 320 (2.03) Fluidized bed, up Aschaffenburg 1600 (10.14) , Hydrogen and CO 2 - flow, DENITROPUR® Fixed bed, up Autotrophic Föhr Island 90 (0.57) Austria Ethanol, Heterotrophic 180 (1.14) Obersiebenbrunn Fixed bed, down - flow, BIODEN® Poland Czestochowa Fixed bed, down - flow, BIODEN ® Ethanol, Heterotrophic 500 (3.17) 1 2009 Dördelmann . In their review of potable water treatment methods for the removal of nitrate, Mateju et al. (1992), Kapoor & Viraraghavan (1997), Soares (2000) , and Shrimali & Singh (2001) discuss previous researc h and applications of biological denitrification. Problems associated with the application of conventional biological denitrification to potable water treatment include additional post - treatment for removal of biomass and dissolved organics, the potential for incomplete denitrification, increased capital costs, and nditions (Kapoor & Viraraghavan 1997). To address these concerns, sensitivity to environmental co several treatment configurations using biological denitrification have been developed. Technical Report 6: Drinking Water Treatment for Nitrate 89

106 3.4.1 B - Design Considerations iological Denitrification 11 summarizes key design considerations in the application of biological denitrification . Table 11 . Summary of d esign c onsiderations for b Table d enitrification for nitrate removal in potable water . iological  Substrate and nutrient dosing Pretreatment  pH adjustment Carbon adsorption for organic carbon removal  o Carbon adsorption is not always required Residual substrate removal can be accompli o shed via biological filtration Post - Treatment Aeration  Filtration   Disinfection  Possible pH adjustment  Substrate and nutrient addition Chemical Usage  Coagulant/polymer use to meet turbidity standards  Disinfection  Historically operator intensive o Operator demands can be minimized with latest design configurations , health microbe Constant monitoring required to assure efficient removal, etc.  Monitoring of nitrite and ammonia will also be necessary due to the potential for  incomplete denitrification O&M cals  Management of chemi Waste sludge storage and disposal  New plants can be highly automated   Historically viewed as operationally complex o More unit processes than IX New design configurations can minimize complexity (e.g., fixed bed) o con siderations include (Dördelmann 2009): Important process  Dosage requirements of substrate and nutrients  Reactor configuration and governing equation of the biological process  System Aeration to remove nitrogen gas and provide oxygen Components  Filtration to remove particulate matter Acti vated carbon may be used to remove substrate residual and avoid DBP  formation (for heterotrophic systems)  Disinfection Waste Sludge disposal Biological solids and residual organic matter  – Management and  No waste brine or concentrate as in separa tion processes Disposal  Requires anoxic conditions Chemical management   Few examples for nitrate removal in the U.S. Limitations treatment requirements  Post - Operator training   Intermittent use of wells may be challenging due to the need for acclimation of isms microorgan 90 Technical Report 6: Drinking Water Treatment for Nitrate

107 Water Quality Anoxic conditions are required for denitrification to occur. In the presence of oxygen (> 0.1 mg/L) bacteria preferentially reduce oxygen rather than nitrate, diminishing the efficiency of the process. For all configurations, the opt imal growth of microbes must be considered. Control and monitoring of water quality characteristics including temperature, pH, salinity, and oxidation reduction potential (ORP) can cation system (WA DOH 2005). be fundamental to the stability and efficiency of the biological denitrifi o o – 8) and temperatures below 5 For biological denitrification, near neutral pH is preferred (7 C/41 F can inhibit denitrification (WA DOH 2005). Pretreatment will include addition of substrate and nutrients in the appropriate d ose while post - treatment requirements can include coagulant addition, filtration, gas exchange, and disinfection for the removal of biomass, particulates and substra te resid uals (Panglisch et al. 2005 ; Dördelmann et al. , ). 2006 System Components and Site C onsiderations Important process considerations in the design and operation of BD systems include (Dördelmann 2009):  Dosage requirements of substrate and nutrients Reactor configuration and governing equation of the biological process   trogen gas and provide oxygen Aeration to remove ni Filtration to remove particulate matter   Activated carbon to remove substrate residual and avoid DBP formation (for heterotrophic) Disinfection  nitrate reduction to Conditions should be optimized to ensure complete denitrification. In addition to - nitrite meet the nitrate MCL (45 mg/L as NO , 10 mg/L as N), effluent nitrite levels must not exceed the 3 MCL of 1 mg/L as N. Incomplete denitrification, which can be associated with higher dissolved oxygen (DO) levels, can result in O. Dissolved the production of greenhouse gases (GHG) such as NO and N 2 oxygen levels can be decreased using reducing agents or through the provision of sufficient electron donor to enable depletion of DO (Meyer et al. 2010). System configurations of biol ogical denitrification include: fluidized bed reactor, fixed bed reactor, membrane biofilm reactor, and bio - electrochemical reactors. In situ options (including bank filtration) have also been explored in the research. Fixed Bed flow systems in a pressurized or open flow ors can be in up - flow or down - Fixed bed biological react and sand, plastic, include configuration (Brown 2008). Typical options for growth support media 91 Technical Report 6: Drinking Water Treatment for Nitrate

108 gr 2008). Accumulation of biomass in the media leads to head loss anular activated carbon (Brown - requiring periodic backwashing. Post treatment requirements can include filtration, gas exchange, and , and substra te residuals ( Panglisch et al. 2005 ; disinfection for the removal of biomass, particulates Dördelmann et al. 20 06 ). The fixed bed configuration “is often coupled with pre - ozonation to improve the removal of organic material, which reduces regrowth potential and DBP formation in distribution systems” (Brown 2008 , p. 137 ). (See Soares (2002), Panglisch et al. (2005) , Aslan (2005) , Dö rdelmann et al. (2006), Carollo Engineers (2008) Upadhyaya (2010), Meyer et al. (2010), and City of Thornton (2010) , , in Table A.4 for research examples of the fixed bed configuration.) A detailed case study of the Fixed Bed configuration of biological den itrification in Riverside, CA , is included in S ection 3.4.5 Biological Denitrification - Case Studies . Fluidized Bed The fluidized bed reactor operates in an up - flow mode, resulting in granular growth support media expansion. Fluidizing the granular media offers several advantages over the fixed bed configuration. Flow resistance is minimized and the system does not need to be taken off line for backwashing because accumulated biomass is removed by the fast flowin g feed water and/or “in - line mech anical shearing ). “The biofilm is detached from the support material only upon strong , p. 139 2008 devices” (Brown mechanical effect, thus the excess biomass can be intermittently removed from the reactor, independently of the purified water” (Holló & Czakó 1987 ). , p. 418 Maintenance of the sufficient up - flow velocity can be achieved through recycled flow and reactor volumes are designed for a typical be d expansion of 25 to 30% (Brown 2008). (See Kurt et al. (1987), Holló & Czakó (1987) , and Webster & Togna (2009), in Table A.4 for research examples of the fluidized bed configuration.) A detailed case study of the Fluidized Bed configuration of biological denitrification in Rialto, CA is . included in S ection 3.4.5 Biological Denitrification - Case Studies Membrane Biological Reactor (MBR)/Membrane Biofilm Reactor (MBfR) With the addition of membrane technology to conventional biological denitrification, common concerns of biological treatment can be minimized through physical separation of biomass and substrate from the treated water. In a comprehensive review of membrane bioreactors, McAdam & Judd (2006) present the pros and cons of a variety of MBR configurations ( Table 12 ). MBRs can be designed for autotrophic or heterotrophic denitrification. Pressurized systems have been explored using submerged ultrafiltration membranes or external (sidestream) MBRs, while pressure neutral diffusion sy stems have been implemented with ion exchange membranes a porous membranes (McAdam & Judd nd micro 2006 ; Brown 2008 ). Membrane types include hollow fiber, ion exchange, microporous , and flat sheet. (See Mansell & Schroeder (2002), Ergas & Rheinheimer (2004 ), Nerenberg & Rittman (2004) , Chung et al. in Table A.4 for research examples of MBR City of Thornton (2010) (2007), Meyer et al. (2010), and configurations.) 92 Technical Report 6: Drinking Water Treatment for Nitrate

109 Applied Process Technology, Inc. (APT) has developed an autotrophic membrane biofilm reactor (MB fR) N.D. Early (Gormly & Borg ) using hydrogen gas as the electron donor rather than a carbon substrate. pilot and bench - scale studies of the APT MBfR raised several key challenges requiring optimization of - up of excess biomass (Rittmann 2007). As part of a long the system, including the need to avoid build 19 term pilot - in Glendale, AZ, APT’s MBfR was examined to address high nitrate levels in scale study groundwater. This autotrophic biological denitrification system successfully reduced nitrate levels to below the MCL. Three types of hollow fiber membranes were examined for substrate delivery. Operational concerns highlighted by t his study include (Meyer et al. 2010):  Problems with leaking fibers,  ydrogen sulfide formation due to excessive hydrogen ga s pressure, h  a mmonium generation from biomass decay due to operational interruption and insufficient electron donor, and  n itrite levels above the 1 mg/L nitrite as N limit (incomplete denitrification). To address these concerns, the authors suggest:  The us e of the latest optimized membranes,  c onsistent and adequate nutrient and electron donor supply, - o xidation of nitrite in post treatment if necessary,   s table loading and continuous operation to avoid system upset, and p nce and repair.  arallel reactors to allow for maintena flow The MBfR pilot was one of three biological configurations examined in Glendale. Two up - heterotrophic fixed bed bioreactors were also examined, each with a different media type. Post - tive carbon and ozonation. The two most promising treatment included filtration using biologically ac - flow fixed bed bioreactor with plastic media, were biological treatment options, the MBfR and the up compared with each other and also with an IX system. Overall, using a multi - criteria analysis with ideration of sustainability, the MBfR scored the most favorably regarding benefits, but the least cons favorably regarding life cycle costs. The life cycle costs of both the IX and the fixed bed bioreactor options were lower than that of the MBfR. The Glendal e study highlights several key areas of future - down test where biological treatment processes sit dry for a period of time research including “a shut and then re - - acclimation test where systems are re - start at optimal hydraulic loading rates [and] a re iated after losing al init 2010 , p. 164 ). l viable biomass” (Meyer et al. As a promising technology, with further research and design optimization to reduce costs, the MBfR may become a more feasible treatment option. As of 2008, a demonstration project of th e MBfR for nitrate and d ibromochloropropane (DBCP) removal was under consideration for the City of Fresno, CA, to test performance of the MBfR as an alternative t reatment option (City of Fresno 2008). 19 Project partners and participants: City of Glendale, Arizona, Water Research Foundation, Arizona State University, CH2M HILL, Applied Process Technology, Inc., Intuitech, Inc., KIWA Water Research, and Layne Christensen. - Technical Report 6: Drinking Water Treatment for Nitrate 93

110 12 . Membrane b iological r eactor c onfigurations. Table Relying on diffusion for nitrate transfer through the membrane, extractive i.e., Fixed MBRs) do not directly filter the treatment stream, but they do MBRs ( Diffusive Extraction and provide separation of the denit rification chamber (substrate, biomass , Microporous Membranes associated residuals) and the treatment stream. Key issues are the transfer of biomass or substrate across membrane, the potential for fouling/scaling, and the need for a high transfer rate of nitrate to the den itrification compartment. With the use of an ion exchange membrane rather than a microporous membrane, selectivity for nitrate and decreased mass transfer from the Diffusive Extraction Ion Exchange Membranes denitrification compartment are facilitated; h owever, capital and maintenance costs of ion exchange membranes can be significant and the need to manage membrane fouling persists (McAdam & Judd 2006). gaseous substrate In autotrophic systems, membranes can be used for 2006). Previous problems with ., hydrogen gas) (McAdam & Judd delivery, (i.e limited hydrogen gas transfer have been addressed with the use of hollow fiber membranes for delivery. Substrate passes through the membrane to the Gaseous Substrate Delivery Hollow Fiber Membranes biofilm on t he outer membrane surface. The membrane does not separate the biomass from the treatment stream and does not directly filter the treatment stream (McAdam & Judd 2006); the presence of “sloughed biomass” and biological residuals in the treatment stream req uires further treatment downstream. Pressurized MBRs provide the advantage of direct filtration. Pressure is used to draw the denitrified water through the submerged membrane, leaving Pressure Driven sirable constituents. However, the use of this behind biomass and other unde Direct Filtration MBRs configuration for denitrification is complicated by the fact that aeration is typically used for mixing and to minimize fouling of the e xternal membrane 2008). surface (Brown In Situ Denitrification Bank filtration refers to the withdrawal of surface water through an embankment. The porous media (soil) of the bank serves as a biological reactor providing treatment through “filtration, dilution, iodegradation processes” (Brown 2008 , p. 140 sorption, and b Bank filtration was employed in water ). treatment as early as 1870 along the Rhine River in Germany (Brown 2008). As a recent example, a full - , demonstrated effective nitrate removal with bank filtration of surface water scale study in Aurora, CO South Platte River (Waskom, Carlson & Brauer N.D.). Bank filtration has also been from the , and Des Moines, IA ( Jones et al. implemented to address nitrate impacted waters in Saxony, Germany 2007 ; Grischek et al. 2010 ) . Biological denitrification for remov al of nitrate from groundwater was explored by Hunter (2001), in situ , Schnobrich et al. (2007) denitrification, the . Through in situ Haugen et al. (2002) , and many others subsurface acts as the porous media through which water is filtered. Residual organics and biomass from denitrifiers can thus be removed naturally. (See Hunter (2001), Haugen et al. (2002), and Schnobrich et al. (2007) in Table A.4 for research examples of in situ applicat ion.) See Te chnical Report and 5, Section 2 for a discussion of in situ den itrification , permeable reactive barriers, phytoremediation , . (King et al. 2012) other remediation methods Technical Report 6: Drinking Water Treatment for Nitrate 94

111 Post Treatment Requirements - Filtration/Taste & Odor /Disinfection - With the use of microorganisms and the addition of a carbon substrate, post tr eatment is essential to - meet turbidity standards, to remove biomass and residual organic matter, and to address taste and odor concerns. Post - treatment must include disinfection to address biological contamination and can also include dual media filtratio n and/or activated carbon filtration and aeration (individual state regulations will need to address local requirements). Residuals Management and Disposal In contrast to the concentrated waste stream from removal processes, biological denitrification ha s limited waste demands due to the conversion of nitrate to nitrogen gas. Waste sludge, consisting of biological solids and residual organic matter, requires appropriate disposal; however, with nearly 100% water recovery, the low waste volumes are not a s ignifican t burden (Kapoor & Viraraghavan 1997). Maintenance, Monitoring, and Operational Complexity Because biological denitrification is microbially mediated, to maximize performance, systems should be run continuously, with a consistent supply of substra te and nutrients at the appropriate dosage (Dördelmann 2009). An initial startup period may be necessary for development and acclimation of the 1987; Aslan microorganisms (Holló & Czakó 2005). This may be problematic for intermittent use of wells and was , and ting may be required for acclimation to occur. Backwashing, consistent maintenance regular monitoring of product water quality are also essential. Constituents that should be monitored frequently are nitrate, nitrite, pH, oxygen, turbidity, conducti vity, dissolved organic carbon, and bacterial count (Dördelmann 2009). Operation and maintenance demands of biological denitrification systems typically exceed those of alternative treatment technologies. However, these systems are more sustainable becau se nitrate is reduced to innocuous nitrogen gas rather than concentrated in a waste stream that requires costly disposal (Dördelmann 2009). 3.4.2 Biological Denitrification - Cost Considerations For efficient operation of a biological denitrification sys tem, maintaining optimal conditions for the bacteria is essential, as is balancing the appropriate substrate and nutrient dose and managing pre and post - treatment while meeting necessary potable water guidelines. Factors affecting system cost include faci lity size (flow rate), source water quality (including nitrate concentration), environmental factors (temperature and pH), target effluent nitrate concentration, and possible wasting due to intermittent use of wells and associated acclimation of microorgan isms. Capital costs for biological denitrification include land, housing, piping, storage tanks, O&M equipment, preliminary testing (through extensive pilot studies), permits, and significant operator training. O&M costs include post - treatment, sludge dis posal, chemical use (pH adjustment, substrate and nutrient ), repair, extensive monitoring and maintenance, power, and labor. Costs can be higher in certain dosing treatment requirements. - states, depending on post 95 Technical Report 6: Drinking Water Treatment for Nitrate

112 Very little published cost information fro m existing biological denitrification systems for drinking water - scale systems ( 13 ). Costs have been Table is available in the literature, due to the limited number of full The listed cost information is provided as an ed otherwise. adjusted to 2010 dollars, unless indicat approximate range of costs for specific facilities. Costs for implementing biological denitrification may be very different from those listed here. A thorough cost analysis of design parameters for specific locations would be required for accurate cost estimation. The information gathered through the questionnaire includes detailed costs associated with the individual case studies included in this analysis. A detailed discussion of treatment c 6 Treatment Cost An alysis . osts is included in Section Table 13 . Cost i nformation* for b iological d enitrification of p otable w ater. 5 MGD System Flow** 0.5 – < 0.5 MGD 5+ MGD *** 00 gal) 0.83 [1] 0.61 – 0.80 [2, 3] Annualized Capital Cost ($/10 0.51 – 0.62 [4] *** – O&M Cost ($/1000 gal) 0.30 [1] 0.33 – 0.46 [2, 3] 0.74 0.94 [4] *** – 1.13 [2, 3] Total Annualized Cost ($/1000 gal) 1.03 1.25 – 1.56 [4] 1.13 [1] * Costs have been adjusted to 2010 dollars wit h 7% interest over 20 years, unless indicated otherwise. ** When available, costs are based on actual system flow rather than design capacity. *** Listed costs are based on biological treatment for perchlorate and should be considered only as a rough ate of similar systems for nitrate treatment. estim [2] Webster & Togna (2009). [3] Carollo Engineers (2008). [1] Silverstein (2010), not adjusted to 2010 dollars. [4] Meyer et al. (2010). 3.4.3 Biological Denitrification - Selected Research s Table A.4 of the Appendix list recent research studies relevant to the use of biological denitrification in potable water treatment. - Summary of Advantages and Disadvantages 3.4.4 Biological Denitrification nitrification in comparison with other A summary of advantages and disadvantages of biological de treatment options is listed in Table A.6 of the Appendix. Advantages of the use of biological denitrification for nitrate removal from potable water include high water recovery, no brine or concentrate waste stream ( reduction of nitrate rather than removal to a concentrated waste stream), low sludge waste, less expensive operation, limited chemical input, multiple contaminant removal, and increased sustainability (WA DOH 2005; Brown 2008; Upadhyaya 2010). Disadvantage - treatment requirements for s of nitrate removal using biological denitrification are post the removal of biomass and dissolved organics, high capital costs, potential sensitivity to environmental conditions (although recent pilot tests indicate robust newe r designs), system footprint larger than typical IX systems, high system complexity (may be simplified with new configurations), lack of full - scale systems in the U.S., potential for incomplete denitrification and GHG production, pilot study 2005). WA DOH requirements, and slow start - up (Kap oor & Viraraghavan 1997 ; 96 Technical Report 6: Drinking Water Treatment for Nitrate

113 3.4.5 Biological Denitrification Case Studies - - scale and The following case studies provide detailed information on the design and operation of a pilot scale biological treatment sy stems that can be used for nitrate removal. planned full - System Name: West Valley Water District System Location: Rialto, CA PWSID: CA 3610004 3 System Type: Demonstration System, Full - CASE #1 scale Installation is under construction Treatment Type: Bi ological Denitrification, Fluidized Bed Bioreactor (FBR) Questionnaire completed by: Todd Webster from Envirogen Technologies, Inc., Director of Sales & Bioreactor Applications. Tom Crowley from the West Valley Water District. eration: 2007 – 2008, Full - Demonstration Dates of Op scale Installation Proposed Startup Date: 2012 System Description Treatment Type Fluidized Bed Bioreactor The West Valley Water District 4,000 gpm – 2,000 System Capacity (District) utilizes 2 surface water - (for proposed full scale treatment) sources and 5 groundwater wells. Raw Water Nitrate Of these supplies, one (abandoned well) 19 17 – mg/L as nitrate roundwater well is impacted by g nitrate contamination above the 5 – mg/L as N ~4 MCL, with an average nitrate (pilot) – 27.9 mg/L as nitrate 27.5 18 concentration of mg/L of 6.2 – 6.3 mg/L as N nitrate as nitrate ( ~4 mg/L of Ra w Water Perchlorate nitrate as N). The well source is g/L (pilot) u 50+ the Chino Basin which has an 400K estimated capacity of 300 – ac. ft. For the immediate future this well has been abandoned due to nitrate contamination, while feasible treatment options are being explored. The primary water quality concern in the District is perchlorate explored principally to address perchlorate levels typically in the contamination. Biological denitrification has been range of 50 – 53 μg/L. However, simultaneous nitrate reduction in the biological denitrification process makes this system an appropriate example as a nitrate treatment alternative. Un like the removal processes of IX, RO , and EDR, biological denitrification allows for the destruction of nitrate through reduction to innocuous nitrogen gas. The fluidized bed configuration maximizes media surface area for - - the growth of denitrifying bacter ia. After a successful 1 scale year demonstration study on well Rialto #2, a full Fluidized Bed Bioreactor (FBR) is expected to be online in early 2012. Technical Report 6: Drinking Water Treatment for Nitrate 97

114 2009 13 FBR configuration . ( Source: reprinted with permission, Webs ter & Togna Figure . . ) “The system is the fluidized bed bioreactor (FBR). The contaminated feed water is pumped from the wellhead and fed directly into a recycle line of the reactor. The feed and recycle water enters the vessel through an inlet header at t he bottom of the reactor and is distributed thro ( Figure 13 ). The fluid passes ugh lateral piping and nozzles upward through the media, causing the media to hydraulically expand approximately 28% of the settled bed height. Through a self - inoculating process from the contaminated feed water, microorganisms attach on to the fluidized media. Adequate quantities of electron donor (i.e., acetic acid) and nutrients are added to the reactor. Utilizing this electron donor and the nutrients, the attached microorganisms perform an oxidation/reduction reaction in consuming all of the dissolved oxygen, nitrate, and perchlorate. As the microorganisms grow, the amount of attached microbes per media particle also increases. Sin ce the microbes primarily consist of water, the volume of the microbe/media particle increases, but the specific density decreases. This allows the media bed to taminant removal. expand and fluidize further such that longer hydraulic retention times can be achieved for con The treated fluid flows into a submerged recycle collection header pipe and the effluent collection header pipe at the top of the reactor. A portion of the fluid exits the FBR system to a post - aerator while the balance is recycled back to the suction of the influent pump. An in - bed biomass separation device controls bed height growth by physically separating biomass from the media particles. Typically, a bed expansion of 40 – 60% of the settled bed height is targeted. Any excess bioma ss that is separated from the media exits the system through the effluent coll ection system” (Webster & Togna 2009 , p. 6 ). Source Water Quality Co contaminants: Perchlorate -  Nitrate – mg/L as nitrate (mg/L as N)  – 27.9 (6.2 – 6.3) 53 μg/L, spiked to – Demonstration well: 27.5 Demonstration well: 50 o o l el Abandoned nitrate impacted w o 1000 μg/L (with appropriate substrate and 18 ( ~4 )  Average: nutrient adjustments) Abandoned nitrate impacted well: a few ppb o Treatment Technology Selection The fluidized bed bioreactor was selected for perchlorate removal; however, nitrate reduction is also - accomplished. Due to the levels of nitrate and perchlorate, biological denitrification was deemed to be more cost - Generally, with nitrate levels well above the MCL (e.g., chemical technologies. effective than alternative physico 100 mg/L as nitrate), removal processes can become more costly. Biological denitrification can be the more feasible option for source water containing co . contaminants and/or high levels of nitrate - For the Rialto #6 and West Valley Water District #11 wells, the fluidized bed reactor and IX were considered. Two additional pilot studies were performed to assess nitrate and perchlorate treatment using a packed bed bioreactor and zero valent iron. Technical Report 6: Drinking Water Treatment for Nitrate 98

115 D emonstration System Parameters   Design Capacity Water Recovery: > 99% 50 gpm capacity o Bed Expansion  Pretreatment o  28% of settled bed height Substrate and nutrient addition o – Max: 40 o 60%  Acetic Acid:  Media: Sand or GAC 25%) – 16.2 mg/L as C (+20 Cleaning requirements  15 mL/min, 50% acetic acid Biomass separator o o Nutrient addition bed cleaning eductor - n I o  Phosphoric Acid:  Manufacturer: 0.3 mg/L as P Envirogen Technologies, Inc. 10.5 mL/min Monitoring: Throughout the demonstration  treatment  Post - using online nitrate and perchlorate Aer o ation tank analyzers  Increase dissolved oxygen Clarifier/multimedia filter o GAC filtration o o UV disinfection versus chlorination Scaling up from 50 gpm to 1000 gpm - Demonstration System Cost Estimation ars) (1000 gpm) Capital Costs (2008 doll Total Equipment Costs : $1,966,000 $570,140 : Total Contractor Costs Total Home Office Costs : $664,000 Total Installed Capital Costs (1000 gpm) : $3,200,140 Amortized Capital Cost ($/AF) : $128 (30 years, 4.9% bonding rate) O & M Costs (2 008 dollars) (1000 gpm) Electricity ($/yr) $87,600 : Chemicals ($/yr) : $133,187 $20,000 Maintenance ($/yr) : Total Operating Costs ($/yr) $240,787 : Operating Costs ($/AF) : $149 Total Cost (2008 dollars) (1000 gpm) $277 : Total Annualized Cost ($/AF) Technical Report 6: Drinking Water Treatment for Nitrate 99

116 Proposed Full scale System Parameters - - e Figure 14 Se  Design Capacity Water Recovery: > 99% expected  2000 gpm capacity o Bed Expansion  Expandable to 4000 gpm o o 28% of settled bed height – 60% Max: 40 o  Pretreatment o Substrate and nutrient addition  Media: Sand or GAC – Acetic Acid: 10 15 mg/L as C   Cleaning requirements treatment  Post - o Biomass separator tank Aeration o bed cleaning eductor - o In  Increase dissolved oxygen Manufacturer:  o Clarifier/multimedia filter Envirogen Technologies, Inc. Chlorine disinfection o toring: N/A  Moni  Treatment system footprint  Waste Volume: o Treatment system: 2 FBRs 0.3 gpm waste per 2000 gpm treated  14’ diameter x 24’ height o Residuals handling system  DAF for solids removal o Total system footprint  For 4000 gpm 30’ 180’ x 1  For 2000 gpm   25% less than above Journal AWWA , reprinted from 2009 . (May Figure 14 FBR treatment s ystem schematic . ( Source: Webster et al. 2009) by permission. Copyright © 2009 by the American Wat er Works Association . ) 100 Technical Report 6: Drinking Water Treatment for Nitrate

117 Proposed Full scale System Cost Estimation - Capital Costs (Total with explanation or component costs) Treatment and Monitoring $1.8 million for the 2 FBR vessels, the 2 post - aeration vessels, nitrate Equipment: and perchlorate monit oring, 1 DAF, chemical feed systems, pumps, valves, additional components Filtration System: $800,000 for the 2 clarifier and multimedia filters O & M Costs (Total with explanation or component costs) Unavailable because system is not yet operating. P lease see above for estimated O & M costs from the full - scale demonstration study. Residuals Management Unlike removal technologies, the use of the FBR results in destruction of nitrate and perchlorate, rather than the transfer of these constituents to a concentrated waste stream requiring disposal. Technology Benefits and Drawbacks Benefits Drawbacks Capable of perchlorate removal and   Limited application in the U.S. handling high nitrate levels Large system footprint  Reduction of nitrate rather than just   More complicated permitting removal  Extensive pilot study necessary High water recovery and limited waste  Start  up time (up to a month) -  ing costs than removal Lower operat  Increased operator attention technologies  Sensitivity to system interruption Additional Information The full - scale system is being planned specifically to address perchlorate contamination, but will reduce nitrate as he system should be fully operational with discharge to t and well. Construction is currently underway by Mid - 2012. With approval from the California Department of Public Health (CDPH) it is expected groundwater in Early that distribution of treated drinking water will begin - 2013. Regarding operator training, for this system, the operators already held the required certification due to operation of a surface water treatment plant; existing experience, additional training would likely be needed. however, for other utilities lacking this pre - Regarding permitting, the 97 005 process from the CDPH Policy Memo 97 Guidance for the Direct Domestic 005, (“ - - Use of Extremely Impaired Sources”) was followed as guidance for the FBR installation. The 97 - 005 process is case scenarios and a 6 month demonstration of strenuous and involved, requiring analysis of failure and worst - pro FBR would be appropriate vary with the per operation of the treatment plant. The system sizes for which a n FBR can be more dependent on load than on flow concentration of contaminants. Feasible application of a n r tested for systems as small as 7 to 12 gpm and across perchlorate capacity. The FBR has been implemented o concentrations from 12 to 13000 ppb. Typical flow rates are 250 to 5000 gpm, but with high loads FBRs can become more feasible for lower flow rates. Technical Report 6: Drinking Water Treatment for Nitrate 101

118 * Source Webster, T.S., Gua rini. W.J . and Wong, H.S. (2009) Fluidized bed bioreactor treatment of perchlorate - laden gro undwater to potable standards. Journal of the American Water Works Association , 101 (5). Final Report: Demonstration of a Full Webste ized Bed Bioreactor for the Scale Fluid - r, T.S. and Togna, P. (2009) 0543. Treatment of Perchlorate at Low Concentrations in Groundwater. ESTCP Project ER - Webster, T.S. and Crowley Completed questionnaire and personal communication. , T. (2010) Unpublished sources used in the development of the case studies are not reflected in the References section of this * report. 102 Technical Report 6: Drinking Water Treatment for Nitrate

119 System Name: Western Municipal Water District (WMWD) Arlington Desalter - System Location: Riverside, CA PWSID: CA3310049 4 System Type: Pilot Study, Full CASE #1 scal e System is c urrently in the Design Phase - Treatment Type: Fixed Bed Bioreactor (FXB) Questionnaire completed by: Joseph Bernosky, WMWD, Director of Engineering Jess Brown, Carollo Engineers, Carollo Research Group Manager Startup Date: Proposed 201 3 System Description Fixed Bed Bioreactor Treatment Type 2.4 MGD (1670 gpm) System Capacity The Western Municipal Water District 89 mg/L as nitrate – 44 Raw Water Nitrate (Western) operates the Arlington Desalter as 20 mg/L as N 10 – a drinking water s upply facility. The original RO water treatment facility was constructed in the late - 1980s for salt management in the Arlington groundwater basin. In 2002 the facility was upgraded and subsequently approved as a potable water supply by the California Dep artment of Public Health (CDPH). Three wells supply raw water to the RO process and two additional wells supply bypass water for blending with the RO permeate to produce the finished water (product water). The current capacity of the Arlington Desalter is 5 million gallons per day (MGD) of RO permeate and approximately 1.3 MGD of untreated bypass for a total product water capacity of 6.3 MGD. Western’s water MP), includes supply reliability program, developed through the Integrated Regional Water Management Plan (IRW plans for expansion of the Arlington Desalter to 10 MGD capacity. A pilot study was conducted at the Arlington Desalter from April through November 2007 that demonstrated the t (biodenitrification) of the RO bypass stream. feasibility of nitrate removal using fixed bed biological treatmen After completion of the initial pilot study, CDPH proposed a turbidity standard for the biodenitrification process. e with the new Supplemental pilot testing conducted between May and September 2008 demonstrated complianc turbidity standard by adding a nitrogen degasification step followed by coagulation and polishing filters. The FXB biological process utilizes a stationary bed of granular activated carbon (GAC) on which biofilms containing nitrate - reducin g bacteria develop. Raw water is drawn from a well amended with an electron donor such as acetic acid. The water is then pumped through the GAC bed. Bacteria in the bed convert the nitrate to nitrogen gas and - time acclimation period is req reducing biological activity, which is done - uired to develop the nitrate water. A one by contacting virgin GAC with raw water and acetic acid for two to three weeks. The denitrifying bacteria used in the system are indigenous to the natural groundwater, meaning the sy stem is naturally seeded with bacteria present in the groundwater. During the pilot, a clone library analysis was performed on the bacteria within the biofilter to classify the various rse community of bacteria. At least 10 different types of denitrifying bacteria present. The analysis revealed a dive denitrifying genera were identified, including Acidovorax, which comprised approximately 37 percent of the total they would be bacteria in the FXB biofilters. The bacteria identified were gram negative, suggesting that negative bacteria tend to have thin cell walls. - particularly sensitive to chemical disinfection, as gram Technical Report 6: Drinking Water Treatment for Nitrate 103

120 Source Water Quality  Nitrate – mg/L as nitrate (mg/L N)  Co - contaminants 75 (17) – Average o o  lorate: 6 g/L Perch 44 (10) – Minimum o DBCP: 0.025 μg/L o – o Maximum 89 (20) Treatment Technology Selection The FXB process has been used successfully for removal of nitrate from drinking water supplies in Europe for using the FXB process at the decades, but has not yet been used full - scale in the United States. Reasons for Arlington Desalter include the following: • Nitrate is not concentrated in a waste stream, as in RO or IX treatment, but is converted to nitrogen gas, which is released to the atmosphere — a harmless emission because the atmosph ere is 78 percent nitrogen. • Nitrate removal efficiencies are high. Typically, greater than 90 percent removal was achieved during the detect levels was possible. Arlington pilot studies and removal to non - The biodenitrification process results in sim ultaneous destruction of some anthropogenic contaminants. • detect levels in the For example, perchlorate, found in the Arlington Desalter supply, was reduced to non - pilot study. in this country; however, it results in Historically, IX treatment has been the process of choice for nitrate removal the replacement of nitrate with chloride in the drinking water supply. It is estimated that the equivalent IX system installed at the Arlington Desalter would add approximately 3 million pounds of salt to the basin annually. The removal of salt was the purpose for construction of the Arlington Desalter in the first place. Treatment System Parameters  Design Capacity Water Recovery: expected 95%  o 1,670 gpm capacity Media: GAC   Pretreatment Treatment system footprint TBD :  o Substrate and nutrient addition  treatment - Post Degasification of o N 2 o Filtration for compliance with SWTR Residuals Management The residuals from this process, primarily biological growth, are accumulated in the filtration process. Once a termined pressure loss is experienced in filtration, the filters are backwashed to remove the accumulated prede solids. The solid laden backwash is then sent to the Santa Ana Regional Interceptor (SARI) line and ultimately disposed of off shore. Technology B enefits and Drawbacks Drawbacks Benefits  Less expensive than other technologies scale applications in operation - No full    No disposal of waste brine More complicated permitting  High water recovery rate: > 95% contaminants - Can remove co  104 Technical Report 6: Drinking Water Treatment for Nitrate

121 Treatment Technolo gy Costs Detailed cost estimates are being developed. Detailed design of this system may occur in 2012, and full - scale construction of this system may be in 2013. As an approximation, the below costs are based on a fixed bed demonstration system for the removal of - perchlorate with a similar empty bed contact 2008). time (EBCT) (Carollo Engineers Capital Costs (2008 dollars) Flow Rate: 1000 gpm 2000 gpm Total: $4,193,000 $7,395,000 Direct Installed Costs: $2,373,000 $4,200,000 O & M Costs (2008 do llars) Total Estimated Annual O&M Costs: $175,000 $348,000 Chemicals ($/yr): $161,000 $323,000 $9,000 $17,000 Other (GAC and Filter Sand) ($/yr): $5,000 $8,000 Power ($/yr): Total Costs (2008 dollars, 2.8% discount rate, 30 - year lifecycle) d Project Costs: Amortize $209,000 $368,000 Estimated Annual Budget: $384,000 $716,000 Total Treatment Costs ($/1000 $0.73 $0.68 gal): Total Treatment Costs ($/AF): $238 $222 Source * Personal communication. Bernosky, J. (2010) Brown, J. (2010) Personal commun ication. Fixed - Caroll o Engineers. Final Report: Bed Biological Nitrate Removal Pilot Testing at the Arlington Desalter Facility. Bed Biological Perchlor Carollo Engineers . (2008) Final Report: Direct Fixed - ate Destruction Demonstration. ESTCP 4. Project ER - 054 Unpublished sources used in the development of the case studies are not reflected in the References section of this * report. 105 Technical Report 6: Drinking Water Treatment for Nitrate

122 3.5 Chemical Denitrification (CD) metals have Chemical denitrification can be accomplished with reduction of nitrate by metals. Various 2+ o been investigated for use in nitrate reduction including aluminum and iron (both Fe ), while and Fe , and rhodium can be used as catalysts in nitrate reduction (Shrimali & Singh 2001). copper, palladium The advantage of chemical denitrificati on over the removal technologies is that nitrate is converted to other nitrogen species rather than simply displaced to a concentrated waste stream that requires disposal. Problems with chemical denitrification of potable water are the reduction of nitrat e beyond , and insufficient nitrate removal (nitrite can be nitrogen gas to ammonia, partial denitrification converted to nitrate with the use of chlorine in disinfection). No full - scale chemical denitrification ates for the removal of nitrate in potable water treatment. systems have been installed in the United St A significant body of research has explored the use of zero valent iron (ZVI) in denitrification. Several ® III ron), (Sulfur Modified I patented granular media options are also emerging, including SMI - ® MicroNose™ Technology, and Cleanit - LC. Based on lab and pilot - scale studies, there is much variation in the configuration of chemical denitrification systems for nitrate removal from potable water. The generic mechanism of denitrification involves the transfer of electrons from an electron donating metal to nitrate. As in biological denitrification, nitrate is reduced in accordance with Eqn. 9. However, in contrast with biological denitrification, using chemical denitrification, the nitrogen in ni trate is often reduced to the 1998 ; Hao least oxidized form, ammonium (Eqn. 9a) ( et al. 2005 ). Huang et al. - - NO NO  NO  N O  N (Eqn. 9)  2 2 3 2 - - + (Eqn. 9a) NO  NO  NH 4 3 2 sing the treatment stream through granular Nitrate is exposed to an electron donating metal by pas , and surface chemistry are important media characteristics related to media. Particle size, surface area the efficiency of nitrate removal. 3.5.1 Zero Valent Iron (ZVI) Due to the extensive research focused on th e use of zero valent iron (ZVI), ZVI will serve as a preliminary example. There is some variation in the use of ZVI. Forms of application include powdered iron, stabilized iron as nanoparticles, iron filings and permeable reactive barriers (PRBs). Rele vant reactions , are liste d in Eqns. 13 to 18 ( Huang et al. 2009). Nitrate can be 1998; Hao et al. 2005; Xiong et al. , 5 or nitrogen gas (Eqn. 1 ) by ZVI. Following nitrate reduced to nitrite (Eqn. 1 3 ), ammonia (Eqn. 1 4 ) n be reduced to ammonia (Eqn. 1 6 ). Nitrate can also be reduced by reduction to nitrite, nitrite can the the hydrogen gas that is produced from corrosion reactions (Eqn. 1 7 ) to ammonia (Eqn. 1 8 ). - o - 2+ + + NO (Eqn. 1 )  Fe + NO 3 Fe + H O + 2H 2 2 3 Technical Report 6: Drinking Water Treatment for Nitrate 106

123 - + + 2+ o  NH + NO + 4Fe 4Fe + 3H ) O + 10H (Eqn. 1 4 2 4 3 o - 2+ - + 6H ) O  N 5 + 5Fe + 12OH + 2NO 5Fe (Eqn. 1 2(g) 3 2 o - + 2+ + + 8H  + NO + NH 6 ) + 2H (Eqn. 1 O 3Fe 3Fe 2 4 2 2+ o + Fe + 2H H + Fe  (Eqn. 1 7 ) 2(g) + + - O + ) 2H 8  NH (Eqn. 1 NO + 3H + 4H 2 2 3 4 The reduction of nitrate by iron is characterized by an increa se in pH and consumption of hydrogen ions. pH is a significant controlling factor for th is treatment method (Hao et al. 2005). The kinetics of nitrate reduction by ZVI have been thoroughly covered in the literature to determine the reaction rate under rious conditions. For example, Alowitz & Scherer (2002) examined the nitrate reduction rates of three va types of iron. Findings indicate that reduction rate increases with decreasing pH. Huang et al. (1998) uction of nitrate to ammonia. Highly pH dependent, investigated the use of powdered ZVI for the red nitrate reduction was kinetically favorable only at a pH below 4. The minimum ratio of iron to nitrate 2 - /mol NO for complete reduction within 1 hour. Nitrate reduction by ZVI can be optimized was 120 m 3 through pretreatment of iron particles. High temperature exposure to hydrogen gas and deposition of copper were explored separately as options for pretreatment of the iron surface (Liou et al. 2005). Both methods resulted in improvement of nitrate reduc tion in almost neutral solutions. The mechanism of improvement is due to the surface chemistry of iron. With a buildup of a surface oxide layer, the availability of sites for nitrate reduction decreases. Hydrogen gas pretreatment reduces the oxide layer , while deposited copper serves as a catalyst for the transfer of electrons. In their investigation of stabilized ZVI nanoparticles, Xiong et al. (2009) found that the end product of denitrification (nitrogen gas versus ammonium) could be controlled by th e iron to nitrate ratio and the use of catalysts. 15 ) Figure . . 2009 Surface c hemistry of ZVI p articles . ( Source: reprinted with permission, Ch iu Examination of the surface chemistry of ZVI particles is of the utmost importanc e to model and understand its use in the reduction of nitrate. Illustrated in Figure 15 , relevant factors include corrosion of ZVI, complexation with water, surface complexation, reduction, precipitation and adsorption. In the , corrosion of ZVI, the formation of “green rusts” and “suspended green particles” is associated with Technical Report 6: Drinking Water Treatment for Nitrate 107

124 ecrease in nitrate (Choe et al. 2004). For nitrate reduction to occur, stabilization of pH and steady d contact with the reducing agent is required. Reduct ion of nitrate by ZVI or by any surface bound species requires access to surface sites. Competition for surface sites can impede nitrate reduction; Moore & Young (2005) examined chloride as a potential competitor. Results indicate a minimal impact on nit rate removal; however, other competing ions could be important regarding both competition for adsorption sites and reduction. 3.5.2 Catalytic Denitrification An extension of chemical denitrification, catalytic denitrification involves metal reduction of ni trate in the presence of a catalyst. Extensive research has investigated catalytic denitrification which may become more readily applicable to potable water treatment wit h further advances (Reddy & Lin 2000; 2002; Lemai Chen et al. 2002; Pintar et al. 2001; Gavagnin et al. 2002; Pirkanniemi & Sillanpaa gnen et al. 2007; Sun et al. 2010.) 2003; Palomares et al. 2003; Pintar 2003; Constantinou et al. 3.5.3 Chemical Denitrification - Design Considerations ign considerations in the application of chemical denitrification for nitrate summarizes key des Table 14 removal from potable water. 108 Technical Report 6: Drinking Water Treatment for Nitrate

125 Table Summary of d esign c . c hemical d enitrification. 14 onsiderations for Pretreatment  pH adjustment  Filtratio n for iron removal  pH adjustment - Treatment Post  Chlorine addition for disinfection and oxidation of iron  Gas stripping or breakpoint chlorination (for ammonia)  pH adjustment (acids and bases) Chemical Usage  Disinfection and oxidation of iron (chlorine) Constant mon itoring required to ensure efficient nitrate reduction  o Nitrate levels o Oxidation reduction potential (ORP) Monitoring of nitrite and ammonia will also be necessary due to the potential  for incomplete denitrification O&M  Management of chemicals o pH adjustment o Disinfection  Waste media and backwash water storage and disposal Ratio of electron donor to nitrate for desired removal   Hydraulic Loading Rate (HLR), Empty Bed Contact Time (EBCT) - Reactor configuration (up - flow, down  f low, in series, in parallel) System  Gas stripping or breakpoint chlorination to remove ammonia (if ammonia is Components/Design the end - product) Parameters  Monitoring equipment  Filtration to remove iron pH adjustment (decreased in pretreatment and increased before distribution)   Disinfection  Spent media disposal Waste Management  Iron sludge management and Disposal  Backwash water  No waste brine or concentrate as in removal processes No examples of full - scale application for nitrate treatment  scale trea o Unknown reliability for full - tment Limitations o Unknown costs and operational complications  Potential for incomplete denitrification Water Quality The performance of chemical denitrification systems can be affected by pH, temperature, potential - contaminants, and the availabili ty of surface sites. The reduction of nitrate by iron is interference by co characterized by an increase in pH and consumption of hydrogen ions. Thus, p H is a significant controlling factor for th is treatment method (Hao et al. 2005). Alowitz & Scherer (2002) examined the nitrate reduction rates of three types of iron. Findings indicate that reduction rate increases with decreasing pH. If nitrate in the water does not come in contact with the electron donor, then reduction will not be possible. The build - up of precipita tes can negatively impact nitrate reduction. The appropriate iron to nitrate ratio will be based on influent and target nitrate concentrations. Product water quality will require monitoring for nitrate, nitrite and ammonia. , Technical Report 6: Drinking Water Treatment for Nitrate 109

126 nsiderations System Components and Site Co Major system components include chemical storage (for pH adjustment and disinfection), the column containing the media, and post - treatment disinfection. With reduction to ammonia, post - treatment ammonia stripping may also be necessary. Desig n constraints include optimal temperature, sufficient EBCT length , avoidance of incomplete denitrification, monitoring requirements (nitrate, nitrite, and ammonia), appropriate iron to nitrate ratio for disinfection and iron and adequate chlorine dosing , . Incomplete denit xidation (DSWA 2010) rification, which can be associated with higher dissolved o oxygen (DO) levels, can result in the production of greenhouse gases (GHG) such as NO and N O. 2 or through the provision of sufficient Dissolved oxygen levels can be decreased using reducing agents e depletion of DO (Meyer et al. 2010, regarding biological denitrification). electron donor to enabl Residuals Management and Disposal In contrast to IX and the membrane technologies, the burdens of brine and concentra te disposal are minimized because nitrate is reduced through chemical denitrification . There is no concentrated brine solution requiring costly disposal. However, disposal of backwash water, spent media, and iron sludge is necessary. Maintenance, Monitor ing , and Operational Complexity With the possibility of incomplete denitrification, monitoring is required to ensure that product water does not contain high levels of ammonia or nitrite. Exposure of these nitrogen species to chlorine in disinfection or o xygen downstream can lead to nitrification (oxidation back to nitrate) in the distribution system, unless controlled. Additional O&M demands include management of chemicals erformance, and waste and chlorine), backwashing the column to maintain flow and p ( e.g., acids, bases , - scale installation for comparison, overall, chemical denitrification management. Despite having no full may potentially be less operationally complex than biological denitrification. 3.5.4 Chemical Denitrification - gies Emerging Technolo Modified Iron (SMI) Media - Sulfur Chemical reduction of nitrate has been demonstrated for potable water treatment using sulfur - mod ified 2010). Certified to the NSF/ANSI Standard 61 for use in drinking water iron granular media (DSWA ® - treatment, SMI is a patented media that is recyclable and offers the advantage of multi ple III - PS 2009). Arsenic and metals can be removed via adsorption ( hexavalent contaminant removal (SMI chromium can also be reduced and precipitated), while nitrate is reduced to ammon ia (Prima ® Envir onmental N.D.) or nitrogen gas (DSWA 2010). “The SMI III - manufacturer believes sulfur modification regulates the environment of reactions to achieve greater and a cons istent nitrate reduction” (DSWA 2010 , p. 9 ). Nitrate reduction is gover ned by the following reactions (SMI - PS 2009): 0 - + 2+ + 4Fe (Eqn. 19) + NO + 3H O + 10H  4Fe + NH 4 3 2 110 Technical Report 6: Drinking Water Treatment for Nitrate

127 - + 2+ 0 5Fe + 12H  5Fe + 2NO + N (Eqn. 20) + 6H O 2 3 2 ® - III Key advantages of SMI are the ability to remove multiple contaminants simultaneously and the limited w aste disposal costs relative to other nitrate removal options (no brine waste stream is produced) (DSWA 2010). Some previous research indicates inconsistent and insufficient nitrate removal to meet potable water regulations (DSWA 2010). s Associates ( ) Damon S. William and the City of Ripon, CA, conducted a pilot study investigating DSWA ® the use of SMI III in potable water treatment. Findings suggest that this treatment option may be - suitable for source nitrate concentrations slightly above the MCL (up t o 70 mg/L as nitrate (16 mg/L as ® flow fluidized bed across a III N)). Phase A of the pilot study was operated with the SMI media in an up - - pH range of 6.0 – 6.8 and with an EBCT of 15 to 30 minutes . P hase B tested improved media displays a process 16 me pH range and with an EBCT of 30 minutes. Figure performance across the sa ® III schematic of the SMI - process. To Waste Chlorine Static Coagulant Fluff Mixer Chlorine pH Adjust pH Adjust (Acid) (Caustic) ® Pressure SMI - III Filter Reactor Static Distribution Mixer System Backwash Wellhead Figure 16 Process schematic for denit rification using SMI - III® . ( Source: DSWA 2010 . ) . flow mode) daily to remove oxidized iron and to avoid media e media was fluffed (backwashed in up Th - ® - III agglomeration. The SMI reactor was followed by coagulation and filtration for the removal of iron, atest nitrate removal (18 mg/L as nitrate, 4 mg/L as and other constituents. In phase A, the gre arsenic , N) was insufficient to meet the 20% safety margin of the project (a goal of 36 mg/L as nitrate, 8 mg/L as N), with a starting nitrate concentration of 60 mg/L as nitrate (13.5 mg/L as N). Problems in ph ase B resulted in operation interruption; however, a maximum nitrate removal of 24 mg/L as nitrate (5.4 mg/L as N) was achieved. When the pH was reduced to 6, the system did produce water with nitrate below wo occasions. Temperature was found to have the most the MCL and the project goal nitrate level, on t Technical Report 6: Drinking Water Treatment for Nitrate 111

128 significant impact on removal efficiency. Nitrate reduction improved with increasing source water temperatures. Design constraints include temperature, EBCT, avoidance of incomplete denitrification, determination of the appropriate iron to nitrate ratio, chlorine requirements for released iron oxidation, clogging from precipitates , and pH. Nitrate reduction typically increases with decreasing pH, increasing EBCT , a nd increasing temperature (DSWA 2010 ). - Cost Considerations SMI , and media and excluded labor and Cost analysis in the City of Ripon report included acid, caustic soda waste management. In both phases, minimal variation in nitrate reduction was found with operation at pH of 6.8. Operation at the higher pH minimizes costs due to prolonging media life a pH of 6 versus a and decreasing chemical input. The production costs for operation at a pH of 6 and 6.8 were estimated ) and $0.88/1000 gal. ($ AF tively (DSWA 287/ AF ), respec to be $2.24/1000 gal. ($729/ 2010). Due to dissolution of the media over time, media disposal is not expected to be necessary. Dissolved iron is oxidized and then removed through filtration. In the Ripon pilot study , backwash water was discharged to the sewer; however, when “...direct sewer disposal is not feasible, the backwash water must undergo solid/liquid separation with the decant liquids recycled to the head of the treatment system and the ). If the waste is . 82 dewatered solids sent to an appropria te landfill for disposal” (DSWA 2010 , p deemed hazardous, disposal can be a major cost consideration. Granular Clay Media Certified to the NSF/ANSI Standard 61 for use in drinking water treatment, MicroNose™ Technology rnia , for nitrate removal in potable water media is currently being examined in Manteca, Califo 2010 treatment (MicroNose™ ). The media consists of “absorbent and permeable pottery granules which function similar to the mucous membrane in the hu man nose” (University of Hawaii 2006). Limited information is a vailable on this emerging technology. The company states, “MicroNose™ Technology media removes heavy metals, such as arsenic, manganese, and lead as well as nitrates through non - chemical processes” (MicroNose™ 2010 ). MicroNose™ offers removal of multiple and a green contaminants conc omitantly and claims to be cost - effective, suitable for nitrate removal , technology. Additional information is needed to assess the design considerations, costs and , water. applications of MicroNose™ for nitrate removal from potable Media Powdered Metal ® - As an emerging technology, Cleanit LC (from North American Höganäs) is a metal powder with the 2010). Certified to the NSF/ANSI Standard 61 for potential to achieve 60 – 90% nitrate removal (Lavis use in drinking water treatm ent, this proprietary iron - based powder could be used for removal of co - contaminants in addition to nitrate, including “arsenic, heavy metals, p hosphates and pathogens” (Lavis ® by the following: a 2010), and potentially hexavalent chromium. Cleanit LC media is characterized - 112 Technical Report 6: Drinking Water Treatment for Nitrate

129 3 – 2100 kg/m , a particle size of 150 – density of 1800 850 microns, and a porosity of 60%. The powder can be used to adsorb arsenic in less than 10 minutes, with an arsenic capacity of 4 – 8 mg/g powder. With an up flow configuration, the treatment stream is pumped through a column containing the media, - maximizing surface contact. The most significant consideration is the EBCT. At the particle surface, nitrate is reduced to nitrogen gas with EBCTs of 10 – 30 minutes. Additional key d esign factors are pH and temperature. In contrast to membrane technologies, the burdens of disposal are minimized because nitrate is reduced rather than removed. There is no concentrated brine solution requiring results indicate nitrate removal over 7 months to below the MCL costly disposal. Preliminary third party (North American Höganäs, data). However, as a new product on the market, further research is ® - required to assess Cleanit LC for the removal of nitrate and other constituents in potable water treatment. Cost Considerations 3.5.5 Chemical Denitrification - For efficient operation of a chemical denitrification system, maintaining efficient nitrate reduction is - treatment to provide ap propriate essential. Optimal performance includes managing pre and post environmental conditions, while meeting necessary potable water guidelines. Factors affecting system cost can include facility size (flow rate), source water quality (including nitrate concentration and co - emperature and pH), and target effluent nitrate concentration. contaminants), environmental factors (t Capital costs for chemical denitrification include land, housing, piping, media, storage tanks, O&M M costs equipment, preliminary testing (through pilot studies), permits, and operator training. O& - treatment, media replenishment and disposal, backwashing, chemical use ( e.g., pH include pre and post adjustment, chlorine), repair, monitoring, maintenance, power, and labor. The availability of published cost information for chemical denitrificat - scale ion is strictly limited to pilot studies, due to the lack of full - scale systems. Cost analysis in the City of Ripon report included acid, caustic soda , and media and excluded labor and waste management. In both phases, minimal variation in nitrate reduction was found with operation at a pH of 6 versus a pH of 6.8. Operation at the higher pH minimizes costs due to prolonging media life and decreasing chemical input. The production costs 1000 gal. ($729/ AF ) and $0.88/1000 gal. for operation at a pH of 6 and 6.8 were estimated to be $2.24/ ($287/ AF ), respectively (DSWA 2010). Due to dissolution of the media over time, media disposal is not expected to be necessary. Dissolved iron is oxidized and then removed through filtration. In the Ripon pilot st udy backwash water was discharged to the sewer . H owever, when “...direct sewer disposal is not feasible, the backwash water must undergo solid/liquid separation with the decant liquids recycled to the head of the treatment system and the dewatered solids se nt to an appropria te landfill for disposal” (DSWA 2010 , p. 82 ). If the waste is deemed hazardous, disposal can be a major cost consideration. A alysis 6 Treatment Cost An . detailed discussion of treatment costs is included in Section 113 Technical Report 6: Drinking Water Treatment for Nitrate

130 3.5.6 Chemical Denitrification Selected Research - lists Table A.5 of the Appendix recent research studies relevant to the use of chemical denitrification in potable water treatment. - Summary of Advantages and Disadvant ages 3.5.7 Chemical Denitrification A summary of advantages and disadvantages of chemical denitrification in comparison with other treatment options is listed in Table A.6 of the Appendix. Advantages of chemical denitrification for nitrate removal from potable water include the convers ion of nitrate to other nitrogen species (no brine or concentrate waste stream), the potential for more sustainable treatment, and the ability to remove multiple contaminants. Problems with chemical denitrification of potable water are the potential reduct ion of nitrate beyond nitrogen gas to ammonia, the possibility of partial denitrification and the associated production of , GHGs, and the lack of full - scale chemical denitrification systems. 3.6 Brine Treatment Alternatives and Hybrid Treatment Systems B rine or concentrate disposal can be a great concern with the use of the removal processes, IX, RO, and ED/EDR, especially for inland communities. Brine treatment and recycling alternatives have the potential to address disposal concerns, improving sustain ability and decreasing operating costs. Hybrid systems, combining different nitrate treatment technologies, have been explored to include the advantages of multiple treatment options, while avoiding their respect ive disadvantages ( Pintar et al. 15 Table iewski et al. lists 2001; Wisn ). 2001 ; Matos et al. 2006; Kabay et al. 2007; Van Ginkel et al. 2008 a selection of additional research related to brine treatment alternatives and the use of hybrid systems nitrate treatment performance. to improve The combination of denitrification methods with removal technologies enables resolution of common problems with each option. The brine waste stream from IX, for example can be treated using biological denitrification to r educe nitrate to nitrogen gas. Biological denitrification of IX waste brine using an up - flow sludge blanket reactor (USBR) was demonstrated at the lab - scale by van der Hoek & Klapwijk (1987). Clifford & Liu (1993) implemented a lab - scale sequencing batch reactor (SBR) for the denitrification of waste brine resulting in the ability to recycle the treated waste brine for 15 regeneration cycles. The SBR process was pilot tested i 1996). With n McFarland, CA (Liu & Clifford denitrification of spent brine fol lowed by reuse, a 95% decrease was achieved in salt waste. Use of a membrane bio - reactor in this context has also been explored (Bae et a l. 2002; Chung et al. 2007 ; Van Ginkel et al. 2008 ). Through the reduction or destruction of nitrate in spent IX brin e, disposal needs may be significantly reduced as treated brine can be repeatedly recycled back for use in resin regeneration. Additionally, through treatment of RO or ED/EDR waste concentrate, removal of nitrate from the waste stream may improve disposal options. 114 Technical Report 6: Drinking Water Treatment for Nitrate

131 Table . Selected r esearch on b rine t reatment a lternatives and 15 ybrid s ystems for n itrate t reatment of p otable h w ater. IX and Catalytic Reduction Pintar et al. (2001) RO and EDR EET Corporation (2008) IX and EDR Kabay e t al. (2007) Wisniewski et al. (2001) ED and MBR IX and Biological Van der Hoek & Klapwijk (1987) , Clifford & Liu (1993), Bae et al. (2002), Denitrification and Van Ginkel e t al. (2008) RO and VSEP Lozier et al. (N.D.) , Goltz & Parker, 2010/2011 , Yu & Kupferle (2008) Dortsiou et al. (2009), Electrochemical destruction of ni and Ionex SG Limited (2011) trate (in IX brine) 3.6.1 Electrochemical Destruction of Nitrate in Waste Brine Research by two different groups is focused on electrochemical destructio n of nitrate for the treatment of waste brine. a system which incorporates IX First, two major companies are collaborating on the development of with electrochemical destruction of nitrate. With the removal of nitrate from spent brine, the brine can ; Goltz & Parker be recycled f or reuse in regeneration (Goltz 2010 2010/2011). Spent brine is treated following IX treatment. With electricity distributed over a high surface area electrode, upon contacting the electrode surface, nitrate in the brine waste i s reduced to nitrogen gas. In the overall reaction, nitrate is reduced and water is oxidized producing nitrogen gas and oxygen. Possible reduction to ammonia/ammonium is accounted for in the process. Laboratory tests have been successful and site tests is underway to examine system performance across a variety of raw water planning of characteristics. The potential benefits of this brine treatment alternative are in the increased mize the process such that the sustainability and the decreased disposal costs. The objective is to opti savings on disposal costs, especially for inland communities, outweigh the increased electricity costs of the process. Second, a company based in the United Kingdom, Ionex SG Limited, has developed and is currently testing a patented brine treatment system utilizing an electrochemical cell for the destructi on of nitrate (Ionex SG Limited 2011). Following resin regeneration, spent brine is passed through the cell and nitrate ). 17 is converted to nitrogen and oxygen gases through i nteraction with a rhodium catalyst ( Figure Treated brine can be subsequently reused for regeneration, minimizing waste brine volume as well as espan of the cell has been laboratory the overall salt consumption of the ion exchange system. The lif tested and is estimated to operate efficiently for at least 15 years through a mechanism of automatic maintenance of the rea ctive surface ( Tucker et al. 2004 ; Ionex SG Limited 2011 ). The brine treatment system is curre ntly being pilot tested at a location in California through collaboration with UC Davis. A indicates that brine disposal costs would be significantly reduced; chemical cost benefit analysis by Ionex th existing chemical costs for typical ion and power costs of the brine treatment system balance wi exchange systems. 115 Technical Report 6: Drinking Water Treatment for Nitrate

132 Figure 17 . Schematic of the brine treatment system developed by Ion ex SG Limited . ( Source: Ionex SG Limited 2011 ) . 3.6.2 Catalytic Treatment of Waste Brine Calgon Car bon is working with the University of Illinois to develop a brine recovery system with the goal of up to 90% brine recovery for reuse using Pd - ba sed catalytic treatment (Drewry 2010). The following information was provided by Dr. Charles Werth (2010) from the University of Illinois. - As illustrated in 18 , “The spent brine solution is equilibrated with hydrogen in a gas Figure liquid membrane, and subsequently treated to remove nitrate in the packed bed catalyst system containing P d - In on granular activated carbon (GAC). The treated brine is then put back into a holding tank and reused when ion exchange br eakthrough occurs again” (Werth 2010). Figure 18 . Schematic of an ion exchange system with brine r egeneration coupled with catalytic treatment of Werth 2010 ) . Source: brine for reuse . ( 116 Technical Report 6: Drinking Water Treatment for Nitrate

133 Advantages of this brine treatment alternative include conversion of nitrate mainly to nitrogen gas, of nitrate with the Pd based avoidance of the need for brine waste disposal, and rapid reduction - catalyst. Disadvantages include the use of flammable hydrogen gas, some conversion of nitrate to ammonia, “the cost of Pd and potential fouling of the catalyst which requires regeneration using a strong oxidant li ite (Chaplin et al. 2007)” (Werth 2010). Catalytic removal of nitrate has not ke hypochlor yet been implemented at the full scale, but “catalytic systems have been used to remove chlorinated - solvents at contaminated groundwater field sites, and they appear to be econom ve ically competiti (Davie et al. 2008)” (Werth 2010). Dr. Werth stated, “Catalytic systems show great promise for removing nitrate from IX brines, but pilot - scale studies are needed to evaluate the economic viability.” 3.7 Residential Treatment (Point - of - Use, Point - of - Entry) Point of - Use (POU) and Point - of - Entry (POE) water treatment devices can be used to address high - nitrate levels and other constituents of concern at the residential scale. A POU treatment device is installed for the purpose of reducing contaminants in drinking water at a single tap, typically the kitchen tap. A POE treatment device is installed for the purpose of reducing contaminants in drinking water entering a house or building. Treatment technologies for POU and POE systems, used t o address nitrate contamination, include IX, , and distillation (Mahler et al. 2007). IX is generally considered more for POE than for POU and RO requires disposal of concentrated waste brine. RO systems require more maintenance and have lower ery, resulting in a lar ger waste volume (Mahler et al. 2007). However, according to the New water recov Hampshire Department of Environmental Servic es (2006), RO “is the most cost - effective method for producing only a few gallons o f treated water per day” (NHDES 2006 ). While distillation can require lower maintenance, energy demands are higher than the other options. Distillation systems are generally intended as POU devices as they remove all minerals and produce water that is aggressive towards plumbing materials. POU units are installed either under the counter or on the counter top, preferably by a licensed professional. The treatment units generally consist of several stages; for example, a POU RO system can a post consist of a pre - treatment filter, an RO stage, and treatment filter. The system can also include a - storage tank to hold treated water and a conductivity meter to indicate when maintenance is required. g. POE units, for the treatment of all water entering a building, are larger and require more pipin Certification to the relevant ANSI/NSF standards by an ANSI accredited third party certifier ensures the safety and performance of the residential treatment systems. In the U.S., the following certifiers have been accredited by ANSI to certify drinking water treatment systems:  Canadian Standards Association International (www.csa - international.org);  International Association of Plumbing & Mechanical Officials (www.iapmo.org);  NSF International (www.nsf.org); and Underwriters Laboratories Inc. (www.ul.com);  117 Technical Report 6: Drinking Water Treatment for Nitrate

134  Water Quality Association (www.wqa.org). Numerous RO devices for nitrate removal are certified to the ANSI/NSF standard specific to RO POU devices: NSF Standard 58 Reverse osmosis drinking water treatment systems (NSF 2009). Additionally, - the Wate r Quality Association, an accredited certifier, lists two POU ion exchange devices for nitrate – removal that are certified to NSF Standard 53: Drinking Water Treatment Units Health Effects (Water Quality Association 2011). All the technologies listed abo ve are capable of reducing nitrate levels ; however, proper maintenance of the treatment equipment is fundamental to ensure the provision of safe drinking water. Additionally, it is important to conduct periodic testing (annually or as recommended by the m anufacturer) using an accredited laboratory on both the influent water and the water produced by the treatment system to verify that it is working effectively. CDPH provides a list of approved POU devices for nitrate treatment consisting predominantly of R O devices (CDPH 20 11 ). Published cost information for POU systems is listed in Table 16 . Based on a survey in Minnesota, “the average cost of nitrate removal systems was $800 to install and $100 per year has EPA . S to maint ain” (Lewandowski 2008 , p. 92A ). Providing a more detailed cost analysis, the U . . developed a cost estimating tool for the us . S . EPA 2007 ; U . S U EPA 2011). e of POU and POE devices ( It is important to note that water systems are responsible for m eeting federal, state , and local requirements and the allowable uses of POU/POE devices vary by state. Current California regulations enable small public water systems to use POU devices to meet the nitrate MCL for up to three years, with certain restrict ions, including the following (California Code of Regulations 2011 , p. 2 ): “...a public water system may be permitted to use point of - use treatment devices (POUs) in lieu of - centralized treatment for compliance with one or more maximum contaminant levels... if ; (1) the water system serves fewer than 200 service connections, (2) the water system meets the requirements of this Article, (3) the water system has demonstrated to the Department that centralized treatment, for the contaminants of concern, is not econo mically feasible within three years of the water system’s submittal of its application for a permit amendment to use POUs, ... no longer than three years or until funding for the total cost of constructing a project for centralized treatment or access to an alternative source of water is available, whichever occurs first...” 118 Technical Report 6: Drinking Water Treatment for Nitrate

135 1 16 . Costs of POU t reatment for n itrate Table emoval . r Upfront Investment Annual Costs Comments Ion Exchange $660 s disposal of brine waste – $2425 Salt costs ($3.30 – $4.40/bag) Require Requires scale removal, higher Distillation electricity $275 – $1650 $440 – $550/yr + energy use - 4 10 gal/d Requires filter replacement, high Reverse Osmosis $1430 r water maintenance, lowe $110 $330/yr + electricity $330 – – 10 gal/d - 2 recovery Costs have been adjusted to 2010 dollars, unless indicated otherwise. 1 Mahler et al. (2007). 119 Technical Report 6: Drinking Water Treatment for Nitrate

136 4 Tulare Lake Basin and Salinas Valley Water Quality Analysis - is a key variable in both As mentioned above in the discussion of treatment technologies, water quality 17 is a summary of water quality data of high nitrate wells across treatment selection and cost. Table treatment (e.g., both the Tulare Lake Basin and the Salinas Valley. Constituents that interfere with - contaminants. sulfate) or affect treatment technology selection (e.g., TDS) are included as well as co Table 17 lists water quality data for all wells with raw water monitoring data in the CDPH w ater quality 2010 having nitrate levels above the MCL, including both abandoned and inactive database from 2006 – wells, to enable consideration of the complete range of scenarios. This water quality summary acteristics encountered in the region of interest. he highlights the wide range of water quality char T GAMA Priority Basin Projects provide additional water quality information including the basins within the study area ( 2011). USGS 17 . S ummary of water quality data of h igh n itrate w ells in the Tulare Lake Basin and Salinas Valley . Table ( CDPH PICME and WQM databases . ) Source: Units Min Max # Samples # PWS # Sources Constituent Average 64.69 45.1 402 1705 Nitrate as NO 159 209 mg/L (as NO ) 3 3 Ammonia 1 - - 2 1 1 mg/L (NH - N) 3 Sulfate 116.43 3.7 2300 1594 78 111 mg/L 1582 6.38 9.6 7.91 87 120 pH (Lab) pH (Field) 7.12 6.52 7.85 1186 3 7 o 19.71 3 24.4 780 5 10 C Temperature 1562 Hardness 344.7 6.7 5000 88 121 mg/L as CaCO 3 ug/L 117 Iron 126.34 0 5700 1586 84 Manganese 0 400 1559 86 83 116 ug/L 11. 100.04 3.5 16000 2502 78 111 mg/L Chloride 1693.31 28 21000 71 4 5 mg/L Silica 4176 TDS (Conductance) 23 43600 856.65 101 136 780.04 120 28700 2143 78 111 mg/L TDS (TDS) Alkalinity (Total) 203.85 32 340 1572 79 112 mg/L as C aCO 3 ug/L Arsenic 3.24 0 137 53 2228 99 Chromium (Total) 0 45 1785 2.9 96 132 ug/L Chromium (Hex) 1.19 0.2 3.5 20 2 2 ug/L Perchlorate 3.09 0 2000 138 2637 107 ug/L Uranium 8.4 58 532 15.64 3 5 ug/L 3.61 1.2 6.6 85 1 1 mg/L TOC NTU 108 Turbidity (Lab) 0.55 0 8.9 1494 74 120 Technical Report 6: Drinking Water Treatment for Nitrate

137 4.1 Water Quality Treatment Interference - yed nitrate treatment method, ion exchange (IX) The most commonly emplo , is highly sensitive to competing anions in the treatment stream. With the use of generic anion exchange resin for n itrate treatment, sulfate will out - compete nitrate for resin sites and, as sulfate levels in source water increase, the resin capacity for nitrate will decrease. Under such circumstances, the cost of more frequent sposal may make ion exchange less feasible. Nitrate dumping regeneration, associated chemicals and di can also be a problem, resulting in effluent nitrate levels higher than influent levels due to sulfate displacing nitrate on the resin. To address these concerns nitrate selective resin can be used, which is , more expensive, but can maintain a higher capacity for nitrate. At the highest levels in the above listed s would likely become more cost - effective. sulfate range, alternative treatment option High hardness, silica, TSS, turbidity, manganese , and iron can impact most nitrate treatment options requiring additional steps for pretreatment such as filtration and anti - scalant addition to avoid resin (IX) - treatment requirements can also be affect ed or membrane fouling (RO and EDR). Cleaning and post , thus increasing system complexity and overall costs. 4.2 Water Quality - Co - contaminants The need for multiple contaminant removal is a key factor in the selection of the most appropriate treatment option. While IX can effectively remove several co - co ntaminants, this technology cannot address all constituents of concern and alternative options should be considered with particularly poor quality waters. IX has been used for simultaneous removal of perchlorate and nitrate , and arsenic and resulting in , removal multiple contaminant nitrate; howev er, specific resins may be required to optimize occurrence of nitrate – 22 trate removal alone. Figures 19 highlight the co - higher costs than IX for ni with other constituents of interest in the Tulare Lake Basin and Salinas Valley. raw water nitrate levels above the MCL (45 mg/L as nitrate) including both active and Figure 19 maps maps high nitrate wells with high arsenic levels for which inactive well data from 2006 – 2010. Figure 20 IX may still be considered or RO may be a more suitable treatment option, depending on the for which priorities of a given system. Figure 21 maps high nitrate wells with high perchlorate levels; again IX may be considered under such circumstances. Alternatively, biological denitrification may be implemented to simultaneously remove both constituents while avoiding the brine disposal problem. Last, Figure 22 maps high nitrate we lls in which at least 1 of 4 major pesticides has been detected (bromacil, simazine, atrazine , and DBCP). The co - occurrence of nitrate and pesticides is important on two fronts. With pesticide levels above the MCL, treatment requirements will change. RO may be implemented to address the multiple contaminants or IX could be used for nitrate and activated carbon for pesticides. Regardless of the selected option, treatment for nitrate and pesticides will be more expensive than occurrence may be indicative of the source of nitrate Additionally, the co - treatment for only nitrate. contamination. 121 Technical Report 6: Drinking Water Treatment for Nitrate

138 ) . Databases Figure 19 . Raw w ater n itrate ICME and WQM evels e xceeding the MCL (45 mg/L as nitrate) . ( Source: CDPH P l Technical Report 6: Drinking Water Treatment for Nitrate 122

139 ) . Databases Figure 20 . Raw w ater h igh n ICME and WQM w ells with h igh a rsenic l evels . (Source: CDPH P itrate Technical Report 6: Drinking Water Treatment for Nitrate 123

140 ) . Databases Figure 21 . Raw w ater h igh n ICME and WQM w ells with h igh p erchlorate l evels . (Source: CDPH P itrate Technical Report 6: Drinking Water Treatment for Nitrate 124

141 . Databases ICME and WQM ) Figure 22 . Raw w ater h igh P itrate w ells with p esticides d etected . (Source: CDPH n Technical Report 6: Drinking Water Treatment for Nitrate 125

142 depth and the Through the investigation of local water quality data, the relationship between well incidence of nitrate and arsenic emerged as a potential concern for water systems drilling deeper wells to reach groundwater with lower nitrate levels. The incidence of nitrate impacted groundwater was suspected to decrease with well depth while the incidence of arsenic impacted groundwater was suspected to increase with well depth. Further analysis of this option requires that well screen depth is known in addition to well water quality. Unfortunately, for most wells with water quality information, ilable (See Technical Report 4 depth information is not ava for additional information (Boyle et al. 2012) . All wells having depth information and arsenic testing data were included , as were all on wells data) ts Inspection Compliance wells having depth information and nitrate testing data (CDPH PICME [Permi Monitoring and Enforcement] and WQM [Water Quality Monitoring] databases). The available dataset leads to a potential bias as it excludes wells with testing data for which depth information was nitrate and depth, the depth to the top of the screened interval was unavailable. For the examination of used; depth categories are based on the minimum screened depth of the well. For the examination of arsenic and depth, w screen and the ell depth was calculated as the sum of the depth to the top of the length of the screen; depth categories are based on the maximum screened depth of the well. Nitrate and arsenic levels were averaged for each well across all available tests from 2006 – 2010 to avoid bias caused by having many samples for some wells and only few samples for others. Constituent concentrations were subsequently averaged across all wells in each depth category. may be due, in part, to the No relationship was found in the Salinas Valley ( Figure 23 ); however this more limited sample size of Salinas Valley wells (192 wells for nitrate analysis, 142 wells for arsenic analysis) in comparison with the sample size of Tulare Lake Basin wells (826 wells for nitrate analysis, nic analysis). In the Tulare Lake Basin, the variation between individual wells within 741 wells for arse each depth category and across depth categories leads to inconclusive results lacking statistical 24 significance, despite the suggestion of the expected trend ( Figure ). However, in the context of the nitrate and arsenic MCLs, results in the Tulare Lake Basin suggest an increase in the incidence of arsenic MCL exceedance with well depth and a decrease in the incidence of nitrat e MCL exceedance with well depth ( Figure 25 ). Additional data are necessary for definitive confirmation of this trend and local and , well design, conditions can vary significantly as water quality varies substantially with well loc ation subsurface geology. 126 Technical Report 6: Drinking Water Treatment for Nitrate

143 - 23 . versus depth to top of screen versus total well depth (deepest water) and [NO ] Figure Salinas Valley [As] 3 . (shallowest water) [1 ft=0.30 meters] - 24 . Tulare Lake Basin [As] versus total well depth (deepest water) and [NO ] Figure versus depth to top of 3 . [1 ft=0.30 meters] screen (shallowest water) 127 Technical Report 6: Drinking Water Treatment for Nitrate

144 . epth d Figure 25 . [1 ft=0.30 Tulare Lake Basin: Incidence of n itrate and a rsenic MCL e xceedanc e with w ell meters] 4.3 Water Quality and Treatment Selection Within the drinking water community, the treatment options typically considered to address nitrate DR, BD, CD) because, contamination are IX and RO. Other technologies are available or emerging (E under some circumstances, the alternatives offer advantages that IX and RO cannot. New technologies will continue to be investigated and developed because no single option is ideal for all situations. There at a lower is not a nitrate treatment option currently available that can address all possible scenarios . The following diagram is a rough guide for treatment technology selection cost than all other options Table based on water quality concerns and possible priorities for a given water source or system ( 18 ). This diagram includes generalizations and is not intended to be definitive. In the selection of nitrate assessed by treatment technologies the unique needs of an individual water system must be professional engineers to optimize treatment selection and design. As Table 18 shows , the most appropriate method to address nitrate contamination can be influenced by influent nitrate concentrations as well as other water quality parameters. Nitrate levels well above the 28 Technical Report 6: Drinking Water Treatment for Nitrate 1

145 MCL may lead to the selection of one treatment option while nitrate levels just above the MCL may be effectively addressed with a different treatment option. lists several scenarios as an 19 - Table more cost example of appropriate options based on influent nitrate level and water system characteristics. 1 Comparison of m Table t reatment t ypes 18 . ajor . Concerns Priorities EDR BD CD RO IX IX RO EDR BD CD trate High Ni High Hardness Not Removal a Major Concern High TDS Reliability Removal Arsenic Training/ Ease of operation Removal Radium and Minimize Capital Uranium Cost Removal Chromium Minimize Ongoing Removal O&M Cost Perchlorate Minimize Footprint Removal Industry Experience Unknown Ease of Waste  Good Poor (blank) Management 1 Ion Exch ange (IX), Reverse Osmosis (RO), Electrodialysis Reversal (EDR), Biological Denitrification (BD), Chemical Denitrification (CD). This table offers a generalized comparison and is not intended to be definitive. There are notable exceptions to the above cla ssifications. 1 19 Table . Influence of n itrate c oncentration on t reatment s election . Option Considerations Practical Nitrate Range 10 – 30% above MCL Dependent on capacity and nitrate level of blending sources. Blend Dependent on regeneration efficiency, costs of disposal , and salt usage. MCL Up to 2X Brine treatment, reuse, and recycle can improve feasibility at even higher IX nitrate levels. se for Dependent on availability of waste discharge options, energy u Up to many X the MCL RO effective than IX for pumping , and number of stages. May be more cost - addressing very high nitrate levels. Dependent on the supply of electron donor and optimal conditions for - Ability to operate in a start denitrifiers. stop mode has not yet been Up to many X the MCL BD demonstrated in full - scale application; difficult to implement for single effective than IX for addressing high well systems. May be more cost - nitrate levels. 1 tants. Based on contact with vendors and environmental engineering consul 129 Technical Report 6: Drinking Water Treatment for Nitrate

146 5 Addressing Nitrate Impacted Potable Water Sources in California In the Tulare Lake Basin and Salinas Valley, current and historical methods to address nitrate and inac tivation; contamination of potable water supplies include well abandonment, destruction , , or inactivation in California blending; and treatment. The incidence of well abandonment, destruction was explored through analysis of the CDPH PICME and WQM databases. Through collaboration with CDPH, systems treating or blending to address n itrate contamination were identified and a survey was conducted to collect additional information, including the costs of treatment. , and Inactivation 5.1 Well Abandonment, Destruction If alternative source wells are available, costly treatment is often av oided through abandonment or inactivation of wells. However, wells must be properly destroyed or abandoned, in accordance with local requirements, to avoid hazardous conditions and the potential for groundwater contamination. The cost of proper well dest ruction and abandonment varies with numerous factors including depth, subsurface conditions, well type , and well construction. The minimum cost to properly destroy a 300 – er Consulting 400 ft well is ~$15,000; use of best practices would increase cost (Aegis Groundwat 20 . 2011) To assess the incidence of well abandonment and inactivation due to nitrate contamination, an analysis of the CDPH PICME and WQM databases was performed. Nitrate records from 2006 – 2010 for wells labeled as abandoned, destroyed , o r inactive were examined. Wells with at least one nitrate test above the MCL and wells including “NO3” or “Nitrate” in the well description were flagged. Table 20 lists the , resulting number of nitrate impacted we lls abandoned, destroyed or inactive in the study area and also 26 . There is evidence of mislabeling in the Figure across California; locations are mapped in PICME/WQM database. Wells missing from this analysis which may have been abandoned, destroyed , or inactivated due to nitrate may have records that are not up to date or may be mislabeled. This analysis utilizes exceedance of the nitrate MCL as an indicator of the reason for well status change; however, a po rtion of these wells may have been abandoned, destroyed , or inactivated for reasons other than nitrate contamination. The purpose of this analysis was to assess the incidence of well owever, the small number abandonment, destruction , and inactivation due to nitrate contamination . H or inactivated due to nitrate, relative to the total number of of wells identified as abandoned, destroyed , wells in these categories (which were abandoned, destroyed or inactivated for any number of reasons) , leads to two poss ible conclusions: the reason for well status change is not consistently identified in the CDPH database or there are simply very few wells in these categories. However, comparison of the frequency of abandonment, destruction , and inactivation of wells due to nitrate within the study area 20 Using best practices to properly destroy wells, “all wells would be perforated top to bottom, a high - grade cem ent - sand concrete would be used, and the concrete would be pressure grouted into the formation and then allowed to fill the well ” 2011). Use of best practices for well destruction would increase the cost above the $15,000 (Aegis Groundwater Consulting mi nimum. Technical Report 6: Drinking Water Treatment for Nitrate 130

147 with those across California indicates that the study area accounts for ~22% of such wells across the state; a disproportionate number as the total number of wells in these categories accounts for only ~13% of the CA total . Analysis of the frequency of abandoned and destroyed wells and their relevance n Technical Report 2, Section 9, including agricultural wells (Viers et al. to nitrogen loading is discussed i . 2012) 1 20 . Incidence of a bandonme nt, d estruction, and i nactivation of n itrate i mpacted drinking water w ells . Table Total Wells Nitrate Impacted Wells TLB SV Study Area Total CA Study Area Total CA 217 9 2,315 Destroyed 1 0 1 2 Abandoned 2 1 3 8 494 2,584 8,253 Inactive 33 2 35 138 1,001 39 1,712 175 Total 36 3 13,152 22% of CA total 13% of CA total 1 Source: CDPH PICME and WQM databases. There is evidence of mislabeling in the PICME/WQM database. Wells missing from this analysis which may have been abandoned, destroyed , or inactivated due to nitrate may have records that are not up to date or may be mislabeled. This analysis utilizes exceedance of the nitrate MCL as an indicator of the reason for well status change; however, a portion of these wells , may have been abandoned, destroyed or inactivated for reasons other than nitrate contamination. 131 Technical Report 6: Drinking Water Treatment for Nitrate

148 ICME and P CDPH a Figure 26 . Location of n itrate i mpacted . (Source: bandoned, d estroyed, and i nactivated w ells WQM Databases . ) Technical Report 6: Drinking Water Treatment for Nitrate 132

149 5.2 Survey of Blending and Treating Sys tems Complementing the detailed cases studies for each of the treatment types (above), a survey was scale application of nitrate treatment in California and specifically in the Tulare - conducted to assess full Lake Basin and Salinas Valley. The survey ( Figure 27 ) was designed to gather detailed information on and cost , treatment type, water quality parameters affecting treatment, details of the treatment system l and the survey packet was emailed information. Systems were contacted via phone and emai whenever possible for ease of response. The survey packet included a letter of introduction, a brief project description, and the digital survey. Whenever possible, systems were contacted via phone and email for clarif ication of submitted responses and to gather additional information. The list of water systems treating and/or blending to address nitrate contamination was developed with assistance from CDPH and analysis of the CDPH PICME and WQM databases. CDPH compile d a list of systems across CA treating and/or blending for nitrate after completing an internal review to ensure the provision of the most comprehensive list possible. Analysis of systems listed in the PICME and WQM databases confirmed treating systems ba sed on nitrate levels and descriptions of individual systems and sources. County regulated systems treating and/or blending for nitrate were subsequently determined survey focus through contact with the individual county health departments and added to the list. The was treating systems, but several blending systems were also included. 133 Technical Report 6: Drinking Water Treatment for Nitrate

150 itrate n igh h b Figure 27 . Digital s urvey d istributed to d rinking w ater s ystems t reating and/or ddress lending to a evels. l Of the 42 systems i dentified as treating for nitrate throughout CA, 26 systems completed the survey. Statewide systems are mapped in Figure 28 and systems in the study area of interest are mapped in Figure 29 . Whenever possible, systems are blending to address the nitrate problem, accounting for Figure . ~56% of the statewide systems in 28 134 Technical Report 6: Drinking Water Treatment for Nitrate

151 CDPH Internal Source: ( n Figure 28 . California d rinking w ater s ystems t reating and/or b lending for itrate . Review of facilities and contact with facilit y and county representatives.) Approximately 70% of treating systems across CA are using IX and ~20% are using RO. Several locations ha ve implemented both RO and IX, primarily to address salinity as well as nitrate. Biological treatment is being implemented at two locations in CA. After successful completion of a one - year demonstration ogical treatment system primarily to address perchlorate is installing a biol , study, a system in Rialto, CA contamination of drinking water. The water system also has a history of nitrate contamination and the 135 Technical Report 6: Drinking Water Treatment for Nitrate

152 . biological treatment system provides the potential to treat their high nitrate source(s) as well Construction of the full - scale biological treatment system is underway. In Riverside, CA , biological treatment has been investigated for the treatment of the RO bypass stream to increase total plant capacity. See S ection 3.4.5 Biological Denitrification - Case Studies , for detailed case studies of these unique systems. Focusing on the Tulare Lake Basin and the Salinas Valley ( Figure 29 ), 23 systems were found to be treating and/or ble nding to address the nitrate problem (10 blending systems, 10 IX systems , and 3 RO systems). and Salinas Valley Basin Drinking . Figure 29 . in the Tulare Lake w ater s ystems t reating or b lending for n itrate (Source: CDPH Inter nal Review of facilities and contact with facilit y and county representatives.) , and RO systems in the study area. Table 21 lists the population ranges and nitrate levels of blending, IX The IX systems cover the w idest population range; however, it is important to note that some large , systems using IX for nitrate treatment also use blending. For each system the minimum, maximum and average nitrate concentration across all active wells were determined, then the av erage of each of those categories across all systems for each of the treatment options was calculated to illustrate the typical , and average nitrate levels for each treatment type. Of these systems, the average maximum, minimum ending systems is slightly lower than that of treating systems; with lower maximum nitrate level of bl 136 Technical Report 6: Drinking Water Treatment for Nitrate

153 nitrate levels and access to a low nitrate source, the possibility of blending can avoid the need for more - Figure have only a single active well and costly treatment. Sixty percent of the IX and RO systems in 29 40% of the blending systems have only two active wells. Water quality changes and increasing nitrate - levels could be particularly problematic for these one - well systems. E ven where blending is and two successfully meeting today’s needs, it may be precarious to assume or expect that water systems can rely solely on blending for compliance into the future. Table 21 . Population and n itrate l evels of s ystems i n the s tudy a rea t reating and/or b lending for n itrate. Average Raw Nitrate (mg/L as nitrate) Avg. Population Range (Total) Min. Max. 40 15 Ion Exchange 25 – 133,750 (261,200) 71 45 – 6,585 (6,760) Reverse Osmosis 75 24 41 32 Blending 45 – 25,500 (83,4 75) 64 3 An example of a blending system in Tulare County ( Figure 30 ) illustrates the complexity of even a simple blending system. This system has seven wells, two of which are high in nitrate (wells 8 and 11 in red). Most of the year the high nitrate wells are inactive, but with high demand in summer, the system blends 22 a high nitrate source with other wells. Table lists nitrate levels, depth , and capacity of the system’s source wells. It is interesting to note that the high nitrate wells have the highest capacity and are actually some of the deeper wells. Increasing nitrate levels in the low nitrate wells would be cause for concern as the system’s blending progr am would be affected. This is one simple example of hundreds of scenarios. Even with this simple blending system, there are several complicating factors including differences in capacity, seasonal variation, and the variability of nitrate levels in wells that are very close together. Extrapolating this concept over the entire study area, the case by case nature of addressing the nitrate problem becomes more apparent. 137 Technical Report 6: Drinking Water Treatment for Nitrate

154 . 30 Wells of a b lending s Figure . ( So urce: Contact with facility/survey .) ystem in Tulare County Table 22 . Nitrate l evel, w ell d epth and w ell c apacity for a Tulare County b lending s ystem. Well # Max Nitrate Total Depth Depth to Top of Capacity Screen (ft) (mg/L as nitrate) (gpm) (ft) 163 255 153 05 10 12 328 30 174 06 07 7.8 296 94 161 250 378 08 78 393 11 398 160 150 09 81 11 400 340 318 180 170 12 11.8 470 Case studies of nitrate treatment systems are included above in the various treatment technology sections. 138 Technical Report 6: Drinking Water Treatment for Nitrate

155 alysis 6 Treatment Cost An In the estimation of treatment costs, there are two major categories to consider: c apital costs and O&M costs. Capital costs refer to the upfront investment required for the design, implementation , and installation of the treatment system. O&M cos ts refer to the annual costs for operating and maintaining the system. Based on U . S . EPA cost estimating procedures, developed through the Technology Design 2005): Panel, capital costs can be further categorized as follows ( U . S . EPA 2000 ; U . S . EPA Costs Construction Costs Process Sitework & Excavation Manufactured Equipment Subsurface Considerations Concrete and Steel Standby Power Electrical & Instrumentation Pipes & Valves Contingencies Interest During Construction Engineering Costs Indirect Costs and Profit Housing Contractor Overhead Engineering Fees Permitting and Administrative Fees Land , Legal, Fiscal Training Piloting Public Education EPA U . S . The cost analysis detailed below was performed in accordance with cost estimation procedures ( U . S . EPA 2000). Total capital costs were converted to annualized capital costs ($/kgal) based on Eqn. 21. Annualized Capital Cost = [Capital Cost ($) * Amortization Factor] / [Flow (MGD) * 1000 kgal/mgal * 365 days/year] (Eqn. 21) 802 was used which corresponds with an interest rate of 5% over 20 years An amortization value of 0.0 (Eqn. 22). N N /((1+i) Amortization Factor = (1+i) – 1)/i) (Eqn. 22) Where i = interest rate and N = number of years Annual O&M costs were calculated based on Eqn. 23 to convert to tal annual O&M costs to $/kgal. O&M Cost ($/kgal) = [O&M Cost ($)] / [Flow (MGD) * 1000 kgal/mgal * 365 days/year] (Eqn. 23) Annualized Capital Cost ($/kgal) and O&M Cost ($/kgal) were summed to determine Total Annualized Cost ($/kgal). 139 Technical Report 6: Drinking Water Treatment for Nitrate

156 U . S . EPA 19 79 Cost Estimating Manual ( U . S . EPA 1979), cost information is included for anion In the exchange for nitrate removal; however, for a more recent set of cost indices, cost information reported for anion exchange and membrane processes for arsenic treatment will b e used here for comparison U . S . EPA 2000). Similarly, in a recent AWWA publication examining the with collected cost information ( national impact of changing the nitrate MCL, cost curves of anion exchange for arsenic removal were used in the estimation of nitrate treatment costs. “USEPA’s cost curves were chosen because they are generally used for developing national compliance costs and because arsenic and nitrate use IX treatment” (Weir & Roberson 2011 comparable regenerable ). Based on the U . S . EPA cost , p. 49 curves of IX for arsenic removal, costs by system size using regenerable IX are listed in Table 23 , ranging from $0.22/kgal for a 10 MGD system to $4.60/kgal for a 0.01 MGD system. Disposal costs were not U . S . included in the cost estimates of IX for arsenic removal. EPA 1 23 . Cost e stimation using U . S . EPA c ost c urves of IX for a rsenic r emoval . Table System Capacity Annualized Capital Cost d Cost Total Annualize O&M Cost ($/kgal) ($/kgal) ($/kgal) 0.22 0.13 10 0.09 0.13 0.23 0.36 2 1 0.26 0.46 0.72 0.70 0.1 0.21 0.91 0.01 0.79 3.81 4.60 1 . U . S . EPA 2000 For the cost analysis detailed herein, treatment cost information was collected from literature, vendors, ct with existing systems. Factors affecting system cost include facility size (flow rate), surveys, and conta source water quality (including nitrate concentration), environmental factors (temperature), and target be a significant portion of O&M costs for the effluent nitrate concentration. Disposal of waste brine can for a more detailed discussion of disposal costs). removal technologies (see Section 6.4 Disposal Costs Capital costs for treatment can include land, housing, piping, storage tanks, O&M equipment, process equipment (i.e., vessels, resin, membranes, media, etc.), preliminary testing (pilot studies), permits, and or membrane replacement (due to loss or training. O&M costs for treatment can include resin, media , chemical use degradation) and disposa l ; waste residuals disposal or treatment (e.g., brine disposal) ; power ; - scalant, pH adjustment) ; repair and maintenance ; (salt, anti and labor. Costs can be difficult to assess due to inconsistencies in how cost information is reported. Comp arison of costs across different systems is not always valid due to differences in influent water quality parameters, system size, waste management options, and system configuration. Published costs do not always include comparable information. The cost information listed in this section is provided as an approximate range of costs. Costs for implementing treatment may be very different from those listed here. A thorough cost analysis of design parameters for specific locations would be required for acc urate cost estimation. The information gathered through the survey includes reported costs associated with treating systems in CA. Assumptions and sources of uncertainty in this analysis of treatment costs include the following: ost information it is unclear if all components are included in capital  For certain sources of c costs (e.g., preliminary planning, pilot testing, installation, administration fees, engineering fees, 140 Technical Report 6: Drinking Water Treatment for Nitrate

157 building cost, storage, etc.) and O&M costs (e.g., pumping, disposal, labor, ener gy, chemicals, etc.). Whenever possible efforts have been made to ensure inclusion of all relevant costs. Many treatment systems blend treated water with a bypass stream; whenever possible costs  were calculated on the basis of total produced water to acco mmodate the blending configuration. For example, an RO system may remove nitrate to very low levels, then blend the permeate with a bypass stream, raising nitrate levels to the distribution goal (and restoring other ions).  with design capacity, but some systems “over design” with a Capital costs generally increase design capacity significantly greater than the actual flow. The calculation of annualized capital costs are based on average flow rather than design capacity whenever possible to provide capital costs per unit of produced water.  O&M costs are based on actual average flow, rather than design capacity, whenever possible.  Costs were adjusted to 2010 dollars.  Costs of drilling a new well were excluded from the treatment cost analysis to ensure appro priate comparison.  Costs of systems in the design phase are anticipated costs.  Several sample costs of electrodialysis and biological treatment systems designed for the removal of other constituents were included; based on communication with water treatmen t engineers and vendors, the costs for treatment for nitrate removal should be similar.  Costs were collected for systems with wide - ranging characteristics including variation in system size, nitrate levels, co ents. contaminants, and other water quality constitu -  Several systems reported renting treatment equipment and/or contracting O&M services, resulting in very different capital and O&M costs.  Given only equipment costs (e.g., from vendors), total capital costs were modeled ba U . S . sed on scaling facto rs ( U . S . EPA 2000). EPA 6.1 Costs by Treatment Type Comparison of the average total annualized cost for IX, RO, EDR, and BD across all system sizes highlights RO as the most expensive option ( Figure 31 ). EDR costs are for a limited number of systems including costs for the treatment of constituents other than nitrate and may not be representative of actual EDR costs for nitrate removal. Based on preliminary estimates, biological treatment has the potential to be cost competitive. Costs for IX, RO , and BD are broken down into three system size categories to illustrate the variability in cost with system size ( Figure 32 , Figure 33 , and Figure 34 , respectively). Note the much higher cost for system sizes less than 0.5 MGD for IX ( Figure 32 ) and RO ). Pre ( Figure 33 liminary estimates of BD treatment costs do not illustrate the same degree of variability with system size ( ); however, BD for nitrate removal from drinking water is an emerging 34 Figure 141 Technical Report 6: Drinking Water Treatment for Nitrate

158 technology and available cost information i s limited. The very high O&M costs for small systems ( Figure 32 for IX and Figure 33 for RO) are representative of low flow systems making the cost per 1000 gallons quite high. This highlights a problem faced by many small systems lacking the benefits of economies of scale; funding may be available for the initial upfront investment, but with high O&M costs long term treatment can become unsustainable. Additionally, due t o insufficient funds for ongoing costs, small water systems can be faced with an inability to retain qualified operators which can lead to MCL (ED/EDR costs are excluded from this violations and insufficient maintenance of the treatment system. comparison due to insufficient cost information.) The variability in cost information reported here is due to many factors, including variability in water quality parameters, site considerations, and the sources of uncertainty in the cost information, as discussed a bove. For example, one very small IX system reported the disposal of waste brine to septic, a low cost option, resulting in significantly lower O&M costs in comparison with other systems. 31 echnologies. t Figure . reatment Average c ost c omparison of n itrate t 142 Technical Report 6: Drinking Water Treatment for Nitrate

159 t itrate n reatment. Figure 32 . Costs of a nion e xchange for . t reatment. Figure 33 itrate Costs of r everse o smosis for n 143 Technical Report 6: Drinking Water Treatment for Nitrate

160 reatment. t ater Figure 34 . Costs of b iological d enitrification in d rinking w 6.2 Costs by System Size As indicated above, system size is a major factor in determining nitrate treatment costs. Larger treatment systems will have higher total capital and O&M costs; however, the cost per unit of produced water generally decreases as system size increases. Large treatment systems have the advantage of and economies of scale. Based on cost information collected from vendors, literature, surveys , relative to system size are illustrated in the below cost curves for IX treatment systems, treatment costs and RO ( Figure 35 ). The development of cost curves for the other technologies was not possible due to insufficient cost information. The higher relative cos t of treatment for smaller systems can be seen the vertical axis, with decreasing system size and increasing cost as the curve sweeps moving toward upward. Again, the total cost for RO treatment is higher than that of IX. 144 Technical Report 6: Drinking Water Treatment for Nitrate

161 . emoval of IX (blue) and RO (red) for nitrate r urve Figure 35 . Cost c 24 Table includes all of the most reliable treatment cost information collected for comparison of cost ranges across system size categories for IX and RO. 145 Technical Report 6: Drinking Water Treatment for Nitrate

162 Table . Summary of a nion e xchange and r everse o smosis c ost i nformation by s ystem s ize. 24 Annualized Costs in $/1000 gallons Design Flow Range Total Combined Cost Capital Cost R System Size (people) (typical average ange (Avg.) O&M Cost Range (Avg.) Treatment Type Range (Avg.) flow range) MGD $/1000 gallons $/1000 gallons $/1000 gallons – 4.60 (1.97) – 0.62 3.81 (1.22) 0.28 Ion Exchange 0.05 – 1.53 (0.75) 0.17 Very Small 0.009 – – (0.002 500) – (25 0.052) Reverse Osm osis 0.47 4.40 (2.43) 0.22 – 16.16 (4.22) 0.69 – 19.16 (6.64) – 2.73 (1.05) – – 0.34 2.63 (0.87) 0.15 0.25 (0.15) – 0.08 Ion Exchange 1.09 – 0.17 Small 3,300) – (501 (0.052 0.39) – 1 0.19 – 1.13 (0.47) 0.23 – Reverse Osmosis 0.58 – 1.34 (0.9 3) 1.15 (0.57) – – 0.36 1.69 (0.84) 2.04 (1.06) Ion Exchange 0.06 – 0.52 (0.19) 0.12 Medium 1.09 – 3.21 – (0.39 1.3) 10,000) – (3,301 1 3.39 (2.59) – 1.35 Reverse Osmosis 0.44 – 0.63 (0.53) 0.91 – 2.76 (1.89) 0.13 1.81 (0.97) – 0.22 1.39 (0.66) – 0.41 (0.26) – 0.09 Exchange Ion Large 30.45 – 3.21 – 15.51) (1.3 100,000) – (10,001 – 0.33 – 1.46 (0.97) 0.40 Reverse Osmosis 2.21 (1.48) 0.73 – 3.67 (2.38) 1 Limited data set for the indicated system size and treatment type. Technical Report 6: Drinking Water Treatment for Nitrate 146

163 6.3 Costs by Water Quality Parameters A s highlighted above in the discussion of water quality (Section 4 Tulare Lake Basin and Salinas Valley - Water Quality Analysis ), if treatment of multiple contaminants is necessary, treatment costs will generally increase. Simila rly , the level of nitrate and water quality parameters that can interfere with treatment can increase O&M costs. 25 lists costs by system size with increasing nitrate levels and Table is intended as an example of nitrate treatment cost estimation based on nitrate concentration in source water. Table 25 is strictly an example and is not intended to be definitive, but only to suggest how The actual costs with increasing nitrate treatment costs might change with inc reasing nitrate levels. The percent increase in O&M costs was level are wide ranging and vary with numerous factors. modeled based on only two sets of vendor data in which estimates were provided based on given Available data were specifically applicable to estimation of O&M increases ni trate levels (low and high). as the nitrate concentration increases from ~1X the nitrate MCL to 2X the MCL. To extrapolate the to predict the O&M increase from 2X the MCL to exercise further, the same percent increase was used 3X the MCL. It is not possible to accurately estimate or generalize how these costs would translate for other IX systems as the two vendors provided cost estimates specifically for a system using a selective resin and a second unique system designed for low brine. Based on the information herein, O&M costs would be expected to increase even more using conventional IX under the given scenario of increasing atment system could also be designed differently for It is important to note that the tre nitrate levels. higher nitrate levels (more or larger vessels, in series/in parallel, different bypass ratios, etc.); this is not included in the table as it would be pure speculation. Technical Report 6: Drinking Water Treatment for Nitrate 147

164 1 . An exercise in the estimation of treatment costs based on appropriate technology for various nitrate levels . Table 25 2 Treatment Type O&M Cost Range (Avg.) System Size (people) Annualized Combined Cost Range (Avg.) Raw Nitrate Level $/1000 gallons $/1000 gal lons 4.60 (1.97) – 0.62 1X MCL Ion Exchange 0.28 3.81 (1.22) – 11.27 (2.88) – 2X MCL Ion Exchange 0.35 – 10.48 (2.13) 0.69 Very Small (25 500) – 0.42 3X MCL 17.15 (3.05) Ion Exchange 0.76 – 17.94 (3.80) – 3X MCL Reverse Osmosis 0.22 – 16.16 (4.22) 0.69 – 19.16 (6.64) 2.63 (0.87) Ion Exchange 0.15 – 0.34 – 2.73 (1.05) 1X MCL 7.33 (1.70) 2X MCL Ion Exchange 0.19 – 7.23 (1.52) 0.38 – Small (501 – 3,300) 3X MCL 0.23 – 11.84 (2.18) Ion Exchange 0.42 – 11.94 (2.36) 3 Reverse Osmosis 3X MCL 0.23 – 1. 15 (0.57) 0.58 – 1.34 (0.93) 2.04 (1.06) – – 0.36 1.69 (0.84) 0.12 1X MCL Ion Exchange 4.65 (1.47) – Ion Exchange 0.15 2X MCL 0.39 – 5.00 (1.60) Medium (3,301 – 10,000) 3X MCL Ion Exchange 0.18 – 7.61 (2.10) 0.42 – 7.96 (2.32) 3 Reverse Osmosis 3X MCL 0.91 – 2.76 (1.89) 1.35 – 3.39 (2.59) 1.39 (0.66) 1X MCL Ion Exchange 0.13 – 0.22 – 1.81 (0.97) 3.82 (1.16) – 0.16 – 0.25 4.24 (1.46) 2X MCL Ion Exchange 100,000) – Large (10,001 0.29 0.20 – 6.26 (1.65) – 6.68 (1.96) 3X MCL Ion Exchange Rev erse Osmosis 3X MCL – 2.21 (1.48) 0.73 – 3.67 (2.38) 0.40 1 This table is strictly an example and is not intended to be definitive, but only to suggest how treatment costs might change with increasing nitrate levels. The many factors including water quality parameters, disposal options, resin , and depends on 175% e stimated increase in O&M costs is wide ranging, 25% – , , and ion exchange system design. As nitrate levels increase, salt, disposal e capacity, resin typ and resin costs for IX will increase (O&M). Reve rse osmosis costs will increase with increasing TDS, but not at the same rate, this cannot currently be estimated. Depending on other water qu ality parameters, the costs of IX n will likely be considered as an option for > 2X the nitrate MCL. Additionally, are predicted to surpass those of RO. In the future, biological denitrificatio increasing the number and/or size of resin vessels to address higher nitrate levels would increase capital costs. O&M costs would still increase; in practice the d be designed to optimize costs. O&M increases were considered here as an example. Actual costs with increasing nitrate lev els for specific system woul systems may vary significantly from listed costs and should be assessed by professional engineers. 2 Increases in O&M are estimated from a limited dataset comprised of vendor cost estimates for IX costs with nitrate levels increasing from just above the MCL to slightly more than double the MCL. All available cost information was included in the 1X MCL scenario as a s tarting point, including systems with nitrate levels above 1X the MCL. 3 Limited dataset for the indicated system size and treatment type. 148 Technical Report 6: Drinking Water Treatment for Nitrate

165 6.4 Disposal Costs Brine disposal costs for drinking water systems in CA using IX for nitrate treatment vary with s everal factors including proximity to a coastal brine line, waste brine volume (e.g., water efficiency), and the water quality characteristics of waste brine (e.g., salinity). The presence of contaminants other than e stream can have a and vanadium) in the wast nitrate (e.g., arsenic, selenium, urani um, chromium , brine disposal options and costs ; disposal to a hazardous waste facility may be significant impact on required at a greater cost . Methods for disposal of waste brine or concentrate reported in the survey of nitrate treatment systems in CA include discharge to a septic tank and leach fields, to a wastewater treatment plant via a sewer connection, to irrigation ponds (for RO concentrate), to a brine line, and to a ing. wastewater treatment plant via truck , in part, to the great distance to the Disposal options are limited in the Central Valley of California due coast. Trucking of waste brine to coastal wastewater facilities, although costly, is sometimes the chosen trucking and disposal of spent IX brine at a coastal wastewater plant disposal option. Typical costs for from the Central Valley can be around $0.15/gallon ($150/1000 gallons of waste brine). East Bay ram for the Municipal Utility District (EBMUD), in Oakland, CA, operates a wastewater management prog disposal of high salinity and high nitrate wastewater. O&M costs for the disposal of waste brine reported in the survey of nitrate treatment systems in CA range from $0.015 to $ 0.05/1000 gallons of treated water. Assuming a high efficiency of 99.5%, O&M disposal costs range from $3 to $11/1000 gallons of waste brine. This is consistent with the results of a Meyer recent research investigation comparing the life cycle costs of several nitrate treatment options. of multiple brine disposal options including evaporation ponds, deep well et al. (2010) discuss the costs and sewer. Based on vendor estimates, results indicate total brine disposal costs (including injection , capital and O&M costs) ranging from approximately $7 to $27/1000 gallons of waste brine disposal to evaporation ponds and approximately $6 to $11/1000 gallons of waste brine disposal to sewer ( Table 26 ). (In the conversion of costs reported by Meyer et al. to the cost per 1000 gallons, an amortization However, it is value of 0.0802 was used which corresponds with an interest rate of 5% over 20 years.) important to note that the study by Meyer et al. (2010) was focused on the evaluation of nitrate aracteristics are an important factor affecting disposal costs. treatment in Arizona; location specific ch 149 Technical Report 6: Drinking Water Treatment for Nitrate

166 1 . Brine d isposal c osts . Table 26 Annualized Capital Cost O&M Cost Total Annualized Cost Total Range Average Cost by Waste Volume ($/1000 gallons) 10.23 5.62 Evaporation Ponds 15.85 7 to 27 Solar Ponds 20.48 18.80 39.27 8 to 88 Well Injection 18.52 30.52 13 to 111 12.00 Sewer 2.40 5.51 7.91 6 to 11 Average Cost by Treated Volume ($/1000 gallons) Evaporation Ponds 0.046 0.015 0.061 0.03 to 0.14 0.063 0.047 0.110 0.07 to 0.20 Solar Ponds 0.03 to 0.33 0.128 0.051 Well Injection 0.077 Sewer 0.034 0.041 0.02 to 0.12 0.007 1 B ased on Meyer et al. 2010. Costs of resin disposal can also vary with water quality parameters other than nitrate; IX resin removes not only nitrate, but other contaminants (e.g., arsenic) which can affect disposal options when resin needs to be replaced. H igh levels of other contaminants on the resin can require disposal at hazardous waste facilities and increase disposal costs , although the impact of co - contaminants is more significant filled. Using on brine disposal costs than on resin disposal costs. N on - hazardous resin can be land - 8 years , regenerable resin, requiring replacement only once every 3 – 2010c) (WA DOH 2005 and Dow the cost of land fill disposal of non - hazardous resin is expected to be minimal compared with the - disposal of other waste residuals (waste brine/concentrate). Service contracts are available with various companies to manage resin replacement and disposal. selection of the most appropriate nitrate treatment option, disposal costs are a significant factor; In the consideration of the pros and cons for the unique conditions of an individual water system is not always straightforward and can be heavily weighted by di sposal options. Although other removal technologies (RO and ED) require concentrate disposal, because IX requires the addition of salt for resin regeneration, the waste stream consists of not only the nitrate and other ions that have been removed from the water, but also the spent brine solution used in regeneration. As nitrate levels in source water increase, IX resin will need to be regenerated more frequently , increasing salt use and brine waste volume. In contrast, although the recovery rate for RO i s significantly lower than that of IX (~80% and >95%, respectively), the nitrate level that can be addressed with RO is theoretically much higher (in accordance with the membrane nitrate rejection rate). Recall that RO is used for desalination as well. A s the nitrate concentration in the treatment stream increases, with appropriate pressure, the RO membrane will continue to reject nitrate, assuming membrane scaling and fouling are properly controlled. In the comparison of IX and RO as nitrate levels incr ease, theoretically there is a tradeoff between the O&M costs for each technology. With increasing nitrate levels, chemical use and waste volume will increase for IX while power use and membrane maintenance may increase for RO. Excluding all other water quality parameters, the nitrate level at which the cost of IX exceeds the cost of RO requires more research. 150 Technical Report 6: Drinking Water Treatment for Nitrate

167 Lastly, several small water systems included in the nitrate treatment survey indicated disposal of waste concentrate to a septic system. This hi ghlights an important tradeoff; while small water systems do not have the advantages of economies of scale, with a low volume waste stream (depending on chemical composition to avoid negatively impacting isposal groundwater), d septic system function or underlying to septic can avoid other, more costly disposal options. Technical Report 6: Drinking Water Treatment for Nitrate 151

168 7 Guidance for Addressing Nitrate Impacted Drinking Water 7.1 Checklist for the Selection of Mitigation Strategy The followin g checklist is intended as a general guide for the selection of promising mitigation strategies . for nitra te impacted drinking water (adapted from U . S , WA DOH 2005a, and WA DOH EPA 2003 b 2005b). Questions regarding mitigation strategy development should be directed to the Department of 21 Public Health. concentration of distributed water and determine compliance status. Quarterly 1. Monitor nitrate monitoring may be necessary and public notification requirements must be met. 2. Develop long - term compliance schedule with the Department of Public Health. 3. Determine all pertine nt water quality characteristics (nitrate, arsenic and other co - contaminants, pH, TDS, sulfate, etc.). 4. Assess non - treatment options (e.g., removing well from service, blending, consolidation, 36 development of new sources, etc.). See Decision Tree 1 ( Figure ) and Technical Report 7 (Honeycutt et al. 2012). treatment options are not feasible, determine evaluation criteria for treatment (e.g., If non - 5. te and local requirements) effluent nitrate goal, operator certification, water demand, and sta and assess treatment options. Choose optimal approach to addressing nitrate impacted source(s). See Decision Trees 1 and 2 ( 36 and Figure 37 ). Figure - level cost estimates for capital and O&M costs. 6. Develop preliminary or planning Assess design considerations. See Table 4 and Table 6 for details on IX and RO, respectively. 7. pectively. Considerations for ED/EDR and BD are listed in Table 8 and Table 11, res Pilot test the selected solution (engineering professional required). 8. Develop construction - level cost estimates for capital and O&M costs (engineering professional 9. required). 10. Examine funding options and attain funding (e.g., Drinking water state revolving fund (DWSRF) loan). See Technical Report 8 (Canada et al. 2012). 11. Implement the selected solution. This may include the development of a pre - design report, design, obtaining appropriate permits, construction, inspections, and start - up tasks (en gineering professional required). 12. Monitor the system to ensure safe operation and the consistent supply of compliant drinking water (engineering professional may be required). 21 Additional information can be found at http://www.cdph.ca.gov/certlic/drinkingwater/Pages/default.aspx and - DrinkingWater_Jan - June%202012.pdf. http://www.rcac.org/assets/.online%20materials/CA 152 Technical Report 6: Drinking Water Treatment for Nitrate

169 7.2 Decision Trees . S EPA Figure 36 . Decision Tree 1 - . Opt ions to address nitrate impacted drinking water s ources (adapted from U 2003 b , W A DOH 2005a, and WA DOH 2005b). Technical Report 6: Drinking Water Treatment for Nitrate 153

170 xchange (adapted from USEPA 2003, WA DOH 2005a, and WA DOH 2005b). Anion e . Figure 37 - Decision Tree 2 Technical Report 6: Drinking Water Treatment for Nitrate 154

171 Sum mary and Conclusions 8 Current full - scale nitrate treatment installations in the United States consist predominantly of  . ion exchange ( IX ) and reverse osmosis ( RO ) Other technologies are available because, under While vantages that IX and RO cannot. some circumstances, the alternatives offer ad electrodialysis ( EDR ) is a feasible option for nitrate removal from potable water, the application of EDR is generally limited to high TDS and/or high silica waters. The use of biological denitrification (BD) to addr ess nitrate contamination of drinking water is more common in Europe than in the U.S. However, this option is emerging in the U.S. and two full - scale systems may become a feasible nitrate are expected with in a few years. Chemical denitrification (CD) atment option in the future; however, the lack of current full tre scale implementation suggests - the need for further research, development and testing. New technologies will continue to be investigated and developed because no single option is ideal for all situations. A single treatment solution will not fit every community; however, the provision of safe drinking water for all communities can be achieved using currently existing technology.  X for nitrate treatment of Brine reuse and treatment are vital to the continued reliance on I potable water. The low brine technologies offer a minimal waste approach and current research and development of brine treatment alternatives seem to be lighting the path toward future progress. In regions with declining water quality and insufficient water quantity, the need to address  multiple contaminants will increase in the future, suggesting the future dominance of technologies capable of multiple contaminant removal. In this context, for any individual water stem, the most appropriate technology will vary with the contaminants requiring source or sy mitigation. Although complex, analysis of the optimal treatment option for pairs and groups of contaminants will assist in the treatment design and selection. In such scenari os, the best treatment option for nitrate may not be the most viable overall.  Currently and into the future, selection of the optimal and most cost - effective potable water treatment options will depend not only on the specific water quality of a given wate r source, but also on the priorities of a given water system. If land is limited, the typical configuration required for biological treatment may not be feasible. If disposal options of brine waste are treatment or development of brine recycling costly or limited, implementation of denitrification and treatment may be the most suitable option. When deciding on nitrate treatment, the characteristics of the water system must be taken into  account as well. With consideration of economies of scale, many ru ral small water systems ford to install treatment. Even with financial assistance to cover capital costs, the long cannot af term viability of a treatment system can be undermined by O&M costs that are simply not an become more affordable through consolidation of sustainable. For such systems, treatment c multiple small water systems into larger combined water systems that can afford treatment as a 155 Technical Report 6: Drinking Water Treatment for Nitrate

172 - treatment options alone, like conglomerate. With a continued decline in water quality, non blending or drilli ng a new well, may become insufficient measures for a water system to provide an adequate supply of safe and affordable potable water. Especially in rural small communities, perhaps the most promising approach will be a combination of consolidation and tr eatment. Alternatively, separate small treatment facilities can be consolidated under a single agency. For additional discussion on the comparison of alternative water supply options and associated costs see Technical Report 7 . (Honeycutt et al. 2012) Wh ile current cost considerations are commonly the driving force in selecting nitrate treatment,  it is essential to consider the long term implications of current industry decisions. For example, it may be cost - effective for a particular system to utilize c onventional IX currently, but future water quality changes (e.g., increasing nitrate levels, co - contamination, high salt loading), discharge regulations, or disposal fees may lead to an unmanageable increase in costs. ing water treatment is being addressed with brine Environmental sustainability in drink treatment alternatives and denitrification options. It is important to approach the future of drinking water treatment with the mindset that environmental sustainability and economic sustainability are tig htly interwoven. Centralized treatment may not be feasible for widespread rural communities ; another approach  to consider is centralized management (e.g., design, purchasing, and maintenance) to minimize costs. Use (POU) and Point Entry (POE) of treatment equipment is an important option to  Point - of - - - consider, especially for the provision of safe drinking water from private wells. Unless connecting to a nearby public water system becomes an option, users relying on domestic wells have two main alternatives : drilling a new well to attain safe drinking water or installing a POU or POE device for the treatment of contaminated water. The use of POU and POE treatment equipment by small public water systems is currently only a temporary option in California and reliance on these devices for the long - term would require regulatory changes. While POU and POE treatment equipment has been shown to effectively address nitrate and other these devices to ensure the supp properly maintain ly of contaminants, it is important to consistently safe drinking water. 156 Technical Report 6: Drinking Water Treatment for Nitrate

173 9 Literature Cited Aegis Groundwater Consulting. (2011) Personal communication with Chris Johnson, Principal Hydrologist. July 6, 2011. Scherer, M.M. (2002) Kinetics of nitrate, nitrite, and Cr(VI) Alowitz, M.J. iron metal. & reduction by , Environmental Science & Technology , 299 – 306. 36 Ameridia Division of Eurodia Industrie. (N.D.) Electrodialysis for Nitrate Removal in Drinking Water , http://www.ameridia.com/html/pw.html. Applied Research Associates. (N.D.) Lar ge - Scale Demonstration of Perchlorate Removal Using Weak Base Anion Resin , http://www.ara.com/perchlorate/ pdfs / p erchlorate - t echnology.pdf ARA ( ) and Purolite . (N.D. a ) Perchlorate Ion Exchange Process using Weak Applied Research Associates , Base Anion Resin http://www.purolite.com/customized/uploads/pdfs/ p erchlorate%20 b rochure.pdf ARA ( Applied Research Associates ) and Purolite. (N.D.b) The “WIN” Process: Performance Test. ARA Applied Research Associates ) and Purolite. (N.D.c) Regenerable Weak Base Anion Ex change Process ( for Treatment of Perchlorate in Drinking Water. Arnold, K. (2004) Low - cost solution to treating nitrates in drinking water. Water Quality Products , 9 , 8 article520 - - http://www.wqpmag.com/ l ow - c ost - s olution - to - t reating - n itrates - in ater d rinking - w Aslan, Ş. (2005) Combined removal of pesticides and nitrates in drinking waters using biodenitrification Process Biochemistry , 40 , 417 – 424. and sand filter system. AWWA ( American Water Works Association ) . (1990) Water Quality and Treatment. McGraw - Hill, NY . AWWA American Water Works Association ) . (1995) Electrodialysis and Electrodialysis Reversal. Manual ( of Water Supply Practices – M38 , 1st Ed . , AWWA, Denver, C O . Desalination of Seawater AWWA ( American Water Works Association ) . (2011) Manual of Water . Supply Practices M61 , 1 st Ed . AWWA , Denver, CO. – Bae, B.U., Jung, Y.H., Han, W.W., and Shin, H.S. (2002) Improved brine recycling during nitrate removal using ion exchange. Water Research , 25 , 3330 – 3340. Banasiak, L. & Schafer, A. (2009) Removal of boron , fluoride and nitrate by electrodialysis in the presence of organic matter. , 334 , 101 – 109. Journal of Membrane Science Basin Water. (2007) Water Online - http://www.wateronline.com/article.mvc/Basin - Water - To - Provide , 0001 - Re - Groundwater - Nitrate 157 Technical Report 6: Drinking Water Treatment for Nitrate

174 N.D. ) Cleaning your RO. Hydranautics , Bates, W.T. ( http://www.membranes.com/docs/papers/02_cleaning.pdf BESST, Inc. (2008) BESST August 2008 Profiling Services G uide: Edition #1 , http://www.besstinc.com/pdf/ wsw %20 p resentation%20083008.pdf Bilidt, H. (1985) The use of reverse osmosis for removal of nitrate in drinking water. Desalination , 53 , 225 – 230. sWeek Black, J. (2003) Treasure from the deep: D rinking water. Busines . Boodoo, F. (2004) Multi , Contaminant Control with Ion Exchange. Purolite - m ulti c c ontaminant%20 http://www.purolite.com/customized/uploads/pdfs/ ontrol.pdf - Boyer, T.H. & Singer, P.C. (2006) A pilot - scale evaluation of magnetic ion exchange treatment fo r removal of natural organic material and inorganic anions. Water Research , 40 , 2865 – 2876. Boyle, D., King, A., Kourakos, M., Fogg, G.E. & Harter, T. (2012) Groundwater G., Lockhart, K., Mayzelle, Nitrate Occurrence. Technical Report 4 in: Addressing Nitra te in California’s Drinking Water with a Focus on Tulare Lake Basin and Salinas Valley Groundwater. Report for the State Water Resources Control Board Report to the Legislature . Center for Watershed Sciences, University of California, Davis. Brown, J.C. (2 008) Biological Treatments of Drinking Water. National Academy of Engineering. Frontiers of Engineering: Reports on Leading - Edge Engineering from the 2007 Symposium. National Academies Press, Washington DC. , Calgon Carbon Corporation. (2003 ) ISEP® for Ni trate Removal . Continuous Ion Exchange Separations. Calgon Carbon Corporation. (2004) ISEP® System – Calgon Carbon Corporation. (2007) Case History: City of Chino: Their Road to Nitrate Removal. Section 64417. California Code of Regulations. (2011) Title 22, Article 2.5, E., Honeycutt, K., Harter, T., Jenkins, M.W., Jessoe, K., & Lund, J.R. (2012) Regulatory and Canada, H. Funding Options for Nitrate Groundwater Contamination. Technical Report 8 in: Addressing Nitrate in California’s Drinking Water with a Fo cus on Tulare Lake Basin and Salinas Valley Groundwater. Report for the State Water Resources Control Board Report to the Legislature. Center for Watershed Sciences, University of California, Davis. Cannella, A. (2009) State of the City Address 2009. City of Ceres, CA. Mayor Anthony Cannella , http://www.ci.ceres.ca.us/newsroom/20090723.pdf. Bed Biological Perchlorate Destruction Carollo Engineers. (2008) Final Report: Direct Fixed - - Demonstration. ESTCP Project ER 0544. 158 Technical Report 6: Drinking Water Treatment for Nitrate

175 CDPH California Department of Public Health ) . (2010) PICME (Permits Inspection Compliance ( D atabases. Monitoring and Enforcement) and WQM (Water Quality Monitoring) ( California Department of Public Health CDPH . (2009) List of Certified Water Treatment Devices ) Certified for the Reduction of Nitrates. California Department of Public Health, Drinking Water Program. ( California Department of Public Health CDPH ) List of Certified Water Treatment Devices ). (2011 Certified for the Reduction of Nitrates. California Department of Public Health, Drin king Water Program. – Concentrate Management . Central Arizona Salinity Study, Phase II 2006) Prepared in partnership with ( the City of Glendale, City of Mesa, City of Phoenix, City of Scottsdale, City of Tempe, Arizona – American Water Company, City of Ch andler, City of Goodyear, City of Peoria, City of Surprise, City of Tucson, Town of Buckeye, Town of Gilbert, Queen Creek Water Company, Brown and Caldwell, and the Bureau of Reclamation http://phoenix.gov/ water /ph2concen.pdf. , Burke, J.E. (1995) Nitrate removal and water quality improvements with Cevaal, J.N., Suratt, W.B. & Desalination reverse osmosis for Brighton, Colorado. 103 , 101 – 111. , Chabani, M . , Amran, A. & Bensmaili, A. (2006) Kinetic modeling of the adsorption of nitrates by ion exchange resin. , 125 , 111 – 117. Chemical Engineering Journal based & Werth, C.J. (2007) Regeneration of sulfur - fouled bimetallic Pd - Chaplin, B.P., Shapley, J.R., catalysts. 5497. – , 5491 41 Environmental Science & Technology , Chen, Y., Zhang, Y. & conditions or maximizing catalytic reduction efficiency Chen, G. (2003) Appropriate Water Research of nitrate into nitrogen gas in groundwater. 37 , 2489 – 2495. , Chen, W.H., Haunschild, K . & Lund, J.R. (2008) Delta Drinking Water Quality and Treatment Costs: Technical Appendix H. Appe Comparing Futures for the Sacramento - San Joaquin Delta , ndix to Public Policy Institute of California. Chen, B., Nam, S., Westerhoff, P.K., Krasner, S.W., & Amy, G. (2009) Fate of effluent organic matter and DBP precursors in an effluent dominated river: A case study of wastewater impact on , 1755 downstream water quality. Water Research, 43 – 1765. Chiu , P. (2009) Simultaneous removal of viruses and DBP precursors (NOM) from drinking Presentation: of - The water using Zero Valent Iron. Ameri can Water Works Association Chesapeake Section (CSAWWA), Milford, DE. June 16, 2009. University of Delaware , Department of Civil & Environmental Engineering http://www.csawwa.org/pdf_files/csawwaseminar2009_2.pdf , Choe, S., Liljestrand, H.M. valent iron under different pH Khim, J. (2004) Nitrate reduct ion by zero - – 19 – , 335 342. regimes. Applied Geochemistry , 159 Technical Report 6: Drinking Water Treatment for Nitrate

176 Chung, J., Nerenberg, R., Rittmann, B.E. (2007) Evaluation for biological reduction of nitrate and & - based membrane biofilm reactor. Journal of perchlorate in brine water using the hydrogen 133 , 157 , 164. Environmental Engineering – City of Fresno. (2008) Report to the City Council , c ouncil d ocs/ a http://www.fresno.gov/ genda9.30.2008/1k.pdf City of Thornton. (2010) Biological Nitrate Removal Pre T reatment System for a Drinkin g Water - Water Research Foundation Final Project Update Application # 4202. . & Weber, W.J. Clifford, D. Jr. (1978) Nitrate Removal from Water Supplies by Ion Exchange. United , States Environmental Protection Agency, Office of Research and Development, Cinc innati, OH. Clifford, D. & Weber, W.J. , Jr. (1983) The determinants of divalent/monovalent selectivity in anion Reactive Polymers exchangers. , 77 – 89. , 1 Clifford, D., Lin, C., Horng, L., & Boegel, J. (1987) Nitrate Removal from Drinking Water in Glendale , Arizona. EPA - 600 - S2 - 86 - 107 , United States Environmental Protection Agency . Clifford, D. Liu, X. (1993) Biological denitrification of spent regenerant brine using a sequencing batch & reactor. , 27 , 1477 – 84. Water Research xchange with and without Brine Reuse. Presentation Clifford, D. (2007) Nitrate Ion E EPA Workshop on , Inorganic Contaminant Issues . Clifford, D. A., Sorg, T. & Ghurye, G. (2010) Ion Exchange and Adsorption of Inorganic Contaminants , Water Quality and Treatment , 6th Ed . , pp. 12.1 - 12.97, Edz wald, J.K. ( ed. ). AWWA and McGraw - Hill, . NY Conlon, W.J., Blandon, F.A. & Moody, J. (1995) Cost comparison of treatment alternatives for the removal of nitrates and DBCP from Southern California groundwater. Desalination , 103 , 89 – 100. ta, C., Constantinou, C., Cos Efstathiou, A. (2007) The remarkable effect of oxygen on the N & selectivity 2 of water catalytic denitrification by hydrogen. & Technology , 41 , 950 Environmental Science – 956. Contaminant Removal News. (2007) Regulator lifts eight - year old administ rative order against city, eight months after installation of ion exchange treatment plant. Contaminant Removal News , 1 , http://www.purolite.com/ elId/606350/ is vars/default/customized/uploads/pdfs/ m c c ook%20pri r nt.pdf Davie, M.G., Cheng, H., Hopkins, G.D., L eBron, C.A., & Reinhard, M. (2008) Implementing heterogeneous catalytic dechlorination technology for remediating TCE - contaminated groundwater. 8915. – 42 , 8908 Environmental Science & Technology , 160 Technical Report 6: Drinking Water Treatment for Nitrate

177 Della Rocca, C., Belgiorno, V. & Meric, S. (2006) An heterotr ophic/autotrophic denitrification (HAD) Process Biochemistry approach for nitrate removal from drinking water. 41 , 1022 – 1028. , Dördelmann, O., Panglisch, S., Klegraf, F., Hell, F., Moshiri, A., Emami, A. & Höll, W. (N.D.) Nitrate R r in Iran – An Assessment of Three Different Treatment Processes emoval from Drinking Wate Based on Pilot - Scale Investigations. Dördelmann, O., Buchta, P. Panglisch, S., Klegraf, F., Moshiri, A., and Emami, A. (2006) Heterotrophic Denitrification in Drinking Water Treatment – Resul ts from Pilot Plant Experiments in 442 Progress in Slow Sand and Alternative Biofiltration Processes Mashhad/Iran , - , pp.433 .). G , R. ( e d imbel IWA Publishing, London, UK Dördelmann, O. (2009) Full - scale Biological Denitrification plants in Germany, Austria and Poland. Seattle, WA AWWA Water Quality Technology Conference (WQTC). . Dortsiou, M , Katsounaros, I., Polatides, C., & Kyriacou, G. (2009) Electrochemical removal of nitrate . from the spent regenerant solution of the ion exchange. , 248 , 923 – 930. Desalination ) How to Design an Ion Exchange System a , Dow Chemical Company. (2010 http://www.dow.com/liquidseps/design/ixdesign_regen.htm, http://www.dow.com/liquidseps/design/ixdesign_efficiencies.htm Dow Chemical Company. (2010b) CADIX Software , http://www.dow .com/liquidseps/design/cadix.htm , ow Ion Exchange Resins Dow Chemical Company. (2010c) D Removal of Nitrate http://dow - – answer.custhelp.com Dow Chemical Company. (2010d) DOWEX Ion Exchange Resins , http://www.inaqua.de/ rod/ion/pdf_en/204_ d owex_en.pdf p Dow Water and Process Solutions. See Goltz, Robert. Drewes, J.E., Bellona, C.L., Xu, P., Amy, G.L., Filteau, G. & Oelker, G. (2008) Comparing Nanofiltration and Reverse Osmosis for Treating Recycled Water. Water Research Foundation Final Report, # 91212. ry, C. (2010) Drew Personal communication with Charles Drewry, Calgon Carbon Representative. DSWA ) and City of Ripon, California. (2010) City of Ripon, California Damon S. Williams Associates ( – Integrated Nitrate and Arsenic Treatment De monstration. ID No. P50 - Proposition 50 Project 3910007 055. - Dvorak, B.I. & Skipton, S.O. (2008) Drinking Water Treatment: Reverse Osmosis. NebGuide , University of Nebraska, Lincoln , http://www.ianrpubs.unl.edu/epublic/live/g1490/build/g1490.pdf Hung, A., Rosenstock, Dzurella, K.N., Medellin - Azuara, J., Jensen, V.B., King, A.M., De La Mora, N., Fryjoff - K., Lund, J.R. & Pettygrove, G.S. T.S., Harter, T., Howitt, R., Hollander, A.D., Darby, J., Jessoe, 161 Technical Report 6: Drinking Water Treatment for Nitrate

178 (2012) Nitrogen Source Reduction to Protect Groundwater Quality. Technical Rep ort 3 in: Addressing Nitrate in California’s Drinking Water with a Focus on Tulare Lake Basin and Salinas Valley Groundwater. Report for the State Water Resources Control Board Report to the Center for Watershed Sciences, University of Califo rnia, Davis. Legislature. & Edwards, M., Jacobs, S. Dodrill, D.M. (1999) Desktop guidance for mitigating Pb and Cu corrosion by - products. Journal of the American Water Works Association , 91 , 66 – 77. Edwards, M. & Triantafyllidou, S. (2007) Chloride - to - sulfate mass rati o and lead leaching to water. Journal of the American Water Works Association , , 96 – 109. 99 , EET Corporation. (2008) How Does HEEPM Work? http://www.eetcorp.com/heepm/heepmworkcont.htm Elyanow, D. & Persechino, J. (2005) Advances in Nitrate Removal. GE – Ge neral Electric Company, Water & Process Technologies , http://www.gewater.com/pdf/technical%20papers_cust/ a mericas/ e nglish/ tp 1033 en .pdf Engle, D. (2007) The HERO Treatment. Onsite Water Treatment: The Journal for Decentralized Wastewater Treatment Systems http://www.foresterpress.com/ow_0711_ultra.html , Envirogen Technologies, Inc. ( 2002 Case Study: Hi - Desert Water District, California. ) 2004 High Efficiency Ion Exchange for Arsenic and Nitrate. Envirogen Technologies, Inc. ( ) Envirogen Technologies, Inc. ( 007 ) Case Study: City of Pomona, California. 2 Envirogen Technologies, Inc. ( 2009 ) Image provided via personal communication February, 2012 . Envirogen Technologies, Inc. (2010) Envirogen Technologies Starts Up Uranium and Nitrate Removal System for East Vall ey Water District of San Bernardino County. Ergas, S.J. & Rheinheimer, D.E. (2004) Drinking water denitrification using membrane bioreactor. Water Research, 38 , 3225 – 3232. Excel Water Technologies, Inc. (2007) Reverse Osmosis Membrane Applications , ://www.excelwater.com/eng/b2c/our_product_membrane.php http & Gabelich, C.J., Gerringer, F.W., Franklin, J.C., Gao, J., Cohen, Y. Suffet, I.H. (2004) Reverse Osmosis Pretreatment: Challenges with Conventional Treatment . Proceedings of the 2004 AWWA ACE Meeting, Orlando, F L . Gavagnin, R., Biasetto, L., Pinna, F. & Strukul, G. (2002) Nitrate removal in drinking waters: the effect of tin oxides in the catalytic hydrogenation of nitrate by Pd/SnO catalysts. Applied Catalysis B: 2 38 99. – , 91 Environmental , 162 Technical Report 6: Drinking Water Treatment for Nitrate

179 GE General Electric Company ) . (N.D.) GE and Public Water Supplies: Representative Installations (RO, ( NF, EDR). GE Water & Process Technologies. ( General Electric Company ) . (2008) GE 2020 EDR Systems: Electrodialysis Reversal Technology. GE GE Water & Process Tech nologies. ( General Electric Company ) . (2010) Electrodialysis Reversal (EDR ), GE http://ge.ecomagination.com/products/electro - - reversal.html dialysis ( General Electric Company GE . (2010b) HERO: High - efficiency RO process boosts recovery ratios, cuts ) cleanin g frequency. GE Power & Water, Water & Process Technologies , http://www.gewater.com/products/equipment/spiral_membrane/ hero .jsp Ghafari, S., Hasan, M. & Aroua, M.K. (2008) Bio - electrochemical removal of nitrate from water and 99 – A review. wastewater urce Technology , , 3965 – 3974. Bioreso Goltz, R. & Parker, G.R. (2010) Abstract submittal for 2011 AWWA Annual Conference and Exposition . An Integrated Selective Ion Exchange Electrochemical Destruct Process for Nitrate in Ground water. Goltz, R. (2010) Personal Communication with Robert Goltz , Dow Water and Process Solutions . Gormly, S. & Borg, C. (N.D.) Presentation: Introducing an Advanced Technology for Sustainable Nitrate Removal for Ground Water Treatment , http://www.c3p.org/Workshop%202007%20documents/ w ater%20 ab.pdf h s he rwin%20 g ormly_ apt %20 s ustainable%20 Grischek, T., Schoenheinz, D., Syhre, C. & Saupe, K. (2010) Impact of decreasing water demand on bank filtration in Saxony, Germany. , 3 , 11 – Drinking Water Engineering and Science 20. Guter, G.A. (1982) Rem oval of Nitrate from Contaminated Water Supplies for Public Use: Final Report. EPA - 600/2 - 82 - 042 , United States Environmental Protection Agency, Cincinnati, OH. ater Guter, G.A. (1983) United States Patent No. 4479877 Removal of N itrate from W . S upplies U sin g a nion T ributyl A mine S trong B ase A xchange E R esin , http://www.freepatentsonline.com/4479877.html e, Ion Exchange Guter, G.A. (1995) Nitrate Removal from Contaminated Groundwater by Anion Exchang Sengupta, A.K. ( Technology: Advances in Pollution Control. ). Technomic Publishing, ed. Lancaster, PA. Hao, Z. Xu, X. & Wang, D. (2005) Reductive denitrification of nitrate by scrap iron filings. Journal of Zhejiang University , 6B , 182 – 186. Science Haugen, K.S., Semmens, M.J. & Novak, P.J. (2002) A novel in situ t echnology for the treatment of nitrate , 3497 36 3506. – contaminated groundwater. Water Research, Technical Report 6: Drinking Water Treatment for Nitrate 163

180 & Baumgartner, G. (1998) Experience with full - scale electrodialysis Hell, F., Lahnsteiner, J., Frischherz, H. for nitrate and hardness removal. Desalination , 11 7 , 173 – 180. Helfferich, F.G. (1995) . Dover Publications, Inc., NY. Ion Exchange Hilkert, E.J., Young, T.M., Green, P.G. & Darby, J.L. (2010) Costs of arsenic treatment for potable water in California and comparison to USEPA affordability metrics. Journal of American Water Resources 1254. Association , 46 , 1238 – Holl, W. (1995). CARIX Process – A Novel Approach to Desalination by Ion Exchange , Ion Exchange Technology: Advances in Pollution Control. Sengupta, A.K. ( ed. ). Technomic Publishing, Lancaster, PA. H olló, J. & Czakó, L. (1987) Nitrate removal from drinking water in a fluidized - bed biological denitrification bioreactor. Acta Biotechnology , 7 , 417 – 423. Honeycutt, K., Canada, H.E., Jenkins, M.W. & Lund, J.R. (2012) Alternative Water Supply Options for Ni Addressing Nitrate in California’s Drinking Water trate Contamination. Technical Report 7 in: with A Focus on Tulare Lake Basin and Salinas Valley Groundwater. Report for the State Water ed Sciences, University of . Center for Watersh Resources Control Board Report to the Legislature California, Davis. Howe, K.J. (2004) Technical Challenges to Concentrate Disposal from Inland Desalination. The University of New Mexico, Department of Civil Engineering. Proceedings Identifying Technologies to Improve Regional Water Stewardship: A Conference Series Featuring Intersections of Technology and Water Management North Middle Rio Grande Corridor . – Huang, C., Wang, H. & Chiu, P . (1998) Nitrate reduction by metallic iron. Water Research , 32 , 2257 – 2264. Hunter, W.J. (2 001) Use of vegetable oil in a pilot - scale denitrifying barrier. Journal of Contaminant Hydrology, – 53 , 119 131. Humbert, H., Gallard, H., Suty, H., and Croue, J. (2005) Performance of selected anion exchange resins for the treatment of high DOC content su Water Research , 39 , 1699 rface water. 1708. – Hydranautics. (2001) Commercial RO Technology , http://www.es.flinders.edu.au/~mattom/Utilities/salcon.html I onex SG Limited. (2011) Personal Communication. Johnston, R. & Heijnen, H. (N.D.) Safe Water Technology for Arsenic Removal. Technologies for Arsenic Removal from Drinking Water. Jones, C.S., Hill, D. & Brand, G. (2007) Use a Multifaceted Approach to Manage High Source - Water 22. – , 20 Nitrate. American Water Works Association , Opflow 164 Technical Report 6: Drinking Water Treatment for Nitrate

181 matya, S., Arar, O. Yuksel, U. (2007) Removal of nitrate from ground water by Kabay, N., Yuksel, M., Sa & a hybrid process combining electrodialysis and ion exchange processes. Separation Science and Technology 42 , 2615 – 2627. , Kapoor, A. & Viraraghavan, T. (1997) Nitrate removal from drinking water – R eview. Journal of Environmental Engineering , 123 , 371 – 380. Keller, M. (2000). Removing nitrate by ion exchange. Water Technology , 23 . Kemper, J.M., Westerhoff, P., Dotson, A. Mitch, W.A. (2009) Nitrosamine, dimethylnitramine, and & chloropicrin formation during strong base anion - exchange treatment. Environmental Science & 32 , 466 – 472. Technology, Khan, I.A. & Spalding, R.F. (2004) Enhanced in situ denitrification for a municipal well. Water Research , 38 , 3382 – 3388. Killingstad, M. W., Widdowson, M.A., & Smith, R.L. (2002) Modeling enhanced in situ denitrification in groundwater. Journal of Environmental Engineering , 128 , 491 – 504. Kim, J. & Benjamin, M.M. (2004) Modeling a novel ion exchange process for arsenic and nitrate removal. W 38 , 2053 – 2062. ater Research, King, A., Jensen, V., Fogg, G.E. & Harter, T. (2012) Groundwater Remediation and Management for California’s Drinking Water with a Focus on Nitrate. Technical Report 5 in Addressing Nitrate in: lley Groundwater. Report for the State Water Resources Control Tulare Lake Basin and Salinas Va Board Report to the Legislature . Center for Watershed Sciences, University of California, Davis. & Kneifel, K., Luhrs, G., Wagner, H. (1988) Nitrate removal by electrodialysis for brewing water . Desalination 68 , 203 – 209. , Koch Membrane Systems, Inc. (2008) About Reverse Osmosis – Filtration Overview, A KMS Leadership Category , http://www.kochmembrane.com/sep_ro.html by electrodialysis. & Koparal, A.S. Ogutveren, U.B. (2002) Removal of nitrate from aqueous solutions International Journal of Environmental Studies 59 , 323 – 329. , Kurt M., Dunn, I.J., & Bourne, J.R. (1987) Biological denitrification of drinking water using autotrophic - organisms with H – in a fluidized Biotechnology bed biofilm reactor. and Bioengineering , 29 , 493 2 501. Lavis, T. (2010) Cleanit® - LC : Water Treatment Media. Höganäs. ( Lawrence Livermore National Laboratory Nitrate Working Group. (2002) Nitrate Contamination in LLNL ) sment and Resource Protection: California Groundwater: An Integrated Approach to Basin Asses Draft. 165 Technical Report 6: Drinking Water Treatment for Nitrate

182 Lemaignen, L., Tong, C., Begon, V., Burch, R., Chadwick, D. (2002) Catalytic denitrification of water & 48. - , 75 , 43 – Catalysis Today with palladium based catalysts supported on activated carbons. nt and Purification Holding B.V. (2009) History of Water Treatment , Lenntech Water Treatme - water - http://www.lenntech.com/history treatment.htm Lenntech Water Treatment and Purification Holding B.V. (2009b) Ion Exchange Calculator , - http://www.lenntech.com/calculators/exchange/ion exchange.htm - Lenntech Water Treatment and Purification Holding B.V. (2009c) Scaling and Anti calants , S http://www.lenntech.com/antiscalants.htm Lenntech Water Treatment and Purification Holding B.V. (2009d) Reverse Osmosis , http://www.lenntech.com/library /reverse osmosis/whatisro.htm - – Lenntech Water Treatment and Purification Holding B.V. (2009e) SDI , Silt Density Index http://www.lenntech.com/sdi.htm Lenntech Water Treatment and Purification Holding B.V. (2009f) Reverse Osmosis Post - treatment: Neutraliza tion/Remineralization , http://www.lenntech.com/ro/bwro - remineralization.htm Lewandowski, A.M. (2008) Costs of groundwater nitrate contamination. Journal Tech Transfer Briefing. of Soil and Water Conservation , 63 , 92A. L iou, Y.H., Lo, S.L., Lin, C.J., Hu , C.Y., Kuan, W.H. & Weng, S.C. (2005) Methods for accelerating nitrate - reducing pretreatment and - neutral pH: Effects of H reduction using zerovalent iron at near 2 copper deposition. & Technology , 39 , 9643 – Environmental Science 9648. Liu, X. & Clifford, D. A. (1996) Ion exchange with denitrified brine reuse. Journal of the American Water Works Association , 88 , 88 – 99. Lozier, J.C., Erdal, U.G., Lynch, A.F., & Schindler, S. (N.D.) Evaluating Traditional and Innovative Concentrate Treatment and Disposal Method s for Water Recycling at Big Bear Valley, California. CH2M HILL and Big Bear Area Regional Wastewater Agency (BBARWA ), http://www.vsep.com/pdf/articles/Big Bear - - RO - Reject - Report.pdf Mahler, R.L., Colter, A., Hirnyck, R. (2007) Quality Water for Idaho – N itrate and Groundwater. & University of Idaho, College of Agricultural and Life Sciences. Mansell, B.O. & Schroeder, E.D. (2002) Hydrogenotrophic denitrification in a microporous membrane 4690. bioreactor. 36 , 4683 – Water Research, Mateju, V., Cizinska, S., K rejci, J. and Janoch, T. (1992) Biological water denitrification – A review. 14 183. – , 170 Enzyme and Microbial Technology , Technical Report 6: Drinking Water Treatment for Nitrate 166

183 Matos, C., Velizarov, S., Crespo, J. and Reis, M. (2006) Simultaneous removal of perchlorate and nitrate he ion exchange membrane bioreactor concept. Water Research , 40 from drinking water using t , 231 – 240. McAdam, E.J. & Judd, S.J. (2006) A review of membrane bioreactor potential for nitrate removal from 148. drinking water. , 196 , 135 – Desalination McAdam, E.J. & Judd, S.J. (2008) Imme rsed membrane bioreactors for nitrate removal from drinking – water: Cost and feasibility. Desalination , 231 , 52 60. Meyer, K. (2009) Full scale Biological Denitrification Installations. City of Glendale. - E., Tang, Y., & Scott, R., CH2M H ILL . (2010) Meyer, K.J., Swaim, P.D., Bellamy, W.D., Rittmann, B. Biological and Ion Exchange Nitrate Removal: Performance and Sustainability Evaluation. Water Research Foundation. MicroNose Technology, Inc. (2010) http://www.micronose.com/mnose - b1.30.html Midaoui, A.E., Elhan nouni, F., Taky, M., Chay, L., Sahli, M.A.M., Echihabi, L., & Hafsi, M. (2002) Optimization of nitrate removal operation from groundwater by electrodialysis. Separation and Purification Technology , 29 , 235 – 244. – rinking Water Protection Series: Nitrate Contamination Minnesota Department of Agriculture. (N.D.) D What is the Cost ?, http://www.mda.state.mn.us/protecting/waterprotection/~/media/files/ waterprotection/dwps2.ashx protecting/ & Young, T.M. (2005) Chloride interactions with iron surface s: Implications for perchlorate Moore, A.M. and nitrate remediation using permeable reactive barriers. Journal of Environmental Engineering , 131 , 924 – 933. Nataraj, S.K., Hosamani, K.M. & Aminabhavi, T.M. (2006) Electrodialytic removal of nitrates and Journal of Applied Polymer hardness from sim ulated mixtures using ion - exchange membranes. 99 , 1788 Science , 1794. – National Academy of Engineering. (2008) - Edge Engineering Frontiers of Engineering: Reports on Leading from the 2007 Symposium. National Academies Press, Wash ington , DC. Rittmann, B.E. (2004) Hydrogen Nerenberg, R. & - based, hollow - fiber membrane biofilm reactor for reduction of perchlorate and other oxidized contaminants. Water Science and Technology , 49 , 223 – 230. Small Water – Newkirk, D. (2010) Guidance Manual for Evaluati on of SRF Capital and O&M Costs Systems. Prepared for the California Department of Public Health. Technical Report 6: Drinking Water Treatment for Nitrate 167

184 NHDES New Hampshire Department of Environmental Services ) . (2006) Environmental Fact Sheet: ( , http://des.nh.go v/organization/commissioner/pip/factsheets/ Nitrate/Nitrite in Drinking Water 3 9.pdf dwgb/documents/dwgb - - SED Nitrate Removal Technology , http://www.nitron.co.il/ Nitron, Ltd. (2009) Nitron – echnology.php#ben t Nitron, Ltd. (2009b) The Use of Selective Electro - from Ground and Surface Dialysis for Nitrate Removal Water. White Paper, September 2009. - effective Nitron, Ltd. (2010) Presentation: Nitrate removal plants for ground and surface water. A cost - and environmentally friendly solution: Nitrate Removal with SED. Nolan, B.T., Hitt, K.J., & Ruddy, B.C. (2002) Probability of nitrate contamination of recently recharged ground waters in the conterminous United States. Environmental Science and Technology , 36 , 2138 – 2145. - LC North American Höganäs. (N.D.) Cleanit® 100% Green Solution for Nitrat e Removal from : Groundwater . Cleanit® - LC Brochure . North American Höganäs, Environment/Water Division. for Nitrate Removal from North American Höganäs (N.D.) Cleanit® - LC : 100% Green Solution . Groundwater. Cleanit® - LC Data . North American Höganäs, Environm ent/Water Division. (American National Standards Institute) O smosis . (2009) Standard 58: Reverse NSF International/ ANSI D rinking ater T reatment S ystems. W ® 2008 ) Application Bulletin: Nitrate Removal with MIEX Treatment. Orica Watercare. ( ® 2008 b) Technical Note: Counter Current MIEX Orica Watercare. ( Treatment System for High Capacity Ion Exchange Applications. Pacific Institute. (2011) The Human C osts of Nitrate - C ontaminated Drinking Water in the San Joaquin Valley. Pacific Institute, Oakland, CA. Palomar es, A.E., Prato, J.G., Marquez, F., & Corma, A. (2003) Denitrification of natural water on 41 supported Pd/Cu catalysts. Applied Catalysis B: Environmental , 13. , 3 – Panglisch, S., Dördelmann, O., Buchta, P., Klegraf, F., Moshiri, A., Emami, A., Fakhraei, M .R. Höll, W. & German Joint Co – A (2005) Nitrate Elimination from Raw waters - - operation Project , n Iranian Proceedings BMBF & UNESCO - RCUWM Workshop in Berlin . “Innovations in Water and Wastewater Technology.” Pintar, A., Batista, J. & Levec, J. (2001) Int egrated ion exchange/catalytic process for efficient removal of nitrates from drinking water. Chemical Engineering Science , 56 , 1551 – 1559. Pintar, A., Batista, J. & Levec, J. (2001) Catalytic denitrification: direct and indirect removal of nitrates 510. – , 503 66 from po table water. Catalysis Today , 168 Technical Report 6: Drinking Water Treatment for Nitrate

185 Pintar, A. (2003) Catalytic processes for the purification of drinking water and industrial effluents. , 77 – 465. Catalysis Today , 451 Sillanpaa, M. (2002) Heterogeneous water phase catalysis as an environmental Pirkanniemi, K. & A application: Chemosphere , 48 , 1047 – 1060. review. Prato, T. & Parent, R.G. (1993) Nitrate and Nitrite Removal from Drinking Water Supplies with Electrodialysis Reversal. GE Power & Water, Water & Process Technologies , http://www.gewater.c om/pdf/ t echnical%20 p apers_ c ust/ a mericas/ e nglish/ tp 1072 en .pdf Prima Environmental. (N.D.) Innovative and Emerging Remediation Technologies , http://www.primaenvironmental.com/emerging.html AWWA Ann ual Conference Proceedings , Rash, V.D. (1992) Nitrate Removal at Des Moines Water Works. pp. 155 Vancouver, BC, 178. - Reahl, E.R. (2006) Half a Century of Desalination with Electrodialysis. GE – General Electric Company, Water & Process Technologies , http://www.gewater.com/pdf/technical%20Papers_cust/americas/ e h/ tp 1038 en .pdf nglis Reddy, K. & Lin, J. (2000) Nitrate removal from groundwater using catalytic reduction. Water Research , 34 , 995 – 1001. http://www.remco.com , Remco Engineering. (N.D.) Water Systems and Controls /ro_quest.htm Rittmann, B.E. (2007) The membrane biofilm reactor is a versatile platform for water and wastewater treatment. , 12 , 157 – Environmental Engineering Research 175. Rohm and Haas Company. (2007) Municipal Potable Water Systems: Low waste Nitrate removal for Potable Water Systems. Rudolph, D. (2010) Presentation: Quantifying P erformance of R egional S cale R eductions in N utrient adose A S ource W ater P rotection through V pplications for Z one M onitoring. Toward Sustainable Groundwater in Agriculture Conf erence , Burlingame, CA. University of Waterloo, Department of Earth and Environmental Sciences. Sahli, M.A.M., Annouar, S., Mountadar, M . & Elmidaoui, A. (2008) Nitrate removal of , Soufiane, A. rodialysis. brackish underground water by chemical adsorption and by elect , 227 , Desalination 327 – 333. Samatya, S., Kabay, N., Yüksel, Ü., Arda, M. & Yüksel, M. (2006) Removal of nitrate from aqueous solution by nitrate selective ion exchange resins. , 66 , 1206 – Reactive & Functional Polymers 1214. Sargent (2006) Cost effective arsenic reductions in private well water Michaud, J., Boyle, K. & Smith , A. - 1245. – 1237 42, in Maine. Journal of the American Water Resources Association, 169 Technical Report 6: Drinking Water Treatment for Nitrate

186 Schnobrich, M.R., Chaplin, B.P., Semmens, M.J. Novak, P.J. (2007) Stimulating hydrogenotrop hic & denitrification in simulated groundwater containing high dissolved oxygen and nitrate concentrations. , 41 , 1869 – 1876. Water Research N itrate Scott, J. (2010) Remedies for ontamination A nything but Q uick, C heap. California Watch , C http://californiawatch .org/remedies - nitrate - contamination - anything - quick - cheap Seidel, C., Summers, R.S., Landi, N., Zachman, B. & Frey, M. (2004) The MIEX® Process: A Technology for , Proceedings of the American Water Works Association Removing Inorganic Contaminants Annual Co American Water Works Association , Denver, CO . nference, Orlando, FL. Seidel, C., Gorman, C., Darby, J. & Jensen , V . (2011) An Assessment of the State of Nitrate Treatment Alternatives. Prepared for AWWA by Jacobs Engineering Group, Inc. and University of California, Davis , http://www.awwa.org/files/ r esources/ tecp rojects/ tecreportn itrates finalr evised07272011.pdf Severn Trent Services. (2010) Personal Communication with Dennis Bitter on Brine Destruction. Singh, K.P. (2001) New methods of nitr ate removal from water. , Environmental Pollution Shrimali, M. & 112 359. – , 351 Silverstein, J. (2010) Presentation: Anaerobic biological treatment for removal of inorganic contaminants , Workshop on Biological Drinking Water Treatment, IWA Leading Edge from drinking water . Technology Conference University of Colorado, Boulder , Department of Civil, Environmental, and Architectural Engineering. SMI - PS, Inc. (2009) Sulfur Modified Iron (SMI III®) New Technology – Nitrate Destruction/Arsenic - Removal , http://www.stancounty.com/p ublicworks/pdf/water - summit/sulfur - modified - iron - 12 16 2008.pdf - Soares, M.I.M. (2002) Denitrification of groundwater with elemental sulfur. , 36 , 1392 – Water Research 1395. Soares, M.I.M. (2000) Biological Denitrification of Groundwater. Water, Air, and Soil Pollution , 123 , 183 – 193. State Water Resources Control Board, Division of Water Quality, GAMA Program (2010) Groundwater , http://www.swrcb.ca.gov/gama/docs/coc_nitrate.pdf Information Sheet: Nitrate/Nitrite Sun, D., Yang, J., Li, J., Yu, J., Xu, X., & Yan g, X. (2010) Novel Pd - Cu/bacterial cellulose nanofibers: Preparation and excellent performance in catalytic denitrification. Applied Surface Science , 256 , – 2244. 2241 Tarabara, V. (2007) Coupled Effects of Scaling and Colloidal Deposition on the performanc e of Reverse Osmosis Membrane Filters. National Water Research Institute. 170 Technical Report 6: Drinking Water Treatment for Nitrate

187 Ten States Standards. (2007) , 2007 Edition. Water Supply Recommended Standards for Water Works – blic Committee of the Great Lakes Upper Mississippi River Board of State and Provincial Pu , Health and Environmental Managers. Health Research Inc., Health Education Services Division, Albany, N Y . & Crisp, G. (2009) Potable water HERO process at Yalgoo. Thomson, J., Slunjski, M., Fabig, M., Water Practice and Technology 4 . , aite, M.J. & Hayden, B.E. (2004) Electrocatalytic reduction of nitrate on activated Tucker, P.M., W rhodium electrode surfaces. , 34 , 781 – 796. Journal of Applied Electrochemistry University of Hawaii. (2006) UH R esearcher I ntroduces B reakthrough T echnology for A rsenic - F ree D rinking ater. The University of Hawaii System , W ttp://www.hawaii.edu/news/article.php?aId=1358 h & Upadhyaya, G., Jackson, J., Clancy, T.M., Hyun, S.P., Brown, J., Hayes, K.F., Raskin, L. (2010) Simultaneous removal of nitrate and arsenic from drinking w ater sources utilizing a fixed - bed bioreactor system. Water Research , 44 , 4958 – 4969. U S . EPA ( United States Environmental Protection Agency ) . (N.D.) National Primary Drinking Water . Regulations: Consumer Factsheet on N itrates /N itrites, http://www.epa.gov/ safewater/ pdfs/factsheets/ioc/nitrates.pdf . (1979) Estimating Water Treatment Costs: ) ( EPA United States Environmental Protection Agency U . S . Volume 2 Cost Curves Applicable to 1 to 200 mgd Treatment Plants. . S . EPA ( United States Environmental Protectio U ) . (1998) Small System Compliance Technology n Agency - Microbial Contaminants Regulated Before 1996. Office of Water, September List for the Non - , EPA 815 R 1998 - 98 - 002 , http://www.epa.gov/safewater/standard/tlstnm.pdf U . S . EPA ( United States Environmental P rotection Agency ) . (1994) Drinking Water Treatment for Small Communities: A Focus on EPA’s Research , http://www.p2pays.org/ref/02/01998.pdf ) . (2000) Technologies and Costs for Removal United States Environmental Protection Agency U . S . EPA ( of Arsenic Drinking Water. from U . S . EPA ( United States Environmental Protection Agency ) . (2003) Revised Guidance Manual for Selecting Lead and Copper Control Strategies . Office of Water. S U . ( . EPA United States Environmental Protection Agency ) . (2003 b ) Arsenic Treatment Technology - Evaluation Handbook for Small Systems. EPA 816 R - 03 - 014 , Office of Water. . (2005) Technologies and Costs Document for ) United States Environmental Protection Agency U . S . EPA ( al Stage 2 Disinfectants the Final Long Term 2 Enhanced Surface Water Treatment Rule and Fin and Disinfection Byproducts Rule. 171 Technical Report 6: Drinking Water Treatment for Nitrate

188 U S . EPA ( United States Environmental Protection Agency ) . (2007) Cost Evaluation of Point - of - Use and . - - Entry Treatment Units for Small Systems: Cost Estimating Tool and User Guide. EPA Point of - B - 07 - 001. 815 U . S . EPA ( United States Environmental Protection Agency ) . (2010) Basic Information about Nitrate in Drinking Water , http://water.epa.gov/drink/contaminants/basicinformation/nitrate.cfm - U . EPA ( United States Environmental Protection Agency ) . (201 1) Cost Evaluation of Point S of - Use and . Point - of - Entry Treatment Units for Small Systems: Cost Estimating Tool and User Guide. Cost Estimating Tool , 2011 Update. USGS United States Geological Survey ) . (2011) GAMA ( Groundwater Ambient Monitoring and ( Assessm ent Program ) Priority Basin Project Reports , http://www.swrcb.ca.gov/gama/ report_depot.shtml#priority 21 Water Research , – , 989 Van der Hoek, J.P. & Klapwijk, A. (1987) Nitrate removal from ground water. 997. & Rittmann, B.E. berts, D.J., Lehman, S.G., Adham, S.S., Van Ginkel, S.W., Ahn, C.H., Badruzzaman, M., Ro (2008) Kinetics of nitrate and perchlorate reduction in ion - exchange brine using the membrane biofilm reactor (MBfR). 42 , 4197 – 4205. Water Research , & Reis, M. (2002) Ion exc hange membrane bioreactor for selective removal of Velizarov, S., Crespo, J. Biotechnology nitrate from drinking water: Control of ion fluxes and process performance. , 18 , 296 – Progress 302. g, A.M., Kourakos, Viers, J.H., Liptzin, D., Rosenstock, T.S., Jensen, V.B., Hollander, A.D., McNally, A., Kin N - Hung, A., Dzurella, K. Canada, H.E., Laybourne, S., ., G., Lopez, E.M., De La Mora, N., Fryjoff McKenney, C., Darby, J., Quinn, J.F. & Harter, T. (2012) Nitrogen Sources and Loading to Groundwater. Technical Report 2 in: te in California’s Drinking Water with a Addressing Nitra . Report for the State Water Focus on Tulare Lake Basin and Salinas Valley Groundwater Resources Control Board Report to the Legislature. Center for Watershed Sciences, University of California, Davis. Volokita, M., Balkin, S., Abeliovich, A. & Soares, M.I.M. (1996) Biological denitrification of drinking water using newspaper. Water Research , 30 , 965 – 971. Walker, L.G. (N.D.) Presentation: Nitrate Effects on Public Water System Wells. California Department of alth, Drinking Water Technical Programs Branch , http://www.swrcb.ca.gov/rwqcb5/ Public He water_issues/irrigated_lands/long_term_program_development/15mar09_advisory_wrkgp_inf osession/nitrate_effects_pws_wells.pdf Wang, F. & Tarabara, V.V. (2007) Coupled effects of colloidal deposition and salt concentration 293 polarization on reverse osmosis membrane performance. Journal of Membrane Science , , 123. – 111 Technical Report 6: Drinking Water Treatment for Nitrate 172

189 WA DOH Washington State Department of Health ) . (2005) Guidance Document: Nitrate Treatment ( Water Systems. DOH PUB. #331 309. Alternatives for Small - WA DOH ( Washington State Department of Health ) . (2005b) Guidance Document: Arsenic Treatment for - 210. Small Water Systems. DOH PUB. #331 & Brauer, R. (N.D.) Presentation: Innovative Systems for Tre ating Impaired Waskom, R., Carlson, K., Waters: Natural and Enhanced Purification Processes to Remove Organics and Nutrients. Colorado State University. E - Water Corporation. (2007) Quarterly ewsletter , September : Mainstream , N http://www.watercorporation.com.au/_files/ p ublications r egister/6/ m ainstream_ s eptember07. pdf , Water Corporation. (2009) Desalination in Western Australia www.watercorporation.com.au Water Quality Association. (2011) Gold Seal Certified Product Listings , http://www.wqa.org/sitelogic.cfm?id=1165§ion=3 uality Products. (2003) Nitrate removal by ion exchange. Water Quality Products , 8 . Water Q Webster, T.S., Guarini. W.J. & Wong, H.S. (2009) Fluidized bed bioreactor treatment of perchlorate - laden groundwater to potable standards. ks Association , 101 , Journal of the American Water Wor 137 – 151 . Webster, T.S. & Togna, P. (2009) Final Report: Demonstration of a Full - Scale Fluidized Bed Bioreactor for the Treatment of Perchlorate at Low Concentrations in Groundwater. ESTCP Project ER - 0543. , 0) Nitrate Occurrence Survey Roberson, J.A. (201 https://www.surveymonkey.com Weir, L. & Weir, L. and Roberson, J.A. (2011) National compliance costs from a potential revision to the nitrate regulation.” , 103 (3), 44 - 54. Journal of the American Water Works Association & von Gottberg, A.J.M. (2005) Five Billion Gallons Later – Operating Experience at City of Werner, T.E. – General Electric Company, Water & Process Technologies , Suffolk EDR. GE http://www.gewater.com/pdf/ t echnical%20 p apers_ c ust/ a mericas/ e nglish/ tp 1031 en .pdf Werth, C. ( 2010) Personal C ommunication. ( WHO (2004) Nutrient World Health Organization ). M inerals in D rinking - W ater and the P otential H ealth emineralized and C onsumption of D emineralized and R C A ltered M ineral onsequences of C ontent D rinking - W ater: Consensus of the M eeting , http://www.who.int/water_sanitation_health/ dwq/nutconsensus/en Wisniewski, C., Persin, F., Cherif, T., Sandeaux, R., Grasmick, A. & Gavach, C. (2001) Denitrification of drinking water by the association of an electrodialysis process and a membran e bioreactor: 205. – 139 , 199 F easibility and application. Desalination , 173 Technical Report 6: Drinking Water Treatment for Nitrate

190 & Pan, G. (2009) Rapid and controlled transformation of nitrate in water and brine by Xiong, Z, Zhao, Z. stabilized iron nanoparticles. , 11 , 807 – 819. Journal of Nanoparticle Research & , Shon, Z.H., Lee, G., Moon, B., Noh, B. . Sung, N. (2001) Parametric studies on the performance Y oon, T Korean Journal of Chemical Engineering , 18 , 170 – 177. of anion exchange for nitrate removal. & Yu, J. Kupferle, M.J. (2008) Two - stage sequential electroc hemical treatment of nitrate b rine wastes. 8 385 , – , 379 Water Air and Soil Pollution: Focus 174 Technical Report 6: Drinking Water Treatment for Nitrate

191 10 Appendix 10 .1 Tables of Selected Research Table A. 1 . Selected research on the use of ion exchange (IX) for nitrate removal. Dosing and sorption kinetics of Purolite A 520E (SBA resin). Influent nitrate concentration of 22.6 mg/L as N - Samatya et al. /L). With no sulfate and chloride the treated bed volumes before breakthrough (BV) and resin (100 mg NO 3 ate selective resin Nitr IX – - mg NO (2006) V and 126.4 /g resin, respectively. With sulfate and chloride concentrations 10x capacity were 451 B 3 - /g resin, respectively. that of nitrate, BV and resin capacity were 120 BV and 33.6 mg NO 3 found to be a suitable choice for nitrate removal with influent nitrate concentration Chabani et al. Amberlite IRA 400 resin was IX – Adsorption kinetics - 18 mg NO 4.07 mg/L as N (1 – – /L) and a 96% removal efficiency. of 0.23 (2006) 3 Kim & Benjamin explored. For the sulfate specific resin examined, findings Competition between sulfate and nitrate ions was Sulfate competition IX – (2004) indicate that sulfate selectivity increased with ionic strength. Yoon et al. on nitrate removal. IX – Co - contaminants Compared two different anion exchange resins and studied the influence of other ions (2001) Quaternary amine groups and nitrosamines carcinogenic DBPs, 3 resins examined. With no disinfectants ack of significant (chlorine and chloramine) “release 2 – 10 ng/L”, max of 20 ng/L nitrosamine. “The l IX – exchange treatment system after multiple regeneration cycles indicates Resin residuals and nitrosamine release in full - scale anion - Kemper et al. that releases may eventually subside.” Precursors can be a problem with downstream chloramine use. DBPs (2009) – 100 ng/L (type 1) and 400 ng/L (type 2). Possible problem with IX in POU Upstream disinfection rel eases of 20 with influent containing chlorine/chloramines. atment goal was a decrease in Panglisch et al. 3 - year pilot examined RO, IX and biological denitrification in Mashhad, Iran. Tre (2005) nitrate concentration from 26 mg/L as N (115 mg/L as nitrate) to 9 mg/L as N (40 mg/L as nitrate). Raw water Dördelmann et with electrical conductivity of 1550 uS/cm, pH of 7.2. Optimal treatment options were deemed to be biological treatment consisted of cartridge filtration. Two IX columns denit - al. (2006, and rification and RO. In the IX pilot plant pre IX – Pilot study N.D. ) - current regeneration. System containing ~200 L of nitrate selective resins were run in parallel with counter characteristics: Dördelmann 3 /h, Specific Flow , Bed Volume (BV): 200 L, Bed Depth: 1.6 m, Flow Rate: 1.0 to 3.0 m Column diameter: 0.4 m (2009) Rate: 5 to 15 BV/h, Resin replacement every 5 years. Johnston & Heijnen (N.D.). and See Also Clifford & Weber (1978) , Gute r (1982) , Clifford (1987) , Rash (1992) , Clifford (2007), 175 Technical Report 6: Drinking Water Treatment for Nitrate

192 . Table A. 2 Selected research on the use of reverse osmosis (RO) for nitrate removal. on, CO: Use of RO in Brighton, CO, for potable water treatment including nitrate removal began operating in 1993 (11000 sq Bright er ft., 4 MGD, total cost of $8,253,000) after the performance of preliminary pilot study (4 membranes were tested). Source wat - (with TDS 800 te concentration of 13 to 23 mg/L nitrate - N or 58 to 102 mg/L nitrate as NO – nitra – 1140 mg/L, hardness 370 480 3 – RO ). Problems with biofouling of membrane and cartridge filters (slime forming Pilot Study and - ) minimized with anti Pseudomonas mg/L as CaCO 3 Cevaal et al. (1995) - scale Installation - scalant use. Pretreatment: acid, anti Full scalant and filtration. Post - treatment: CO on, caustic and zinc stripping (degasifier), disinfect i 2 orthophosphate (ZOP) addition. Blending ratio (untreated%:treated%) 20:80(winter) and 60:40 (summer). Waste discharge to , pressure rface water. Additional characteristics: Flux rate < 14.2 gal/sq. ft./day (helps control fouling), 2 stages: 32 : 16 vessels su of 231 psi, each vessel with 6 8” diameter membranes. Cleaning required every 2 months. efficiency reverse osmosis (HERO). Water recovery rates as high as 95%. - online in 2007 using high Yalgoo, Australia: RO plant, in RO HERO – Water Corporation More than 85% waste reduction (a low as 10% the waste of conventional RO.) Removes both nitrate and silica from brackish (2009) Australia water. Inland Empire, CA: Cow Power in Inland Empire, CA: $80 mil. effort is powered by methane gas produced from the high populatio scale uses n RO – Full - Black (2003) th of cows in that region. Online since 2002, this plant provides 1/5 methane of regional demand. - year pilot examined RO, IX and BD in Mashhad, Iran. Treatment goal of 9 mg/L as N (40 mg/L as nitrate) (influent Mashhad, Iran: 3 Panglisch et al. s nitrate). Results indicate BR and RO were best choices. Automated RO pilot plant with a nitrate of 26 mg/L as N, 115 mg/L a (2005) 3 “capacity [of] 3 m - treatment: acid, anti - /h (RO permeate).” Pre scalant, automated filtration (50 um), and cartridge filtration Dördelmann et al. RO – Pilot study e, 2 parallel membranes, second stage, concentrate from stage one to second membrane before (1um). 2 stages: in the first stag ) N.D. (2006 and comparing IX, RO mixing with stage 1 permeate. Post - treatment: Blending, CO2 removal. New membranes every 5 years. RO was least expensive Dördelmann and BD 3 Lower operator demand than BD. RO . of DW), due to low cost of electricity .6 kWh/m despite having highest energy demands (0 (2009) f produced better quality water (lower TDS, Ca, Mg, Na, K, Cl, nitrate, sulfate and bicarbonate), but the greatest percentage o waste. 60 mg/L as nitrate), target of 9 mg/L as N (40 mg/L as nitrate) with waste Milan, Italy: Influent of 11 – 14 mg/L as N (50 – 3 scale stage RO plants (7 to 58 m - /hr permeate), RO – Elyanow & Full - concentrate < 30 mg/L as N (132 mg/L as nitrate) (for sew er disposal). Series of 13 1 Persechino, (2005) with blending 77 – installations 88% water recovery (note the disposal limitation). Pretreatment: anti - scalant (2.0 – 3.5 mg/L). Cleaning only hs. every year and a half (SDI<1) and requires only 3 tec for salt Alternative RO – Alternatives to conventional disposal measures of RO waste brine, including reuse for industrial processes, processing (e.g., Howe (2004) disposal options production), or use in energy generation (“solar br ine pond”). p Analysis of complex source water, membrane fouling and ways of anticipating changes in flux and rejection rates. Relationshi between constituents and how RO treatment is af fected by colloids, silica, concentration polarization (higher salt concentration near membrane surface), back diffusion, and scaling. Typical colloidal constituents: “ Membrane Tarabara (2007) sulfur, silica, and ferric and aluminum RO – scaling and fouling, hydroxides.” Results indicate colloidal fouling may amplify other scaling/fouling factors. “By recognizing possible interferences Wang & Tarabara (2007) between rejected salts and colloids deposited on the membrane surface, the work explored the phenomenon of coupling between colloidal interaction tion of RO membranes” (Tarabara nd scaling as main factors limiting the applica colloidal fouling, concentration polarization, a 2007). RO Gabelich et al. Investigation of the interaction of upstream residuals on RO treatment (membrane and additiv e chemicals). Influence of coagulant – Influence of upstream processes (2004) residuals on colloidal fouling, and disinfectants on membrane oxidation. 176 Technical Report 6: Drinking Water Treatment for Nitrate

193 3 . Selected research on the use of electrodialysis/electrodialysis reversal for nitrate removal. Table A. 3 – 23 mg/L as N (80 – /h capacity. Influent nitrate concentration 18 100 mg/L as Austria: 2 - year pilot starting in 1990 with 1 m nitrate), design for 36 mg/L as N (160 mg/L as nitrate). Plannin g began in 1996 for full - scale installation with seasonal operation ED in commencing in 1997. Disposal options: sewer or irrigation reuse. “Monovalent selective anion exchange membranes.” 3 stacks Pilot study and – Hell et al. (1998) 3 3 stages offer capacity from 48 to 144 m parallel – Full scale installation /h. Co mplete automation. Effluent nitrate of 9 mg/L as N (40 mg/L as nitrate) - nitrate (capable of product water nitrate of 5.7 mg/L as N, 25 mg/L as nitrate) and 23% hardness reduction. Nitrate selectivity: 66% removal with 25% desalination. Morocco: Pilot study using a nitrate selective membrane for nitrate removal from an influent level of 20 mg/L as N (90 mg/L a s Pilot study – ED Midaoui et al. – nitrate) to acceptable levels, in water with a TDS concentration of 800 mg/L. Analysis of oper ating parameters to minimize Optimization precipitation, scaling and associated chemical use. Variation of voltage, flow rate and temperature. Ion removal increased with (2002) increasing temperature. Delaware: EDR has be en successfully employed for the removal of nitrate from potable water in Delaware. The nitrate Full - EDR Prato & Parent – concentration in treated potable water was 1 mg/L as N (4.4 mg/L as nitrate), just 7.5% of the 13.5 mg/L as N (60 mg/L as nit rate) scale (1993) influent nitrate concentrati on. 3 stage system, 88% demineralization (TDS reduction), 90% water recovery, and pH decrease from installation 6.2 to 5.4. for nitrate. Limited to no anti scalant - Barcelona, Spain: GE EDR plant with 50 MGD capacity (260,000 households), not specifically use. “Compared to a typical RO treatment facility producing 3.8 million gallons of water per day, GE’s EDR technology, opera ting at EDR Full – scale GE (2010), GE - 83% efficiency, is designed to eliminate the need for over 28,000 pounds of anti s - calant, reducing operating costs by > $100,000 (N.D.) installation per year at typical 2008 chemical prices.” *Previous EDR systems for nitrate removal were installed in Arizona, Delaware, Japan, Italy, Bermuda, and Israel. Full - scale EDR – scale EDR plants in Italy, France, Switzerland, Netherlands, and Austria. Capacity Europe: As of ~2005, Ameridia/Eurodia full - Ameridia/Eurodia - 0.925 MGD and water recovery from 93 – 98%. Information on full ranging from 0.032 scale installations, pilot studies and costs. – ons installati – Multiple ED nd Research investigation using synthetic waters to determine the impact of organic matter in the removal of nitrate, fluoride a Banasiak & Schäfer ough nitrate removal was the least affected, due to smaller “hydrated ionic radius.” boron. Fouling led to decreased flux, alth contaminant removal, role of Removal of boron and fluoride was enhanced by the presence of organic matter, while nitrate removal was not enhanced and (2009) simply decreased over time with the decrease in me mbrane flux. organic matter Research investigation using synthetic waters to assess nitrate removal under different operating conditions. The impact of Nataraj et al. ies, including nitrate. Alternating anion and cation different voltage (from 40 to 50 V) was examined across several ionic spec Pilot study ED – (2006) exchange membranes were used. Results indicate 94% nitrate removal, with a reduction in the removal rate at 50 V due to back diffusion and fouling. Morocco: Nitrate removal from brackish water. Comparison of ED using a monovalent membrane and adsorption on chitosan. ED – Pilot study ED comparison wi th Sahli et al. (2008) successfully removed nitrate. Adsorption can remove nitrate, but not likely feasible. Adsorption can be used to remo ve nitrate from ED waste concentrate. Highlights concerns regarding waste concentrate disposal from ED. adsorption Dördelmann Mashhad, Iran: Comparison of IX, RO, BD, and ED. ED pilot study was started in 2007. (2009) Technical Report 6: Drinking Water Treatment for Nitrate 177

194 Table A. . Selected research on the use of biological denitrification for nitrate removal from potable water. 4 - Multiple Biological Configurations Pilot Testing The City of Glendale, AZ, has investigated three configurations of biological treatme nt to address high nitrate levels in groundwater wells (Meyer et al. 2010). An autotrophic MBfR using was compared with two heterotrophic fixed bed bioreactors each with different media (plastic versus granular activated carbon). were used as electron donor for the autotrophic system and the heterotrophic systems, respectively. Post - Hydrogen gas and ethanol treatment included filtration using biologically activated carbon and ozonation. The fixed bed bioreactor with plastic media and the MBfR perfo well, with product water rmed - criteria analysis found the MBfR to be most favorable regarding benefits, but the least favorable meeting or exceeding potable water standards. Multi astic media had the lowest life cycle cost. The MBfR costs were greatest. regarding costs. Including comparison with IX, the fixed bed bioreactor with pl An investigation of biological treatment options in the City of Thornton, CO, funded by the WaterRF, examined two packed bed bioreactors and a moving bed ™ 2010, Project # 4202) to address nitrate impacted source water. Nitrate levels were successfully decreased by each of ) (C ity of Thornton biofilm reactor (MBBR the three pilot systems from an influent concentration of 10 mg/L nitrate as N to an effluent concentration of < 2 mg/L n itrate as N. Operation at high and low strate dose optimization in temperatures was tested with examination of seeding for low temperatures. The study highlights the need for nutrient and sub biological treatment systems. Substrates Numerous alternative substrate options have been explored in the literature including newspapers, vegetable oil, cotto n, and formate (Volokita et al. 1996; Hunter 2001; Killingstad et al. 2002 ; and Della Rocca et al. 2006 .) Fixed Bed See Riverside, CA Case Study (Carollo En gineers 2008). A fixed bed heterotrophic denitrification pilot study was implemented in Mashhad, Iran by Dördelmann et al. (2006) using two parallel fixed beds containing expanded clay media. Acetic acid and ferrous sulfate served as the electron donor a nd nutrient supply, respectively. Post treatment consisted of “aeration, dual media and activated carbon 2006). Influent nitrate levels of 26 mg/L as N (115 mg/L as nitrate) were decreased to < 9 mg/L filtration” (Dördelmann et al. 3 3 as N (40 mg/L as nitra 7 kg NO ). Used for flushing /m d (0.43 lb NO and Dördelmann et al. 2006 /ft ~ d) (Panglisch et al. 2005 te) with a nitrate reduction rate of 3 3 and backwashing, 7% of influent volume was wasted. In practice, a final disinfection step would be required (Dörd elmann et al. N.D). An up - flow, fixed - bed, autotrophic, lab - scale system, using granular sulfur as both substrate and growth surface was explored by Soares (2002). Operated over 3 3 d) was achieved with a one hour hydraulic retention time and a loading rate of a 5 month period, a maximum denitrification rate of 0.2 kg N/m d (0.012 lb N /ft 3 3 0.24 kg N/m d (0.015 lb N/ft d). Sulfur based autotrophic systems would not be appropriate for the treatment of feed waters high in sulfate. Aslan (2005) examined a lab - sc ale packed sand bed system, with ethanol as substrate for the simultaneous removal of nitrate and several pesticides. After 3 days for biofilm development, 93 – 98% nitrate removal was achieved requiring at least a 2 hour residence time. Pesticide remova l required longer residence times (up to 12 hours) for efficient removal. Upadhyaya et al. (2010) investigated the use of a fixed bed biological reactor with granular activated carbon media for the removal of nitrate and arsenic at the - same time. The med ia was biologically activated from use in a separate bioreactor for the removal of nitrate and perchlorate. Reactors were th us biologically active carbon (BAC) reactors. With acetic acid as the substrate, two in series BAC reactors were used to treat syn th etic groundwater. Arsenic levels were reduced from 200 μg/L arsenic in the influent to 20 μg/L in the effluent (still above the arsenic MCL of 10 μg/L) while nitrate levels were decreased from 11 mg/L as N (50 mg/L as nitrate) in the influent to less than 0.045 mg/L as N (0.2 mg/L as nitrate) in the effluent. Used as an electron acceptor by microbes in the sulfide (solids) and also with oxidation of substrate, nitrate was reduced to nitrogen gas. Arsenic was removed from solution with the formation of arsenic adsorption and “surface precipitation on iron sulfides (p. 4958).” 178 Technical Report 6: Drinking Water Treatment for Nitrate

195 Fluidized Bed See Rialto, CA Case Study (Webster & Togna 2009). Kurt et al. (1987) investigated an autotrophic fluidized sand bed reactor using hydrogen gas as substrate. With an influent nitrate concentration of 25 mg/L nitrate le stage reactors to - nitrogen, a maximum nitrate reduction rate of 5 mg/L per hour was attained using a mixed culture. The authors propose multip address the problem of partial denitrification. mix of propionic acid and ethanol in a heterotrophic fluidized sand bed reactor, Holló & Czakó (1987) examined denitrificatio n at the lab - and pilot - scale. Using a - treatment consisted of cartridge filtration, gas exchange, sand filtration, carbon filtration and disinfection. “Nitrate removal capacity of the reactor was 50 Post - 3 - o – /m – /day (3.1 3.7 lb NO 1987). 60 kg NO /gal/d), which could be maintained permanently at temperatures as low as 8 – 10 C as well” (Holló & Czakó 3 3 - Diffusive Extraction and Microp MBR orous Membranes (See also Pilot Testing above) Mansell & Schroeder (2002) assessed hydrogenotrophic denitrification at the lab - scale using a membrane bioreactor (MBR) with a 0.02 micron microporous membrane through which nitrate diffuses to the biological compartment. The membrane prevents mixing of microbes with the water being treated and no carbon substrate was necessary because hydrogen gas was supplied as the electron donor for autotrophic denitrification. Prev ious issues regarding the transfer of h ydrogen gas to the water and safety concerns due to explosive nature of hydrogen gas have been addressed with the development of “membrane dissolution systems” (Mansell & Schroeder 2002). Hydrogen gas was delivered to the biomass with silicone tubing. Re sults indicated reduction of nitrate levels − − 96% removal. Measured HPC indicated minimal – from a maximum of 40 mg/L NO N in the treated water, with 92 - N in the feed water to 3.2 mg/L NO - 3 3 biomass transfer to the treated water compartment. Ergas & Rhein h tubular heimer (2004) studied denitrification of potable water using a membrane bioreactor (MBR) in which feed water is passed throug acrylonitrile membranes, nitrate diffuses through the membrane and denitrification occurs on the exterior membrane surfac e. The mean transfer to the biofilm 2 2 was 6.1 g NO N/ft - N/m d (0.6 g NO ). The ultimate methanol (substrate) loading rate of 1.1 g/d resulted in a mean concentration of nitrate in the treated - 3 3 - N/L. A mathematical model of ni potable water of 5.2 mg NO - trate mass transfer was developed. A removal efficiency of 99% was achieved with a starting 3 - concentration of 200 mg NO - N/L. Use of the MBR allows for denitrification with separation of the water to be treated and biological treatment, thereby 3 /L) of the methanol post treatment removal of biomass and dissolved organics. The effluent would require additional treatment, because 8% (30 mg avoiding fer to the effluent stream. crossed the membrane; the authors suggest that further development of the biomass could minimize methanol trans Hollow Fiber Membranes MBR - Gaseous Substrate Delivery – Chung et al. (2007) explored the use of autotrophic denitrification for the treatment of highly concentrated waste from nitra te removal via anion exchange. be significant Using a hyd rogen - based, hollow fiber membrane biofilm reactor, the impact of brine concentration (up to 15%) on nitrate reduction was found to due to microbial inhibition. In the reduction of nitrate, use of hydrogen gas rather than an organic substra te offers an inexpensive alternative for potable water the use of carbon substrates . treatment systems. Biomass production is decreased and there is no need to remove substrate residual, as there would be with ptor, hollow fiber MBfRs with hydrogen gas as electron donor can effectively decrease the levels of multiple Using nitrate as the primary electron acce contaminants including perchlorate, chromate an d arsenate (Nerenberg & Rittman 2004). Low levels of these oxidized species without a primary elect ron acceptor can limit biological reduction; however, in the presence of nitrate, reduction can occur. When the concentrations o f nitrate and the oxidized species of oval was achieved while removal of perchlorate, chromate and were 1.1 mg/L as N (5 mg/L as nitrate) and 1 mg/L, respectively, 99% nitrate rem st intere arsenate reached 36%, >75%, and >50%, respectively. Technical Report 6: Drinking Water Treatment for Nitrate 179

196 Bioelectrochemical Denitrification ors (BERs). In a BER, hydrogenotrophic denitrification occurs Ghafari et al. (2008) provide a review of biological denitrification with a focus on bioelectrical react as hydrogen gas is produced at the cathode and utilized as the electron donor by denitrifiers, while nitrate is reduced to ni trogen gas. Previous BER research is c and heterotrophic examples across a range of nitrate levels and generally in synthetic waters. With additional research, B ERs discussed including autotrophi may become a feasible alternative for nitrate removal from drinking water. In situ Denitrification Permeable reactive barrie an rs (PRBs) can be used to directly treat nitrate contaminated groundwater. Hunter (2001) examined the use of vegetable oil as electron donor in biological denitrification. The use of an insoluble substrate minimized biomass blockage, a problem common w ith the use of soluble substrates like ethanol, methanol and acetate. The barrier was composed of soybean oil - coated sand and effectively decreased the nitrate levels from a eeks, with a flow rate 1100 L/week. After 15 weeks, insufficient oil remained starting concentration of 20 mg/L nitrate - N to below the MCL for a period of 15 w for denitrification. High chemical oxygen demand, TSS and turbidity in the effluent of the reactor indicate a longer sand be d was needed; however, the author in s om the barrier, application of this type of biological reactor would decrease these factors naturally. With a withdrawal point far enough fr suggests that itu encountered in this study was the subsequent potable water treatment requirements would be limited to disinfection. The most significant problem exhaustion of substrate. An effective means of substrate addition must be found (injection for example), but this was not ex plored. The estimated life of the PRB was 2.5 – 12.5 years depending on several key factors inc luding flow, nitrate concentration and dissolved oxygen concentration. With hydrogen gas as the substrate in autotrophic denitrification, Haugen et al. (2002) examined hydrogenotrophic denitrifica tion in a lab scale experiment intended to imitate in situ treatment. Denitrification kinetics, the feasibility and longevity of substrate delivery via tubular membranes and post - treatment water quality were investigated. Delivery of hydrogen gas through tubular membranes minimized the risks associated with util ization of this - - - flammable/explosive gas. The reactor was tested over 155 days. An initial influent nitrate concentration of 8.2 mg NO N/L was doubled to 16.4 mg NO - N/L. 3 3 2 e tubular membrane bioreactor. A denitrification rate of 169 mg N/h/m After adjustment of parameters, complete nitrate removal was achieved using th (membrane surface area) was attained with a hydrogen gas pressure of 1.44 atm (lower pressures resulted in incomplete reducti on. The greatest hydrogen gas 2 - 2 transfer across the membrane s. The simulated groundwater velocity was 0.3 m/d resulting in 14 minutes of membrane contact (flux) was 1.79 x 10 /m mg H 2 for buffer in the current time. Additional considerations for application of this treatment method include: the lower temperature of groundwater, the need study, the depth limitation, nutrient requirements, and the difference between aquarium rocks and subsurface porous media. I ntermediate denitrification products and end products (ammonia and nitrogen gas) were not measured in th is study; however, the authors suggest the high nitrate to substrate ratio would result in reduction to nitrogen gas. Schnobrich et al. (2007) simulated in situ nitrate removal via hydrogenotrophic Denitrification. With hydrogen gas delivery through a me mbrane module influence of pH, nutrient consisting of a fiberglass membrane wound in a spiral fashion and attached to polyethylene membranes. The study examined the ate conditions, the porous media was extracted from an aquifer requirements and the feasibility of appropriate levels of hydrogen gas delivery. To simul in situ o flow columns were operated in series, only the first of which included a membrane fed with hydrogen gas. and the system was operated at 10 C. Two up - Including that required for the re duction of dissolved oxygen, the total concentration of hydrogen gas required for complete denitrification was 11.2 mg H /L. 2 xpected naturally. be e Overall, this study demonstrated effective substrate delivery and nitrate removal under conditions more similar to what would Technical Report 6: Drinking Water Treatment for Nitrate 180

197 . Selected research on the use of chemical denitrification (CD) for nitrate removal. 5 Table A. ow a pH of Reduction of nitrate to ammonia using ZVI powder was highly pH dependent with optimal kinetics bel Huang et al. 2 - - /mol NO CD using ZVI for complete reduction within 1 hour. 50 mg NO /L, 4. The minimum ratio of iron to nitrate was 120 m 3 3 (1998) 100% removal. Xiong et al. ontrolled by the iron to Found that the end product of denitrification (nitrogen gas versus ammonium) could be c – ZVI CD (2009) nitrate ratio and the use of catalysts. Choe et al. In the corrosion of ZVI, the formation of “green rusts” and “suspended green particles” is associated with CD – ZVI stabilization of pH and steady decrease in nitrate. (2004) Examined the nitrate reduction rates of three types of iron. Findings indicate that rate increases with decreasing Alowitz & CD – ZVI Scherer (2002) pH. articles. High temperature exposure to Nitrate reduction by ZVI can be optimized through pretreatment of iron p Liou et al. CD ZVI – hydrogen gas and deposition of copper were explored separately as options for pretreatment of the iron surface. (2005) pact on nitrate removal; however, Examined chloride as a potential competitor. Results indicate a minimal im Moore & Young ZVI – CD other competing ions could be important regarding both competition for adsorption sites and reduction. (2005) for the removal of based granular media that has been commercially developed - III® is a patented, iron “SMI - - DSWA and City nitrate, co contaminants including uranium, vanadium and chromium, and other compounds from water. It is SMI CD - of Ripon (2010) foreseen that the greatest benefit of this technology is that it does not produce a costly brine stream as do the currently accept ed nitrate removal technologies of ion exchange and reverse osmosis.” – CD Reddy & Lin 2000; Pintar et al. 2001; Gavagnin et al. 2002; Lemaignen et al. 2002; Pirkanniemi & Sil lanpaa 2002; Chen et al. 2003; Catalytic 2010. 2007; and Sun et al. Denitrification Palomares et al. 2003; Pintar 2003; Constantinou et al. 181 Technical Report 6: Drinking Water Treatment for Nitrate

198 A. 6 Table . Advantages and disadvantages of the five major treatment options for nitrate removal. Advantages Disadvantages  The disposal of waste brine,  Years of industry experience, The potential for nitrate dumping specifically  for Multiple contaminant removal,  selective resin use for high sulfate waters, - non Selective nitrate removal,   The need to address resin susceptibility to ange Ion Exch Financial feasibility,  hardness, iron, manganese, suspended solids, Use in small and large systems,  organic matter, and chlorine, and and  The possible role of resin residuals in DBP The ability to automate.  formation. The disposal of waste concentrate,   uality product water, High q  Typically high capital and O&M costs,  Multiple contaminant removal, ddress membrane susceptibility to The need to a   Desalination (TDS removal), Reverse hardness, iron, manganese, suspended solids,  Feasible automation, Osmosis silica, organic matter, and chlorine, Small footprint, and  High energy demands, and   Application for small and POU  The lack of control over target constituents applications. (complete demineralization).  Limited to no chemical usage,  The disposal of waste concentrate, Long lasting membranes,   The need to address membrane susceptibility to  Selective removal of target hardness, iron, manganese, and suspended solids, species, High maintenance demands,  Electrodialysis/ Flexibility in removal rate through   Costs (comparable to RO systems), Electrodialysis voltage control, Rever sal  The need to vent gaseous byproducts,  Better water recovery (lower ion with high recovery, The potential for precipitat  waster volume),  High system complexity, and  Feasible automation, and  Dependence on conductivity.  Multiple contaminant removal.  The need for substrate and nutrient addition, High monitoring needs,   Significant post - treatment requirements, High water recovery,  High capital costs,  No brine or concentrate waste  Sensitivity to environm ental conditions  stream (nitrate reduction rather (sometimes), than removal to waste stream), Biological Large system footprint (sometimes),   Low sludge waste, Denitrification  High system complexity (sometimes, can be ive operation, Less expens  comparable to RO),  Limited chemical input, scale systems in the U.S., - Lack of full  Increased sustainability, and   The possibility of partial denitrification, Multiple contaminant removal.  Permitting and piloting requirements, and   - S lower initial start up, which could cause challenges for wells with intermittent run time. The potential reduction of nitrate beyond nitrogen  gas to ammonia,  Conversion of nitrate to other  The possibility of partial denitrification, nitrogen species (no brine or  The possible dependence of performance on pH concentrate waste stream), and t emperature, The potential for more sustainable  Chemical  The possible need for iron removal, and Denitrification treatment,  scale chemical denitrification - The lack of full High  water recovery (higher than systems resulting in: ® RO according to Cleanit - LC), and o Unknown reliability,  Multiple contaminant removal. Unknown costs, and o Unknown operational complications. o 182 Technical Report 6: Drinking Water Treatment for Nitrate

Related documents